Cellular Image Classification 1st Edition

Xiang Xu
Visit to download the full and correct content document:
https://textbookfull.com/product/cellular-image-classification-1st-edition-xiang-xu/
More products digital (pdf, epub, mobi) instant
download maybe you interests ...

Fuzzy Machine Learning Algorithms for Remote Sensing

Image Classification 1st Edition Anil Kumar

https://textbookfull.com/product/fuzzy-machine-learning-
algorithms-for-remote-sensing-image-classification-1st-edition-
anil-kumar/

Image analysis, classification and change detection in

remote sensing: with algorithms for Python Fourth
Edition Canty

https://textbookfull.com/product/image-analysis-classification-
and-change-detection-in-remote-sensing-with-algorithms-for-
python-fourth-edition-canty/

Image Analysis, Classification, and Change Detection in

Remote Sensing: With Algorithms for Python 4th Edition
Morton John Canty

https://textbookfull.com/product/image-analysis-classification-
and-change-detection-in-remote-sensing-with-algorithms-for-
python-4th-edition-morton-john-canty/

Digital Audio Watermarking Fundamentals Techniques and

Challenges 1st Edition Yong Xiang

https://textbookfull.com/product/digital-audio-watermarking-
fundamentals-techniques-and-challenges-1st-edition-yong-xiang/
The Zebrafish Cellular and Developmental Biology Part A
Cellular Biology 4th Edition H. William Detrich

https://textbookfull.com/product/the-zebrafish-cellular-and-
developmental-biology-part-a-cellular-biology-4th-edition-h-
william-detrich/

Analytics in Smart Tourism Design Concepts and Methods

1st Edition Zheng Xiang

https://textbookfull.com/product/analytics-in-smart-tourism-
design-concepts-and-methods-1st-edition-zheng-xiang/

Technological Economics Shoubo Xu

https://textbookfull.com/product/technological-economics-shoubo-
xu/

WHO Classification of Tumours 5th Edition Digestive

System Tumours Who Classification Of Tumours Editorial
Board

https://textbookfull.com/product/who-classification-of-
tumours-5th-edition-digestive-system-tumours-who-classification-
of-tumours-editorial-board/

Proceedings of the Third International Forum on

Decision Sciences 1st Edition Xiang Li

https://textbookfull.com/product/proceedings-of-the-third-
international-forum-on-decision-sciences-1st-edition-xiang-li/
Xiang Xu · Xingkun Wu
Feng Lin

Cellular
Image
Classification
Cellular Image Classification
Xiang Xu Xingkun Wu
•

Feng Lin

Cellular Image Classification

123
Xiang Xu Feng Lin
School of Computer Engineering School of Computer Engineering
Nanyang Technological University Nanyang Technological University
Singapore Singapore

Xingkun Wu
Zhejiiang University
Hangzhou, Zhejiang
China

ISBN 978-3-319-47628-5 ISBN 978-3-319-47629-2 (eBook)

DOI 10.1007/978-3-319-47629-2
Library of Congress Control Number: 2016955784

© Springer International Publishing AG 2017

This work is subject to copyright. All rights are reserved by the Publisher, whether the whole or part
of the material is concerned, specifically the rights of translation, reprinting, reuse of illustrations,
recitation, broadcasting, reproduction on microfilms or in any other physical way, and transmission
or information storage and retrieval, electronic adaptation, computer software, or by similar or dissimilar
methodology now known or hereafter developed.
The use of general descriptive names, registered names, trademarks, service marks, etc. in this
publication does not imply, even in the absence of a specific statement, that such names are exempt from
the relevant protective laws and regulations and therefore free for general use.
The publisher, the authors and the editors are safe to assume that the advice and information in this
book are believed to be true and accurate at the date of publication. Neither the publisher nor the
authors or the editors give a warranty, express or implied, with respect to the material contained herein or
for any errors or omissions that may have been made.

Printed on acid-free paper

This Springer imprint is published by Springer Nature

The registered company is Springer International Publishing AG
The registered company address is: Gewerbestrasse 11, 6330 Cham, Switzerland
Preface

This book introduces new techniques for cellular image feature extraction, pattern
recognition and classification. We use the antinuclear antibodies (ANAs) in patient
serum as the subjects and the indirect immunofluorescence (IIF) technique as the
imaging protocol to illustrate the applications of the described methods. Throughout
the book, we provide evaluations for our proposed methods on two publicly
available human epithelial (HEp-2) cell datasets: ICPR2012 dataset from the
ICPR’12 HEp-2 cell classification contest and ICIP2013 training dataset from the
ICIP’13 Competition on cells classification by fluorescent image analysis.
First, the reading of imaging results is significantly influenced by one’s quali-
fication and reading systems, causing high intra- and inter-laboratory variance. We
describe a low-order LP21 fiber mode for optical single cell manipulation and
imaging staining patterns of HEp-2 cells. A focused four-lobed mode distribution is
stable and effective in optical tweezer applications, including selective cell pick-up,
pairing, grouping or separation, as well as rotation of cell dimers and clusters. Both
translational dragging force and rotational torque in the experiments are in good
accordance with our theoretical model. With a simple all-fiber configuration, and
low peak irradiation to targeted cells, instrumentation of our optical chuck tech-
nology will provide a powerful tool in the ANA-IIF laboratories. We focus on the
optical, mechanical and computing systems for the clinical trials. Computer pro-
grams for GUI and control of the optical tweezers are also discussed.
Next, we introduce the Bag-of-Words (BoW) framework which is one of the
most successful image representations. To reduce the inevitable information loss
caused by coding process, we study a linear local distance coding (LLDC) method.
The LLDC method transforms original local feature to more discriminative local
distance vector by searching for local neighbors of the local feature in the
class-specific manifolds. Then we encode and pool the local distance vectors to get
salient image representation. Combined with the traditional coding methods, this
method achieves a higher classification accuracy.
Then, we study a rotation invariant textural feature of pairwise local ternary
patterns with spatial rotation invariant (PLTP-SRI). It is invariant to image rota-
tions, meanwhile it is robust to noise and weak illumination. By adding spatial

v
vi Preface

pyramid structure, this method captures spatial layout information. While the
proposed PLTP-SRI feature extracts local feature, the BoW framework builds a
global image representation. It is reasonable to combine them together to achieve
impressive classification performance, as the combined feature takes the advantages
of the two kinds of features in different aspects.
Finally, we design a co-occurrence differential texton (CoDT) feature to repre-
sent the local image patches of HEp-2 cells. The CoDT feature reduces the infor-
mation loss by ignoring the quantization while it utilizes the spatial relations among
the differential micro-texton feature. Thus it can increase the discriminative power.
We build a generative model to adaptively characterize the CoDT feature space
of the training data. Furthermore, we exploit a discriminant representation for the
HEp-2 cell images based on the adaptive partitioned feature space. Therefore, the
resulting representation is adapted to the classification task. By cooperating with
linear support vector machine (SVM) classifier, our framework can exploit the
advantages of both generative and discriminative approaches for cellular image
classification.
The monograph is written for those researchers who would like to develop their
own programs, and the working MATLAB codes are included for all the important
algorithms presented. It can also be used as a reference book for graduate students
and senior undergraduates in the area of biomedical imaging, image feature
extraction, pattern recognition and classification. Academics, researchers, and many
others will find this to be an exceptional resource.
Enjoy the read.

Singapore Xiang Xu
China Xingkun Wu
Singapore Feng Lin
August 2016
Contents

1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .... 1
1.1 Background . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .... 1
1.1.1 Clinical Problems: A Case Study on Autoimmune
Diseases . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .... 1
1.1.2 Cellular Imaging: A Case Study on Indirect
Immunofluorescence . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3
1.2 Computer-Aided Diagnosis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6
1.3 Experimental Datasets in the Book . . . . . . . . . . . . . . . . . . . . . . . . . 8
1.3.1 The ICPR2012 Dataset . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8
1.3.2 The ICIP2013 Training Dataset . . . . . . . . . . . . . . . . . . . . . . 10
1.4 Structure of the Chapters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12
2 Fundamentals . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15
2.1 Optical Systems for Cellular Imaging . . . . . . . . . . . . . . . . . . . . . . . 15
2.1.1 Laser Scanning Confocal Microscope . . . . . . . . . . . . . . . . . 16
2.1.2 Multi-photon Fluorescence Imaging . . . . . . . . . . . . . . . . . . 20
2.1.3 Total Internal Reflection Fluorescence Microscope . . . . . . . 22
2.1.4 Near-Field Scanning Optical Microscopy Imaging
Technology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25
2.1.5 Optical Coherence Tomography Technology. . . . . . . . . . . . 29
2.2 Feature Extraction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31
2.2.1 Low-Level Features. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31
2.2.2 Mid-Level Features . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 38
2.3 Classification . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 39
2.3.1 Support Vector Machine . . . . . . . . . . . . . . . . . . . . . . . . . . . 39
2.3.2 Nearest Neighbor Classifier . . . . . . . . . . . . . . . . . . . . . . . . . 40
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 41

vii
viii Contents

3 Optical Systems for Cellular Imaging . . . . . . . . . . . . . . . . . . . . . . . . . 45

3.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 46
3.2 Optical Tweezer . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 47
3.2.1 Introduction to Optical Tweezers . . . . . . . . . . . . . . . . . . . . 47
3.2.2 Gradient and Scattering Force of Optical Tweezers. . . . . . . 48
3.2.3 Three-Dimensional Optical Trap . . . . . . . . . . . . . . . . . . . . . 49
3.3 Low-Order Fiber Mode LP21 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 51
3.3.1 Fiber Mode Coupling Theory . . . . . . . . . . . . . . . . . . . . . . . 51
3.3.2 Analysis of Field Distribution in Optical Fiber . . . . . . . . . . 53
3.3.3 Solution to LP21 Mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . 55
3.3.4 Selective Excitation of LP21 Mode . . . . . . . . . . . . . . . . . . . 56
3.3.5 The Twisting and Bending Characteristics
of LP21 Mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 58
3.3.6 Why LP21 Mode? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 60
3.4 Optical Tweezer Using Focused LP21 Mode . . . . . . . . . . . . . . . . . . 61
3.4.1 Fiber Axicons . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 61
3.4.2 Cell Manipulation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 66
3.5 Modeling of Optical Trapping Force . . . . . . . . . . . . . . . . . . . . . . . 68
3.5.1 Force Analysis of Mie Particles in Optical Trap . . . . . . . . . 69
3.5.2 Gaussian Beam . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 72
3.5.3 Simulation of Light Force on Mie Particle . . . . . . . . . . . . . 73
3.6 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 77
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 78
4 Image Representation with Bag-of-Words . . . . . . . . . . . . . . . . . . . . . . 81
4.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 81
4.2 Coding . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 83
4.2.1 Vector Quantization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 84
4.2.2 Soft Assignment Coding . . . . . . . . . . . . . . . . . . . . . . . . . . . 84
4.2.3 Locality-Constrained Linear Coding . . . . . . . . . . . . . . . . . . 85
4.3 Pooling . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 86
4.4 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 86
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 86
5 Image Coding . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 89
5.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 89
5.2 Linear Local Distance Coding Method . . . . . . . . . . . . . . . . . . . . . . 90
5.2.1 Distance Vector . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 91
5.2.2 Local Distance Vector . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 92
5.2.3 The Algorithm Framework . . . . . . . . . . . . . . . . . . . . . . . . . 93
5.3 Experiments and Analyses . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 94
5.3.1 Experiment Setup . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 95
5.3.2 Experimental Results on the ICPR2012 Dataset . . . . . . . . . 96
Contents ix

5.3.3 Experimental Results on the ICIP2013

Training Dataset . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 98
5.3.4 Discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 99
5.4 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 102
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 102
6 Encoding Image Features . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 105
6.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 105
6.2 Encoding Rotation Invariant Features of Images. . . . . . . . . . . . . . . 107
6.2.1 Pairwise LTPs with Spatial Rotation Invariant . . . . . . . . . . 107
6.2.2 Encoding the SIFT Features with BoW Framework . . . . . . 110
6.3 Experiments and Analyses . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 111
6.3.1 Experiment Setup . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 111
6.3.2 Experimental Results on the ICPR2012 Dataset . . . . . . . . . 112
6.3.3 Experimental Results on the ICIP2013 Training
Dataset . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 113
6.3.4 Discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 115
6.4 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 117
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 117
7 Defining Feature Space for Image Classification . . . . . . . . ......... 119
7.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ......... 119
7.2 Adaptive Co-occurrence Differential Texton Space
for Classification . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ......... 120
7.2.1 Co-occurrence Differential Texton . . . . . . . . . . . ......... 120
7.2.2 Adaptive CoDT Feature Space . . . . . . . . . . . . . ......... 123
7.2.3 HEp-2 Cell Image Representation
in the Adaptive CoDT Feature Space . . . . . . . . . . . . . . . . . 124
7.3 Experiments and Analyses . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 127
7.3.1 Experiment Setup . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 127
7.3.2 Experimental Results on the ICPR2012 Dataset . . . . . . . . . 128
7.3.3 Experimental Results on the ICIP2013
Training Dataset . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 129
7.3.4 Discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 130
7.4 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 132
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 132
8 Conclusions and Perspectives . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 135
8.1 Major Techniques Developed in the Book . . . . . . . . . . . . . . . . . . . 135
8.2 Directions and Future Work . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 136
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 137
Chapter 1
Introduction

Abstract In this chapter, we introduce the background of our study, followed by

the research motivations and objectives. Then, we introduce the publicly available
datasets used to evaluate our proposed methods. Finally, we summarize our contri-
butions and give the structure of the report in the end.

1.1 Background

1.1.1 Clinical Problems: A Case Study on Autoimmune

Diseases

Autoimmune diseases (AD) occur when the immune system mistakenly attacks and
destroys healthy body tissues, which can be organ-specific or systemic. Depending
on the type, an autoimmune disease can affect one or many different types of body
tissue. It can also cause abnormal organ growth and changes in organ function.
They are prevalent diseases affecting a large number of human beings. There are
more than eighty different types of AD such as rheumatoid arthritis, systemic lupus
erythematosus, scleroderma and autoimmune hepatitis. The common autoimmune
diseases are introduced as follows:

• rheumatoid arthritis: chronic inflammation of joints and surrounding tissues;

• systemic lupus erythematosus: affects skin, joints, kidneys, brain and other organs;
• scleroderma: a connective tissue disease that causes changes in skin, blood vessels,
muscles, and internal organs;
• sjogren’s syndrome: destroys the glands that produce tears and saliva causing dry
eyes and mouth; may affect kidneys and lungs;
• autoimmune hepatitis: causes inflammation and liver damage;
• pernicious anemia: decrease in red blood cells caused by inability to absorb vitamin
B-12;
• vitiligo: white patches on the skin caused by loss of pigment;
• psoriasis: a skin condition that causes redness and irritation as well as thick, flaky,
silver-white patches;

© Springer International Publishing AG 2017 1

X. Xu et al., Cellular Image Classification,
DOI 10.1007/978-3-319-47629-2_1
2 1 Introduction

Fig. 1.1 Patient with

systemic lupus
erythematosus (Courtesy of
Google Images)

Fig. 1.2 Patient with

rheumatoid arthritis
(Courtesy of Google Images)

• inflammatory bowel diseases: a group of inflammatory diseases of the colon and

small intestine;
• Hashimoto’s disease: inflammation of the thyroid gland;
• Addison’s disease: adrenal hormone insufficiency;
• Graves’ disease: overactive thyroid gland;
• reactive arthritis: inflammation of joints, urethra, and eyes; may cause sores on
the skin and mucus membranes;
• type 1 diabetes: autoimmune destruction of insulin producing beta cells in the
pancreas.

For instance, Fig. 1.1 shows the patient with SLE showing inflammation and vas-
culitis, while Fig. 1.2 shows a hand of the patient with rheumatoid arthritis.
The cause of autoimmune disease is still not completely understood. There are
many theories about what triggers autoimmune diseases including bacteria or virus,
1.1 Background 3

drugs, chemical irritants, environmental irritants, etc. The one whose family member
with an AD will be more susceptible to developing one. ADs affect many parts of the
body such as joints, muscles, skin, red blood cells, blood vessels, connective tissues
and endocrine glands. Meanwhile, many of ADs have similar symptoms, which are
fatigue, fever and general malaise. It makes them difficult to diagnose. Hence, ADs
have high mortality rates.

1.1.2 Cellular Imaging: A Case Study on Indirect

Immunofluorescence

The antinuclear antibody (ANA), which refers to a unique group of autoantibodies

targeting at the nuclear contents of the cell, has significant relation to specific AD,
such as systemic lupus erythematosus (SLE), systemic sclerosis and rheumatoid
arthritis. The identification of ANAs plays an important role in the clinical medicine
and clinical immunology.
Indirect immunofluorescence (IIF) is the original approach for ANA test described
by Coons and Kaplan more than 50 years ago [2]. Over time, other methods of detect-
ing ANA have been developed such as solid phase immunoassays, e.g. including
enzyme-linked immunosorbent assay (ELISA), and the multiplex platform [3, 10,
23], which is much easier and cheaper compared to IIF. However, the new tests
usually show lower sensitivity [9, 11, 12]. In addition, different commercial ELISA
kits produce different results [22] and they cannot avoid the problem of false positive
results. Recently, the American College of Rheumatology issued a position statement
stressing that the IIF method based on HEp-2 cells for ANA assay remains the gold
standard [13]. HEp-2 cells, originally considered to originate from a human larynx
carcinoma, are now known to have been established from a HeLa cell contamination
[1]. They have many advantages, such as: they are a more sensitive substrate that
allows identification of many patterns; human origin ensures better specificity than
animal tissues; the nuclei are much larger, so complex nuclear details can be detected;
the cell monolayer ensures that all nuclei are visible; cell division rates are higher so
that antigens produced only in cell division are easily located, e.g. centromere and
mitotic spindle patterns; there is no obscuring of the intercellular matrix; antigen
distribution is uniform [1].
The technique for performing the IIF test in the Clinical Immunology Laboratory
of Tan Tock Seng Hospital, Singapore is shown in Fig. 1.3 [14]. Briefly, the process
of IIF is as follows: First, diluted patient’s sample is incubated with the slides con-
taining fixed HEp-2 cells. After a period, unbound antibodies are washed off and
a fluorescein-conjugated anti-human immunoglobulin will be applied. Then, wash-
ing after the second incubation, any unbound secondary immunoglobulin will be
removed. The ANAs are finally revealed as fluorescent cells under the fluorescence
microscope. Both fluorescence intensity and positive staining patterns for each slide
image are identified by highly qualified and skillful physicians.
4 1 Introduction

BIOCHIP slide
reagent tray
BIOCHIPs

Pipe e: 10 μl per field (3×3 mm)

30 μl per field (5×5 mm) diluted samples
70 μl per field (9×7 mm)

Incubate: 30 min

Wash: 1 s flush
5 min cuve e
PBS-Tween

Pipe e: 10 μl per field (3×3 mm)

25 μl per field (5×5 mm) labelled an body
65 μl per field (9×7 mm)

Incubate: 30 min

Wash: 1 s flush
5 min cuve e
PBS-Tween

Embed: max. 10 μl per field (3×3 mm)

max. 10 μl per field (5×5 mm) embedding medium
max. 20 μl per field (9×7 mm)
cover glass

Evaluate: fluorescence microscopy

Fig. 1.3 IIF procedure

In practice, the intensity of fluorescence cells is reported on a scale of values from

0 to 4+ as follows:
• 4+: bright green fluorescence;
• 3+: less bright green fluorescence;
• 2+: diminished fluorescence yet clearly observable;
• 1+: very subdued fluorescence;
• 0: negative.
1.1 Background 5

It is worth noting that negative intensity or level 0 indicates that the patient is
normal. Staining patterns are recorded for samples with positive intensity (i.e., from
1+ to 4+), since different patterns are related to specific diseases. There are more than
thirty kinds of staining patterns in the world, however, the most frequent staining
patterns of HEp-2 cells in clinical practice are as follows [1, 5]:
• Centromere: characterized by 40–60 discrete speckles diffused over all the nuclei
in the interphase and characteristically found in the condensed nuclear chromatin
during mitosis as a bar of closely related speckles;
• Nucleolar: characterized by large coarse speckles (less than six) within the nucleoli
of interphase cells;
• Homogeneous: characterized by a uniform distributed fluorescence of the entire
interphase nuclei and fluorescence of the nuclear chromatin during mitotic;
• Speckled: there are two sub-categories:
– Fine Speckled: characterized by a fine granular nuclear staining of nuclei in the
interphase, which is distributed uniformly;
– Coarse Speckled: characterized by dense, intermediate sized particles in inter-
phase nuclei together with large speckles;
• Cytoplasmic: characterized by staining of the cytoplasm exclusive of the nucleus.
The staining patterns mentioned above are shown in Fig. 1.4. As a matter of fact,
according to different needs of the lab, the categorization of staining patterns can
be changed slightly. In some cases, mixed staining patterns reveal when patient’s
serum contains more than one autoantibody specificity. However, mixed patterns are

Fig. 1.4 Typical HEp-2 cells with different staining patterns

6 1 Introduction

atypical and occur rarely in clinical diagnosis. In most situations, same pattern is
shown in the same slide then it can be classified into common class. In academic
community, only the images with the same staining pattern are used to study.

1.2 Computer-Aided Diagnosis

Recently, Computer Aided Diagnosis (CAD) has become a part of the routine clinical
works. With CAD, radiologists use the computer output as a second opinion to make
the final decisions. The performance by CAD does not have to be comparable to or
better than that by physicians, but should be complementary to that by physicians. In
practice, a large number of CAD systems have been employed for assisting physicians
in the early detection of breast cancers on mammograms.
Worldwide, there is a strong demand for IIF image analysis due to its effectiveness
and high-quality. However, the main drawback of the strategy is the human subjective
evaluation since the classification always depends on highly qualified experts’ skill.
Results are significantly influenced by one’s qualification and reading systems, caus-
ing high-rate intra- and inter-laboratory variance. The low level of standardization
limits the communications between the clinic units and reproducibility of IIF read-
ings. As the demand for applying IIF method in diagnosis of autoimmune diseases
is increasing, lacking of resources and adequately trained personnel is becoming
another bottleneck [6, 17].
To address aforementioned issues, CAD systems are desired for automatically
supporting the IIF diagnosis in many ways. They can augment the physicians’ capa-
bilities, reduce the time and improve the reliability. They free the physicians from
mass screening tasks and enable them to concern only about the most involved cases.
Moreover, the CAD systems can be used as an education tool to train specialized
medical personnel.
Several approaches have been proposed in the recent research for all the major
stages of the IIF diagnostic procedure. The main technologies have been investigated
in the CAD systems are automated preparation of slides with robotic devices [18],
image acquisition [8, 21], image segmentation [15, 16], mitotic cell recognition [4],
fluorescence intensity classification [18, 19] and staining pattern classification [6,
7, 20]. Till now, the highest level of automation in the CAD systems for ANA test
is the preparation of slides by robotic devices conducting dilution, dispensation and
washing operations [4]. In image acquisition stage, images are taken using a fluores-
cence microscope, equipped with both a mercury vapour lamp and a digital camera
to auto-focus; then the digital images are displayed on a computer screen. Images
are analyzed and processed to be more suitable to following classification. Fluores-
cence intensity is classified into negative group and positive group with intensity
level based on the intensity-related features. The positive group is further classified
into several main staining pattern groups based on the pattern-related features.
1.2 Computer-Aided Diagnosis 7

Coarse Fine
Centromere speckled Cytoplasmic speckled Homogeneous Nucleolar

Positive

Positive

Intermediate

Intermediate

Intermediate

ICPR2012 Dataset

Nuclear
Centromere Golgi Homogeneous Nucleolar Membrane Speckled

Positive

Positive

Intermediate

Intermediate

Intermediate

ICIP2013 Training Dataset

Fig. 1.5 Sample images from ICPR 2012 contest dataset and ICIP 2013 contest training dataset
respectively. The rows named “positive” are the patterns with positive intensity, while the “inter-
mediate” rows are the patterns with intermediate intensity
8 1 Introduction

While all aspects of the CAD systems contribute to the automation of IIF pro-
cedure in one way or another, staining pattern classification is proven to be the
most challenging task in the research community due to large intra-class variation
and small between-class variation regardless of its importance. To reduce the vari-
ability of multiple readings, the levels of fluorescent intensity are always generally
classified into three levels named negative, intermediate (with intensity of 1+) and
positive (with intensity of 2+ or more), where intermediate and positive is belong to
the positive group whose pattern need to be further identified. In the following of
this thesis, when we say “positive staining patterns”, we refer to staining patterns
with non-negative fluorescence intensity, which includes positive and intermediate
level. As shown in Fig. 1.5, the cells with intermediate and positive intensity which
are from the same pattern categories have large variations. Particularly, the cells with
intermediate intensity can not be seen clearly. Meanwhile, some categories share
similar shape and texture. Image representation is crucial for HEp-2 cells analysis.
Compared with the signal in general image classification, HEp-2 cells do not contain
abundant structural information. In addition, the features between various HEp-2
cells are much more similar than those between different objects or natural scene
images. In this thesis, we investigate into the feature extraction and machine learning
methods for automatic staining pattern classification of HEp-2 cells.

1.3 Experimental Datasets in the Book

In order to evaluate the performance of our proposed methods in this thesis, we use
two publicly available HEp-2 cells datasets: ICPR2012 dataset from the ICPR’12
HEp-2 cell classification contest dataset and ICIP2013 training dataset from the
ICIP’13 Competition on cells classification by fluorescent image analysis. Some
examples of the datasets are shown in Fig. 1.6.

1.3.1 The ICPR2012 Dataset

The ICPR2012 dataset consists of 1455 HEp-2 cells segmented from 28 slide images
which are obtained by using a fluorescence microscope of 40-fold magnification,
equipped with a 50W mercury vapor lamp and a digital camera using a CCD with
square pixel of 6.45 µm. The resolution of obtained images is 1388 × 1038 pixels
and the color depth is 24 bits. Each image can be categorized into one of six stain-
ing patterns, namely centromere (ce), coarse speckled (cs), cytoplasmic (cy), fine
speckled (fs), homogeneous (ho) and nucleolar (nu). Also, fluorescent intensity (i.e.,
positive and intermediate) is assigned to each image. The cells in the images are man-
ually segmented and annotated by specialists. Then, each cell image and slide image
1.3 Experimental Datasets in the Book 9

Coarse Fine
Centromere speckled Cytoplasmic speckled Homogeneous Nucleolar

ICPR2012 Dataset

Nuclear
Centromere Golgi Homogeneous Nucleolar Membrane Speckled

ICIP2013 Training
Dataset

Fig. 1.6 Samples of the ICPR2012 dataset and the ICIP2013 training dataset with different staining
patterns of HEp-2 cells

is verified by a medical doctor specialized in immunology with 11 years’ experi-

ence. According to the experimental protocol of the ICPR’12 contest, the ICPR2012
dataset is divided into a training set with 721 cells from half of the slide images and
a test set with 734 cells from rest of the slide images. The composition of the dataset
is reported in Table 1.1.
It is worth noting that the similarity of the cells in the same slide image is always
higher than that of the cells from different slide images. To evaluate the generalization
ability of the algorithms, the cells in one slide image can only be used for training or
testing. If the cells for training and testing are randomly selected from the database,
there can be some cells from the same slide image both for training and testing. The
classification accuracy obtained via this strategy is much higher than that by using
the contest instruction, which is unfair. In our experiments, we strictly following the
experimental protocol of the contest.
10 1 Introduction

Table 1.1 Composition of the ICPR2012 dataset. Each table item represents the number of cells
and the number of images which is in parentheses
Type Training set Test set Total
Centromere 208 (3) 149 (3) 357 (6)
Homogeneous 150 (3) 180 (2) 330 (5)
Nucleolar 102 (2) 139 (2) 241 (4)
Coarse speckled 109 (2) 101 (3) 210 (5)
Fine speckled 94 (2) 114 (2) 208 (4)
Cytoplasmic 58 (2) 51 (2) 109 (4)
Total 721 (14) 734 (14) 1455 (28)

1.3.2 The ICIP2013 Training Dataset

The HEp-2 cell images of ICIP2013 dataset is obtained by using a monochrome

high dynamic range cooled microscopy camera which is fitted on a microscope with
a plan-Apochromat 20x/0.8 objective lens and an LED illumination source. At least
two scientists are involved in the labeling process. In dubious cases, a third expert is
asked to adjudicate the conflict between the two scientists. So far, only the training
dataset is available. However, the training dataset is big enough to evaluate different
methods. The ICIP2013 training dataset contains 13596 cells which are categorized
into six classes: homogeneous (ho), speckled (sp), nucleolar (nu), centromere (ce),
nuclear membrane (nm) and golgi (go). The dataset includes two patterns less fre-
quent occurring in the practical clinic as follows [24]:
• Nuclear membrane: characterized by a thin membranous fluorescence around the
nucleus in the interphase cells;
• Golgi: characterized by speckled staining of a polar organelle adjacent to one part
of the nucleus and composed of irregular large granules.
Thus, it offers a more realistic evaluation on the automatic classification algorithms.
We partition the ICIP2013 training dataset into a training set consisting of 6842 cells
from 42 slide images and a test set consisting of 6754 cells from 41 slide images.
See Table 1.2 for detailed information about the dataset.

1.4 Structure of the Chapters

Chapter 2 provides the reader with the fundamentals of cellular imaging, imaging
feature detection and classification. We introduce the application of optical technol-
ogy in imaging, and the analysis and classification of HEp-2 cell images. Our work
focuses on the efficient feature extraction for staining pattern classification of HEp-2
cells. There are various features for image classification. We introduce some widely
1.4 Structure of the Chapters 11

Table 1.2 Composition of the ICIP2013 training dataset. Each table item represents the number
of cells and the number of images which is in parentheses
Type Training set Test set Total
Centromere 1279 (16) 1462 (16) 2741 (32)
Homogeneous 1347 (16) 1147 (16) 2494 (32)
Nucleolar 1273 (16) 1325 (16) 2598 (32)
Speckled 1391 (16) 1440 (16) 2831 (32)
Nuclear membrane 1190 (16) 1018 (15) 2208 (31)
Golgi 362 (2) 362 (2) 724 (4)
Total 6842 (42) 6754 (41) 13596 (83)

used features for describing staining patterns, and some classifiers for these staining
pattern classification.
In Chap. 3, we describe a low-order LP21 fiber mode for optical single cell manip-
ulation and imaging staining patterns of HEp-2 cells. A focused four-lobed mode
distribution is stable and effective in optical tweezer applications, including selective
cell pick-up, pairing, grouping or separation, as well as rotation of cell dimers and
clusters. Both translational dragging force and rotational torque in the experiments
are in good accordance with our theoretical model. With a simple all-fiber configura-
tion, and low peak irradiation to targeted cells, instrumentation of our optical chuck
technology will provide a powerful tool in the ANA-IIF laboratories.
In Chap. 4, we introduce the Bag-of-Words (BoW) framework for image repre-
sentations. Many models transform low-level descriptors into richer mid-level rep-
resentations. Extracting mid-level features involves a sequence of interchangeable
modules. However, they always consist of two major parts: Bag-of-Words (BoW)
and Spatial Pyramid Matching (SPM). The target is to embed low-level descriptors
in a representative codebook space. We introduce the key techniques employed in
the BoW framework including the coding and pooling processes.
In Chap. 5, to reduce the inevitable information loss caused by coding process, we
study a Linear Local Distance Coding (LLDC) method. The LLDC method trans-
forms original local feature to more discriminative local distance vector by search-
ing for local neighbors of the local feature in the class-specific manifolds. Then we
encode and pool the local distance vectors to get salient image representation. Com-
bined with the traditional coding methods, this method achieves higher classification
accuracy.
Chapter 6 is focused on a rotation invariant textural feature of Pairwise Local
Ternary Patterns with Spatial Rotation Invariant (PLTP-SRI). It is invariant to image
rotations, meanwhile it is robust to noise and weak illumination. By adding spatial
pyramid structure, this method captures spatial layout information. While the pro-
posed PLTP-SRI feature extracts local feature, the BoW framework builds a global
image representation.
12 1 Introduction

In Chap. 7, we design a Co-occurrence Differential Texton (CoDT) feature to

represent the local image patches of HEp-2 cells. The CoDT feature reduces the
information loss by ignoring the quantization while it utilizes the spatial relations
among the differential micro-texton feature. Thus it can increase the discriminative
power. We build a generative model to adaptively characterize the CoDT feature
space of the training data. Furthermore, we exploit a discriminant representation
for the HEp-2 cell images based on the adaptive partitioned feature space. There-
fore, the resulting representation is adapted to the classification task. By cooperating
with linear Support Vector Machine (SVM) classifier, our framework can exploit
the advantages of both generative and discriminative approaches for cellular image
classification.
Chapter 8 concludes this monograph with its major techniques developed, and
gives our perspectives on the future directions of research in this field.

References

1. AR Bradwell and RG Hughes. Atlas of Hep-2 patterns and laboratory techniques. Binding
Site, 2007.
2. Albert H Coons and Melvin H Kaplan. Localization of antigen in tissue cells ii. improvements
in a method for the detection of antigen by means of fluorescent antibody. The Journal of
experimental medicine, 91(1):1–13, 1950.
3. Woodruff Emlen and Laurie O’Neill. Clinical significance of antinuclear antibodies. compari-
son of detection with immunofluorescence and enzyme-linked immunosorbent assays. Arthritis
& Rheumatism, 40(9):1612–1618, 1997.
4. P. Foggia, G. Percannella, P. Soda, and M. Vento. Early experiences in mitotic cells recognition
on hep-2 slides. In 23rd international symposium on Computer-based medical systems, pages
38–43, 2010.
5. P Foggia, G Percannella, P Soda, and M Vento. Benchmarking hep-2 cells classification meth-
ods. IEEE transactions on medical imaging, 32(10):1878–1889, 2013.
6. R. Hiemann, T. Büttner, T. Krieger, D. Roggenbuck, U. Sack, and K. Conrad. Challenges of
automated screening and differentiation of non-organ specific autoantibodies on hep-2 cells.
Autoimmunity Reviews, 9(1):17–22, 2009.
7. R. Hiemann, N. Hilger, J. Michel, J. Nitschke, A. Boehm, U. Anderer, M. Weigert, and U. Sack.
Automatic analysis of immunofluorescence patterns of hep-2 cells. Annals of the New York
Academy of Sciences, 1109(1):358–371, 2007.
8. R. Hiemann, N. Hilger, U. Sack, and M. Weigert. Objective quality evaluation of fluorescence
images to optimize automatic image acquisition. Cytometry Part A, 69(3):182–184, 2006.
9. P Kern, M Kron, and K Hiesche. Measurement of antinuclear antibodies: assessment of different
test systems. Clinical and diagnostic laboratory immunology, 7(1):72–78, 2000.
10. Yashwant Kumar, Alka Bhatia, and Ranjana Walker Minz. Antinuclear antibodies and their
detection methods in diagnosis of connective tissue diseases: a journey revisited. Diagn Pathol,
4(1):1–10, 2009.
11. Pier Luigi Meroni and Peter H Schur. Ana screening: an old test with new recommendations.
Annals of the Rheumatic Diseases, 2010.
12. Artemissia Phoebe Nifli, George Notas, Marilena Mamoulaki, Maria Niniraki, Vaso
Ampartzaki, Panayiotis A Theodoropoulos, Mark J Kopnitsky, and Elias Castanas. Compar-
ison of a multiplex, bead-based fluorescent assay and immunofluorescence methods for the
detection of ana and anca autoantibodies in human serum. Journal of immunological methods,
311(1):189–197, 2006.
References 13

13. American College of Rheumatology. Position statement: Methodology of testing for antinuclear
antibodies. www.rheumatology.org/practice/clinical/position/ana_position_stmt.pdf, August,
2011.
14. NO ORDER. Anti-coxsackie virus ifa (iga, igg or igm) biochip mosaic test system test instruc-
tion.
15. G. Percannella, P. Soda, and M. Vento. A classification-based approach to segment hep-2 cells.
In 25th international symposium on Computer-based medical systems, pages 1–5, 2012.
16. Petra Perner, Horst Perner, and Bernd Müller. Mining knowledge for hep-2 cell image classi-
fication. Artificial intelligence in medicine, 26(1):161–173, 2002.
17. P. Soda. Early experiences in the staining pattern classification of hep-2 slides. In 20th IEEE
International Symposium on Computer-Based Medical Systems, pages 219–224, 2007.
18. P. Soda and G. Iannello. A multi-expert system to classify fluorescent intensity in antinuclear
autoantibodies testing. In 19th IEEE International Symposium on Computer-Based Medical
Systems, pages 219–224, 2006.
19. P. Soda, G. Iannello, and M. Vento. A multiple expert system for classifying fluorescent inten-
sity in antinuclear autoantibodies analysis. Pattern Analysis and Applications, 12(3):215–226,
2009.
20. Paolo Soda and Giulio Iannello. Aggregation of classifiers for staining pattern recognition in
antinuclear autoantibodies analysis. IEEE Transactions on Information Technology in Biomedi-
cine, 13(3):322–329, 2009.
21. Paolo Soda, Amelia Rigon, Antonella Afeltra, and Giulio Iannello. Automatic acquisition of
immunofluorescence images: Algorithms and evaluation. In 19th IEEE International Sympo-
sium on Computer-Based Medical Systems, pages 386–390, 2006.
22. Elio Tonutti, Danila Bassetti, Anna Piazza, Daniela Visentini, Monica Poletto, Franca Bassetto,
Patrizio Caciagli, Danilo Villalta, Renato Tozzoli, and Nicola Bizzaro. Diagnostic accuracy of
elisa methods as an alternative screening test to indirect immunofluorescence for the detection
of antinuclear antibodies. evaluation of five commercial kits. Autoimmunity, 37(2):171–176,
2004.
23. SE VOLLSET. Evaluation of diagnostic tests for antinuclear antibodies in rheumatological
practice. Scand. J. Immunol, 52:309–315, 2000.
24. Allan S Wiik, Mimi Høier-Madsen, Jan Forslid, Peter Charles, and Jan Meyrowitsch. Anti-
nuclear antibodies: a contemporary nomenclature using hep-2 cells. Journal of autoimmunity,
35(3):276–290, 2010.
Chapter 2
Fundamentals

Abstract In this chapter, we firstly introduce the application of optical technology in

cellular imaging. Then, we will introduce the fundamentals about analysis and clas-
sification of HEp-2 cell images. Our works focus on the efficient feature extraction
for staining pattern classification of HEp-2 cells. There are countless features existed
for image classification. Firstly we refer some widely used features for describing
staining patterns, then we will introduce some fundamental classifiers for staining
patterns classification.

2.1 Optical Systems for Cellular Imaging

This section introduces application of optical technology in cell imaging, highlight-

ing the basic principle and main features of a list of microscopic techniques, includ-
ing laser scanning confocal microscope, multi-photon fluorescence imaging, total
internal reflection fluorescence microscopy, near-field scanning optical microscopy
imaging, and optical coherence tomography. To an extent, these optical imaging
techniques have demonstrated characteristics of non-invasive, non-ionizing radia-
tion, and various operating modes, providing a variety of real-time quantification of
cellular properties and playing a decisive role in the life science research. Applica-
tions of optical imaging technology in life science have a history of over two hundred
years, and microscopic analysis has been playing an important role in the biomed-
ical field. Recent progress in laser technology has made remarkable achievements
in tissue optics, spectroscopy, imaging technology, optical diagnostics and treatment
technology. In many of state-of-the-art biomedical systems in nano-science, such as
microarray biochips and one-drop blood diagnosis, optical imaging is a core part that
plays a pivotal role in analysis. A new discipline-biomedical photonics has formed,
focusing on optical technology in biomedicine.

© Springer International Publishing AG 2017 15

X. Xu et al., Cellular Image Classification,
DOI 10.1007/978-3-319-47629-2_2
16 2 Fundamentals

2.1.1 Laser Scanning Confocal Microscope

Optical scanning confocal microscope has gained decades of continuous improve-

ment for its high resolution and penetration capability into tissues. In 1957, Marvin
Minsky clarified the basic principles of scanning confocal microscopy techniques for
the first time. But it did not really become available in scientific research until 1985
with the rapid development of computer, laser and image processing technology. Wij-
naedts Van Resaint first succeeded in obtaining a cross-section of the optical probe
with a fluorescent labeled biological material, demonstrating confocal fluorescence
microscopy has an ability to reject defocused light to obtain three-dimensional image
of biological samples. Within the next two years, Oxfordshire company, which was
acquired by Bio-Rad Laboratories later, produced the first commercial laser scan-
ning confocal microscopy, which became widely used in many fields. With high
resolution, it implemented observation of biological processes within living tissue.
Confocal microscopy has the ability to make chromatography, obtaining fluores-
cence image of internal fine structure of cells, or observing physiological signals and
changes in cell morphology Ca2+ , pH value, the membrane potential at the subcellu-
lar level, etc. These advantages make laser scanning confocal microscope a powerful
tool for clinical diagnosis, biomedical biochemistry, cell biology, reproductive biol-
ogy and neurology [11, 19, 21, 23, 24, 34, 37, 42, 46, 50, 57, 60].

2.1.1.1 Fluorescent Probe and Laser Scanning Confocal Microscopy

Imaging Techniques

Confocal laser scanning microscopy is a powerful tool for in situ observation and
analysis in cell biological research involving fluorescent probes. There is a class of
dyes in biological stain that can be excited by ultraviolet or blue-violet light and
other short-wavelength light and emit fluorescence, which is called fluorescent dyes,
also known as fluorescent pigment. Now some applied dyes can also be excited by
longer wavelength light and produce longer wavelength fluorescence, these fluores-
cent dyes are often referred to as fluorescent probes. Fluorescent staining has the
most significant advantages of high sensitivity in observation. The concentration of
non-fluorescent dye usually needs to be 1 % or higher to make the cells to carry
with visible color. Fluorescent dyes, such as fluorescent yellow stained with a con-
centration of 10−5 M, can stimulate visible fluorescence with ultraviolet excitation.
Fluorescent dyes can produce a desired dyeing effect with concentration of from
10−4 M to 10−5 M.
Fluorescent dyes may be divided into three categories according to their chemical
reactivity of fluorescent probes: (1) Alkalinity fluorescent dye, which contains alka-
line chromophore ionized in acidic solution, with fluorescent color being cationic
ions. Acridine dyes such as acridine yellow can stain cells by binding to DNA and
RNA, mainly through embedded in the DNA double helix structure. Ethidium bro-
mide stains by embedding in DNA, emitting red fluorescence under irradiation of
2.1 Optical Systems for Cellular Imaging 17

ultraviolet light. (2) Acid fluorescent dyes that contain acidic chromophore, ionized
in alkaline solution, while fluorescent color ions being anions. (3) Neutral fluorescent
dye, a compound dye that is a mixture of acidic and alkalinity fluorescent dye.
Depending on different applications, the fluorescent probes can further be divided
into many categories, including cell active probes, probe organelles, membrane flu-
orescence shoving needles, accounting probe, the membrane potential probes, ion
probe, pH probe, reactive oxygen probes, immunofluorescence probes, probe caged
compound, cytoskeletal protein fluorescence probes. It is essentially important to
choose a fluorescent probe according to requirements of analysis and application
conditions.
On the other hand, there has been a new nano-material called “quantum dots”,
which have gradually become the latest fluorescent probes for their remarkable opti-
cal efficiency and exhibited a unique advantage in the field of cellular imaging tech-
nology. Study scope of quantum dots involves multiple disciplines. Even its name
varies in different field of research, For example, colloid chemists attributed it to
the colloidal particles (colloid particle), and material scientist named it nanocrys-
tals (nanocrystal); while the solid-state physicists named it quantum dots due to
confinement of electrons in a region of tens nanometer scale. Common semiconduc-
tor quantum dots are divided into quantum dots of iv Group (Si, Ge), iii–v Group
(lnAs, GaSb), and Group II–VI (ZnTe, CdSe, CdS, Zn0). iv Group and iii–v Group
quantum dots are fabricated using photolithography, selective epitaxial growth and
self-assembly method. Group ii vi diselenide, e.g. CdSe, has excellent fluorescence
properties because of its wide bandgap and direct band gap characteristics. Fluores-
cent probe label plays an important role in cell microscopic imaging, as such, ii–vi
Group quantum dots have attracted wide attention in this field [4, 7, 9, 22, 33, 35,
36, 38, 49, 52, 58].
Group ii–vi quantum dots have remarkable advantages over conventional organic
fluorescent dyes on the following aspects: (1) as a multi-electron system, absorption
coefficient of the quantum dots is much higher than that of a single molecule, with
a magnitude of absorption coefficient reaching 105 L * mol−1 * cm−1 under visible
light or ultraviolet light excitation, which makes its fluorescence emission intensity
much higher than the organic dye; (2) by changing the material ratio and size of
the quantum dot, the fluorescence emission wavelength can cover a wide spectral
range from 400 nm to 2 µm; (3) in contrast to organic dye molecules that have a
narrow excitation spectrum, quantum dots have a wide and continuous excitation
spectrum and there is a quantum confinement emission peak (the longest wavelength
can excite quantum dot to emit fluorescent light), any light whose wavelength is
shorter than the quantum confinement peak can efficiently excite the quantum dots,
so it is possible to use a single wavelength light source to excite quantum dot with
different composition and sizes, making them emit different colors of fluorescence for
simultaneous monitoring or distinguishing bio-processes; (4) fluorescence emission
peak of quantum dots in an organic dye has a narrow and symmetrical peak, with
a half-width of only 1/3 of that of fluorescent dyes, and has no tail in long-wave
side; (5) the fluorescence lifetime is longer (about several hundred nanoseconds),
fluorescence bleaching rate is only 1/100 of rhodamine 6G (a popular red fluorescent
18 2 Fundamentals

dye), therefore it can be used for a long lifetime fluorescence microscopy experiments
with a great potential to substitute conventional organic dyes as a new biological
fluorescent probe.
Two papers by Alivisatos and Nie respectively published in the same issue of
Science in 1998 debuted the application of quantum dots in the biomedical fields,
firstly reporting quantum dots in cell imaging as a fluorescent probes instead of con-
ventional dyes [9]. The usage of quantum dots for highly sensitive cellular imaging
has found major advances over the past decade [51]. The improved photostability of
quantum dots, for example, allows the acquisition of many consecutive focal-plane
images that can be reconstructed into a high-resolution three-dimensional image [59].
Another application that takes advantage of the extraordinary photostability of quan-
tum dot probes is the real-time tracking of molecules and cells over extended periods
of time [13]. Antibodies, streptavidin [27], peptides [3], DNA [18], nucleic acid
aptamers [16], or small-molecule ligands [33] can be used to target quantum dots to
specific proteins on cells. Researchers were able to observe quantum dots in lymph
nodes of mice for more than 4 months [6].
Semiconductor quantum dots have also been employed for in vitro imaging of
pre-labeled cells. The ability to image single-cell migration in real time is expected
to be important to several research areas such as embryogenesis, cancer metastasis,
stem cell therapeutics, and lymphocyte immunology.

2.1.1.2 Fundamentals of Confocal Laser Scanning Microscopy

Figure 2.1 shows a description of the basic principles of confocal laser scanning
microscope imaging. The excitation light goes through illumination pinhole and form
a point light source, reflected by a beam splitter by means of exciting filter, focused
by the microscope objective into the three-dimensional sample. By scanning in the
direction perpendicular to the optical axis in the xy plane (focal plane), fluorescence
emitting from illuminated region on the focal plane and on the top and bottom of
the focal plane is collected by objective with the beam splitter and emission filter.
There is a confocal pinhole in front of the detector, the illumination pinhole and
confocal pinhole is conjugate with respect to the focal plane of the objective lens, so
that only the fluorescence emitted from the focal plane can be focused to a confocal
pinhole and goes through the pinhole and reaches the detector either PMT or CCD.
Points above or below the focal plane cannot be imaged into the detect pinhole,
the light emitted outside of the focal plane is blocked by the pinhole, it contributed
very little to confocal images. Thus the confocal images substantially only obtain
light information from the focal plane, namely acquired a 2D image sampled by
the focal plane. If gradually adjusting the position of the longitudinal axis of the
sample, multiple tomographic image of the sample can be generated and each cross-
sectional image of cell or tissue can be clearly displayed. This imaging method is
called a confocal scanning microscope tomography. With three-dimensional image
reconstruction technique, a high resolution three-dimensional image of the sample
can be provided, as in many commercial instruments.
2.1 Optical Systems for Cellular Imaging 19

Fig. 2.1 Imaging principle

of confocal laser scanning
microscopy

2.1.1.3 Characteristics of Confocal Laser Scanning Microscope

If the pinhole is removed from a laser scanning confocal microscope imaging sys-
tem, the defocused light (light from off focal planes) can reach the detector, the
resolution in the depth direction will be greatly reduced, then the non-confocal opti-
cal microscope would be no much different from an ordinary microscope. So the
most fundamental difference between confocal laser scanning microscope and the
ordinary microscope is that there is a pair of conjugate pinholes to limit received
light rays only from focal plane, thus largely eliminating the blur effect of defocused
light. Another remarkable difference lies on the fact that laser scanning system can
acquire an extremely low aberration as the whole optical imaging system needs
only to deal with point-to-point imaging, instead of a complete field of view, which
requires a sophisticated optical aberration reduction for system design. Compared
with a general-purpose optical microscope, a confocal laser scanning microscope
has the following characteristics:
(1) Scattering background of confocal laser scanning microscope is lower than nor-
mal optical microscope, resulting in a high image contrast and detection sensi-
tivity.
(2) Three-dimensional imaging can be reconstructed from layer to layer image
data, forming a light-ray tomography through line by line, point by point three-
dimensional scanning, completely different from computed tomography method.
Image data stored in the computer allows display any two-dimensional cross-
sectional or perspective view of three-dimensional tomography.
20 2 Fundamentals

(3) Utilization of laser light source provides a small size and high brightness illumi-
nation light spot on focal place. Monochromatic laser light essentially eliminates
chromatic aberration of the system. With advances in laser technology, more and
more types of lasers have been fully able to meet the needs of biomedical research
sample.
Commercialization of laser scanning confocal microscope are usually equipped
with powerful image processing software, realizing functions of positioning of mul-
tiple fluorescent labeling, optical sectioning, 3D reconstruction, time series scanning
and other quantitative analysis.

2.1.1.4 Application of Confocal Laser Scanning Microscope

Being equipped with advanced data acquisition, recording and processing algorithm,
laser scanning confocal microscope is becoming one of the most important molecular
cell biology analytical instrumentation in life sciences research, including observing
the structure of living cells and specific molecules, ions of biological histologi-
cal changes, quantitative analysis, and real-time quantitative determination of bio-
systems, which has derived a number of new technologies and new methods such
as adherent cell sorting, laser cell microsurgical techniques, fluorescence recovery
after photo-bleaching technique, intercellular communication, cell membrane fluid-
ity measuring technology and photo-activation.
Laser scanning confocal microscopy has been widely used in the fields of biol-
ogy, biochemistry, physiology, pharmacology, pathology, genetics and immunology
and embryology science, environmental medicine and nutritional science. Some of
examples include: investigation of internal element of the insect digestive tract, the
enteric valve of Apicotermitinae [26], and quantum dots labeled three-dimension
imaging of nanofibrous structure [5].

2.1.2 Multi-photon Fluorescence Imaging

Conventional fluorescence microscopy techniques are based on single-photon excita-

tion fluorescence, in which a fluorescent molecule absorbs a photon, transiting from
the ground state to an excited state, followed with emission of a long wavelength pho-
ton by transiting to the ground state after the energy relaxation. The photon energy
of excitation light used here must be higher than that of the emitted fluorescence
wavelength for energy conservation. If long-wavelength photons to excite the fluo-
rescent molecule, it will not produce fluorescence. In 1931, Maria Goppert-Mayer
predicted that if one photon does not have enough energy to excite the fluorescent
molecule, but in a short time to encounter a second or more photons, the same mole-
cule can simultaneously absorb two or more photons, each of the absorption produces
the molecule excitement with an equivalent energy of the absorbed photon. Conse-
2.1 Optical Systems for Cellular Imaging 21

quently, the fluorescent molecule, after gaining sufficient photon energy, will also
fluoresce after relaxation. This multiphoton excitation process, was not observed
until the 1960s when two-photon excited fluorescence was first discovered in CaF2 :
Eu3+ by Kaiser and Garret [34], there-photon excited fluorescence was observed and
the three-photon absorption cross section for naphthalene crystals was estimated by
Singh and Bradley [35].
As shown in Fig. 2.2, multiphoton fluorescence involves the absorption of multiple
photons to an excited electronic state followed by the relaxation of the molecule to
the ground electronic state through the emission of a single photon. The wavelength
of the emitted photon is approximately equal to the excitation wavelength divided
by the number of photons absorbed.
Two-photon excitation wavelength is twice single-photon excitation wavelength.
For example, to excite the fluorescent probe Indo-l with Ar+ laser, we use 351 nm
laser, while two-photon excitation is necessary to use 700 nm laser. Despite there
is a difference between single-photon and multi-photon excitation process, the flu-
orescence emission spectra are identical. That is, the multi-photon technology can
detect ultraviolet fluorescent probe without the use of an ultraviolet light source.
But to achieve multi-photon excitation, it usually requires ultra-fast femtosecond
laser pulses to produce a high density of photons focused on suitable fluorescent
medium, and induce multi-photon transition with a sufficiently high probability, due
to the fact that the probability that more than one photon can be absorbed simulta-
neously scales with intensity raised to the nth power where n is equal to the number
of photons absorbed. Thus the multi-photon excitation has high spatial local char-
acteristics, only samples in the center area of focus can absorb enough photons to
give off fluorescence, which naturally reduces size of emission spots on sample upon
laser illumination. Based on this principle, multi-photon excitation can obtain clearer
three-dimensional fluorescence image than the single-photon confocal (Figs. 2.3 and
2.4).
In addition, multi-photon fluorescence excitation uses red or infrared light, which
minimizes scattering in the tissue. Further the background signal is strongly sup-
pressed. Both effects lead to an increased penetration depth for these microscopes,
typically 5–20 times deeper than other types of fluorescent microscopes. Two-photon

Fig. 2.2 Single photon

versus multi-photon
excitation
Another random document with
no related content on Scribd:
 Psammoma of Dorsal Cord, 38 of Table (after Charcot).

PROGNOSIS.—The prognosis of spinal tumors is generally very

unfavorable. Syphilitic cases are of course the most hopeful, but
even in these cases it is only when they are recognized early that
much can be expected. A gumma that has grown to any dimensions
will have so compressed the cord that even when the tumor is
melted away by specific treatment its effects will remain.
DURATION AND TERMINATION.—Most cases of spinal tumor last from
about six months to three years. Occasionally death may result, as
from a rapidly-developing sarcoma, in less than six months, and
somewhat more frequently in slowly-developing tumors, or in those
which are held more or less in abeyance by treatment the sufferings
of the patient are prolonged to four or five years or more.
Hemorrhages into or around the growths sometimes take place, and
are the cause of death, or more frequently of a sudden aggravation
and multiplication of severe symptoms. Death sometimes takes
place from the complete exhaustion which results from the disease
and its accompanying secondary disorders, such as bed-sores,
pyelitis, etc. Occasionally death results from intercurrent diseases,
such as pneumonia, infectious fevers, etc., whose violence the
weakened patient cannot well withstand. Sometimes the symptoms
of a rapidly-ascending paralysis appear, probably due to an
ascending myelitis or meningo-myelitis.

COMPLICATIONS AND SEQUELÆ.—Spinal tumors are sometimes

complicated with other similar growths in the brain or the evidences
of the same constitutional infection in other parts of the body. In one
case of cysticercus of the cord sclerosis of the posterior columns
was also present.

TREATMENT.—The treatment of spinal tumors can be compressed into

very small compass. In cases with syphilitic history, or when such
history is suspected, although not admitted, antisyphilitic remedies
should be applied with great vigor. It should be borne in mind,
however, that even in syphilitic cases after destruction of the cord by
compression or softening specific remedies will be of no avail. In
tubercular cases and in those in which the system is much run down
tonics and nutritives are indicated. Bramwell11 advises an operation
in any case in which the symptoms are urgent, in which the
diagnosis clearly indicates the presence of a tumor, when there is no
evidence of malignant disease, when the exact position of the growth
can be determined, and when a vigorous antisyphilitic treatment has
failed to produce beneficial results. As some meningitis, meningo-
myelitis, or myelitis is usually present in cases of spinal tumor,
treatment for the complication will assist in relieving the torments of
the patient. Anodynes, particularly opium and its preparations,
should be used freely in the later stages of the affection. Bromides
and chloral are of little value except in association with opiates.
Operation offers even less hope than in brain tumor, but in very rare
cases should be taken into consideration.
11 Diseases of the Spinal Cord, Edinburgh, 1884.

TABLE OF FIFTY CASES OF SPINAL TUMOR.

Sex
Path. Anat. and
No. and Clinical History. Remarks.
Location.
Age.
1 M. 33. Paresis of forearms, left worse. Paraplegia, then Glioma; T. Whipham,
paralysis of all limbs; paralysis of intercostals. syringo-myelus. Trans. Path.
Contractures of hands, then of feet. Pain and Dilated Soc. London,
stiffness of neck on motion. Wasting of interossei. lymphatics. 1881, xxxii.
Diplegic contractions of legs. Only partial 8-12.
paralysis of sphincters. Sensation perfect. Bed- Entire length of
sores. Duration, thirteen months. cord, and
involving
medulla
oblongata.
Upper four
inches of cord
greatly
enlarged.
2 F. —. Constricting pains about abdomen. Paresis of Glio-myxoma. Schueppel,
legs. Persistent subsultus. Temporary Arch. d.
improvement after labor. General paralysis. In gray columns Heilk., viii.
Scoliosis. from medulla Bd., 1867
oblongata to (quoted by
cauda equina. Rosenthal).
3 M. 15. Paresis of left arm. Pain back of neck. Later, Gelatinous S. Wilks,
paralysis of left arm, and wasting of arm, tumor left side Lectures on
shoulder, and neck muscles. Slight paresis of of cord, and Dis. of
involving in Nervous
 right arm. Prolonged vomiting. Constriction of some parts the System, p.
neck; dysphagia; paralysis of chest. gray matter. 266.

From medulla
to sixth cervical
vertebra.
4 M. 18. Paresis of left leg, increasing; some atrophy. Round-celled Schultze (F.),
Weakness in left arm. Later, numbness in both sarcoma or Arch. f.
legs. Contracture of fingers. Some mental glio-sarcoma, Psychiat.,
confusion. Left hand and leg livid and cold. growing from Berlin, 1878,
Hyperæsthesia of left leg; anæsthesia of right leg, ependyma of viii. 367-393,
perineum, penis, scrotum, rectum, and inguinal central canal, 1 pl.
region, and of left arm. Right arm normal. Islands causing
of heat and cold in leg, and of cold in arms. Left hydromyelia,
ankle clonus. Left pupil contracted. Vomiting. softening, and
Dysphagia. Occipito-cervical pain and contracture secondary
of cervical muscles. Leg contractures and tremor. degeneration.
Later, hyperæsthesia disappeared. Incontinence Dura mater
of urine. Patellar and skin reflexes increased. thickened.
Facial spasm. Amblyopia, optic neuritis, diplopia, Brown exudate
deafness, paralysis of left abducens; pupils in cord and
contracted. Sacral bed-sores. Thick speech. base of brain.

From medulla
oblongata to
dorsal cord.
5 F. 48. Pain in abdomen and down legs, worse on left Tumor J.
side. Tonic spasm in flexors and adductors of (psammoma), Hutchinson,
thighs. No anæsthesia. Two months before death growing from Jr., Tr. Path.
paralysis of sphincters. Great emaciation. dura mater on Soc. Lond.,
right side in 1881-82,
cervical region. xxxiii. 23, 24.

Upper part of
cervical region.
6 —— Pain in arms. Contracture of fingers of right, then Sarcoma of left E. Long Fox,
left side. Numbness in right foot, then upward, post. aspect of Bris. Med.-
 then left foot. Girdle feeling. Priapism and cord; adjacent Chir. Journ.,
dysuria. Complete anæsthesia, later, up to third cord 1883, i. 100-
rib, with paralysis of legs and paresis of fingers. compressed 106, 2 pl.
Respiration diaphragmatic. Legs very jerky. Later, and soft. Belt of
arms paralyzed. yellow
substance
enveloped cord
to cauda
equina.

Between
cervical bulb
and second
cervical
vertebra.
7 F. 31. Pain, stiffness in neck; pain radiating, aggravated Gumma of dura Charles K.
by jarring. Sudden paralysis of both arms; next mater two Mills,
day paralysis of legs, incomplete. Partial inches long, Philada. Med.
anæsthesia. Marked skin reflexes in legs. Patellar with intercurrent Times, Nov.
reflexes retained, weaker on right than left. hemorrhage; 8, 1879, p.
Dyspnœa. Profuse perspiration. Cardiac flattening and 58.
irregularity. Day before death temperature in right softening of
axilla 100°; left, 102.2°. cord, with
secondary
sclerosis.

From first to
fifth cervical
vertebra.
8 M. 34. Pain in back of neck, with stiffness and torticollis. Gumma of dura Charles K.
Paresis of arms; later, of legs. Anæsthesia of mater; caries, Mills,
arms, then of legs; also paræsthesia of legs. Late probably Philada. Med.
symptoms: shortening and great rigidity of neck, syphilitic, of Times, Nov.
with choking sensation (girdle sensation at neck). vertebræ. 8, 1879, p.
Dimness of vision. Atrophy of arms and less of Abscess. Total 58.
legs. Complete paralysis of arms, almost (almost)
complete of legs. Electro-contractility preserved. transverse
 Violent skin reflexes in legs. Involuntary sclerosis of
evacuations and incomplete priapism. Severe cord.
pains in knees and ankles. No acute bed-sores. Secondary
Paroxysms of dyspnœa. Average temp. for two degeneration.
weeks before death, M. 97.9°, E. 98.3°. Some softening
above and
below tumor.
Cervical nerves
compressed
and atrophied.

From second to
fifth cervical
vertebra; most
in front.
9 M. 43. Pain between shoulders. Numbness in right hand At third cervical E. H. Clark,
and arm, with weakness and swelling. Numbness vertebra, to Bost. Med.
in left arm, which spread over chest and right of front of and Surg.
abdomen. Unable at first to lie down. Felt as cord. Journal,
though encased in armor. Pain in back of neck. Destruction of 1859-60, lxi.
Tongue protruded to right. Exaggerated reflexes opposite 209-212.
in legs. Right arm and leg weaker than left. vertebra.
Vertigo. Dysphagia. Sense of constriction about
neck. Breathing impaired.
10 —— No record of symptoms especially referable to the Cysticercus in Geo. L.
cysticercus. Symptoms of tabes dorsalis. substance of Walton, ibid.,
cord. Lesions of vol. cv. p.
tabes dorsalis. 511.

On level with
third cervical
nerve.
11 M. 25. Pain in back of neck; stiffness. Numbness of left Fibro-sarcoma H. A. Lediard,
hand. Gradual loss of power of left arm. Jerking at level of fourth Tr. Path. Soc.
of arm. Paresis of left leg. Constriction of upper cervical nerves. Lond., 1881-
chest. Right limbs involved, and eventual Cord 82, xxxiii. 25-
complete paralysis of trunk and extremities. compressed. 27.
 Severe headache. Last three days absolute
anæsthesia of arms and legs. No
ophthalmoscopic changes. Constipation and
dysuria.
12 F. 25. Œdema of ankles; pain in legs; afterward Fibroma, size Bernhuber,
numbness, formication, and stiffness of legs. hazelnut, under Deutsch.
Painful contractures in upper extremities. Slight pia mater. Klin., Berlin,
left scoliosis. Abdominal pains. Paresis of arms. 1853, v. 406.
Fingers flexed. Fever. Respiration became Between fourth
involved, and bowels and bladder paralyzed. and fifth
Mind clear. Died in attack of suffocation. Duration, cervical
two years and three months. vertebræ.
13 M. 16. Restlessness. Cramps in pharynx on swallowing. Sarcoma. Adamkiewicz,
Excitability. Delirium. Hallucination. Pain in the Arch. de
neck. On touching neck general cramps. Extending from Neurol.,
Grimaces. Salivation. In three days complete fifth to seventh Paris, 1882,
paraplegia. No fever. Sudden change. Pulse 120. cervical nerve iv. 323-336, 1
Pupils alternating. Blepharospasm. Irregular on antero- pl.
respiration. Pulmonary œdema. Suspicion of lateral face of
hydrophobia, because patient had been with cord,
hydrophobic dog; when offered coffee had compressing
symptoms simulating rabies. left half and
penetrating into
right half, so
that anterior
longitudinal
fissure
described arc of
circle around it.
14 F. Paresis and partial anæsthesia in all limbs for Carcinoma. J. W. Ogle,
many months, most marked on left side. Brain Tr. Path. Soc.
and special senses unaffected. Had a tumor at Tumor caused Lond., 1885,
bottom of right side of neck. Extensive bed-sore. partial 6, vii. 40, 41.
absorption of
sixth cervical
vertebra. Cord
compressed
 and twisted.
Right lateral
aspect
especially
affected. Cord
atrophied.

At level of sixth
cervical
vertebra.
15 F. 34. Pain in right foot, and paresis increasing to Sarcoma, T. Whipham,
paraplegia. Paresis of arms. Contractures of legs. growing from Tr. Path. Soc.
Hyperæsthesia in both legs up to crest of ilia. dura mater; Lond., 1873,
Later, great pain; paralysis of sphincters. Bed- nerves passing xxiv. 15-19.
sores. through and
over tumor.
Cord congested
and pushed to
one side. Thin,
but not
softened.
Growth
resembled
psammoma.

Between sixth
and seventh
cervical nerves
of left side.
16 M. 57. Pain in right arm. Numbness in hand, and Myxoma from Pel (P. K.),
paresis. Paresis and coldness of left leg. Some arachnoid. Cord Berlin. Klin.
anæsthesia and wasting of right leg. Later, compressed Wochensch.,
paraplegia. Diminished reflexes. Contractures. and softened 1876, xiii.
Constriction sense about legs and abdomen. on right 461-463.
Triceps, deltoid, and serratus magnus of right postero-lateral
side paralyzed. Incontinence of urine, difficult side.
defecation, decubitus, fever. Abdominal muscles Secondary
 paralyzed. Later, other muscles of arms degeneration.
paralyzed. Complete anæsthesia of legs. Some œdema
Dyspnœa, œdema of lungs. of brain.

At sixth and
seventh
cervical
vertebra on
postero-lateral
surface of cord.
17 F. 35. First, pain in right arm, weakness in right hand. Spindle-cell E. Long Fox,
Then paralysis almost complete in arms, and sarcoma, Bris. Med.-
impaired sensation. In legs paralysis complete, springing from Chir. Journ.,
sensation impaired. Alternate incontinence and arachnoid and 1883, i. 100-
dysuria. Ankle clonus and increased knee-jerks destroying cord 106, 2 pl.
and plantar reflex. Tapping biceps causes reflex by pressure,
in little and ring fingers. No atrophy or bed-sores. except posterior
Cold on one side, hot on other. Pain and little columns. Cord
swelling over sixth cervical vertebra. No eye below tumor
symptoms. Brain clear. Inability to turn head. soft.
Before death respiratory paralysis and bed-sores.
Duration, fifteen months. At sixth cervical
vertebra.
18 M. 50. Paresis in right arm. Stiffness in neck and back. Glioma in right Schueppel,
Paralysis of all extremities gradually developed. half of cord. Old Arch. d.
hemorrhages in Heilk., viii.
adjacent parts Bd., 1867
and in medulla (quoted by
oblongata. A Rosenthal).
more recent
hemorrhage in
dorsal cord.

In lower
cervical region.
19 —— Coldness, numbness, violent pains, first in left Tubercle, large Chvostek,
arm, later in both legs. Paralysis of all limbs and as hazelnut. Med. Press,
 muscles of trunk. Atrophy. Reactions of Consecutive 33-39, 1873
degeneration. Violent leg reflexes. myelitis of (quoted by
adjacent parts Rosenthal).
and left anterior
horn.

In lower
cervical region.
20 M. 45. Interscapular pain. Chest-pressure and dyspnœa. Phlegmon of Mankopff
Paræsthesia and pain in legs. Spastic paralysis. dura mater, (E.), Berl.
Difficulty in stools; bloody urine and dysuria. compressing Klin.
Œdema of legs. Bed-sores. Kypho-scoliosis. Pain cord. Some Wochensch.,
on pressure over spine. Paralysis of left leg, infiltration of 1864, i. 33-
paresis of right, some anæsthesia of both. tissues of throat 46, 58, 65,
Broncho-pneumonia, fever. and mediastinal 78.
space.

From seventh
cervical to
second dorsal
vertebra.
21 M. 22. Pain in back and side of neck and in limbs. Tumor of H. C. Wood,
Marked pain in sternal region on coughing. membrane. “Proceedings
Pressure and jarring cause pain. Rapid loss of Cord beneath of
power in both arms. Feeble and slow movements compressed Philadelphia
of thighs, legs, and feet. Right deltoid and flexors and Neurological
of fingers much wasted. No paralysis of face. degenerated. Society,”
Knee-jerks exaggerated. Later, complete Medical
paralysis, including bladder and rectum. Lower cervical News, vol.
and upper xlviii. No. 9,
dorsal region. Feb. 27,
1886.
22 F. 50. Pain in neck, shoulders, and chest. Stiffness of Secondary Gull, by
neck, back, and arms. Chest fixed; breathing cancer of Wilks, in
diaphragmatic. No paralysis or altered sensation. vertebræ. Lect. on Dis.
of Nerv.
Cervical region. System.
23 F. 40. Severe pain in back. At height complete paralysis Fibro-cyst on Risdon
in legs, some paresis in arms. Variable right side, Bennett, Tr.
anæsthesia. Girdle sensation and mammary pain. between cord Path. Soc.
Lively and distressing reflexes. Contractures in and dura, and Lond., 1855-
legs. Bed-sores and paralysis of sphincter of between 56, vii. 41-45.
bladder. Toward close rigors (pyæmia?). anterior and
posterior
nerves.

Top of dorsal
region.
24 M. 30. Cough, dyspnœa, wasting, simulating phthisis. Tumor, size of Gull, in Guy's
Pain in back of neck and shoulders. Pain in joints; hazelnut, inner Hosp. Rep.,
paresis of legs and bladder. Pain in chest. anterior surface (quoted by
Paresis of arms. Later, increased paralysis, bed- of dura mater. Wilks in
sores, sweating. Flattening and Lectures on
softening of Dis. Nerv.
cord. Syst., p.
264).
Top of dorsal
region.
25 F. 43. Pain in shoulders, chest, and sides. Contractures Fibro-nucleated Gull, by
of legs; heels to nates. No anæsthesia. Later, tumor from Wilks, ibid.
retention of urine and bed-sores. Incessant pain inner surface of
in back and abdomen. dura mater.

Opposite third
dorsal vertebra.
26 F. 43. Pain in chest and shoulder, then in legs. Paresis Fibro-nucleated Wilks, Trans.
of legs. Contractures and jerking of legs. Spasm tumor, size of a Path. Soc.
of abdominal muscles. No anæsthesia. Paresis of bean, from dura Lond., 1855-
bladder and rectum. Wasting and bed-sores. mater. Cord 56, vii. 37-40.
Finally, paresis increased, but never complete compressed
paralysis. Duration, nine months. backward, and
softened.
 Opposite third
dorsal vertebra.
27 M. 24. Paraplegia. Depressed reflexes; girdle symptom. Probable B. G.
Partial anæsthesia. Dysuria. Vomiting. Pulse gumma. McDowell,
weak and intermittent. Partial recovery from M.D., Dubl.
paralysis, and anæsthesia in left leg, and reflexes Middle dorsal Q. J. Med.
in right foot regained. Later, complete paraplegia, region. Sci., 1861,
anæsthesia, and bed-sores. Duration, five xxxii. 299-
months. 303.
28 F. 44. Paresis in legs. Spine hypersensitive and Alveolar G. W. H.
inflexible; least attempt at bending causes great sarcoma. Kemper,
cervico-brachial pain. Paræsthesia; sense of Journ. Nerv.
falling out of abdominal viscera through Eighth and and Ment.
abdominal walls. Pains in extremities increasing, ninth dorsal Dis., xii. No.
and involving right shoulder, intercostals on both vertebræ. 1, Jan., 1885.
sides, and lumbar region. Paralysis of right arm
(first); complete paralysis of leg. Excessive spinal
tenderness. Loss of sensation (partial) in legs,
body, and right arm. Later, dyspnœa, then
dysuria, then complete inability to empty bowels
or bladder. Great tympanites. Girdle sense above
umbilicus, and finally complete paralysis and
anæsthesia below this band. Sense of twisting of
legs and feet, so that latter seemed close to face.
Œdema. Later, paresis of left arm. One small
bed-sore.
29 F. 42. Projection of seventh, eighth, ninth, tenth, and Round-celled E. Long Fox,
eleventh dorsal vertebræ. Numbness below sarcoma. The Brit. Med.
ankles, and early girdle sensation. Peronei and anterior Journ., 1871,
anterior tibial muscles first involved; then all leg- columns soft p. 566.
muscles, then sphincters, then arms. Died in a fit. opposite tumor.
Bodies of
seventh, eighth,
ninth, and tenth
vertebræ soft.

Opposite
 seventh, eighth,
ninth, and tenth
dorsal
vertebræ.
30 F. —. Ill-defined hemiplegia; later, paraplegia, with Gumma and Taylor,
contractures and rigidity. syringo-myelus. Lancet, 1883,
Small cavities p. 685.
in anterior
cornua.

At ninth dorsal
vertebra
anterior aspect.
31 M. 7. Paraplegia, except adductors and rotators of Tubercle (?). Geoghegan,
thigh. Reflex contractures; most intense from Dublin Med.
irritation of penis and scrotum. Rigidity of legs. Cord soft for Press, 1848,
Complete anæsthesia of lower half of body. Later, two inches. xix. 148-151.
anuria, incontinence of feces. Anal sphincter
reflex; figured stools. Cystitis. Pain on percussion Tenth dorsal
in dorsal region. Pain in back. Complete vertebra.
paraplegia. Very late, brain symptoms. Duration,
nine months.
32 F. 46. Fixed pain in left iliac region. Paresis in left leg, Fibroma (?) William
increasing to paraplegia. Formication. Girdle from inner Cayley, Tr.
sensation. Incomplete, increasing to complete, surface of dura. Path. Soc.
anæsthesia of legs. Spontaneous twitchings. Cord hollowed Lond., 1864-
Bladder and sphincter ani paralyzed. Bed-sores. out and 65, vol. xvi.
Duration, one year. softened. 21-23.

Interval
between tenth
and eleventh
dorsal
vertebræ.
33 M. 30. Hyperæsthesia; later, anæsthesia in legs; then Tubercle size of Chvostek,
complete paraplegia. pea. Adjacent Med. Presse,
myelitis. In 33-39, 1873
 lower dorsal (quoted by
region. Rosenthal).
34 M. 31. Ataxia; stiffness of legs and cramps in abdomen Myxoma of Shearman,
and legs. Slight nystagmus. Difficulty in forming dura mater 3 Lond. Lancet,
words. Ataxia of arms. Slight wasting of legs, inches long. vol. ii. 1877,
especially of left. Lumbar pains; abdominal Dura mater of p. 161.
cramps. Dysuria. Impotence. Later, increased brain contained
spastic state of legs. Mind depressed and fluid and lymph.
emotional; attempts at suicide. Anuria. Bed-sores.
Urine albuminous. Duration, one year. Dorsal region,
left side.
35 F. 50. Pains in limbs (thought to be rheumatic). Paresis Cancer of Gull, by
in legs. Hyperæsthesia in right leg; burning pains vertebræ Wilks, Dis.
alternating with sense or coldness. (sarcoma?). Nerv. Syst.

Dorsal region.
36 F. 35. Paresis of left leg; soon of right leg. Pain in back Tumor, osseous H. Ewen, Tr.
and left side. Tonic spasms of legs. Darting pains or fibrous, Path. Soc.
in knees. Partial anæsthesia. Exalted plantar three-fourths of Lond., 1848-
reflexes. Dysuria. Later, complete paraplegia and an inch long, 50, i. 179.
anæsthesia; violent reflexes; severe pain in back. growing from
Bed-sores. Duration, seven and a half years. dura mater.
Cord flattened,
and softened
below tumor.

Lower part of
dorsal cord.
37 F. 28. Weakness in legs. Aching and shooting pains in Tubercle the S. O.
legs. Numbness and formication. Slight spasm in size of cherry, Habershon,
legs. “Felt as if ground was some distance below which had M.D., Guy's
feet.” Œdema of ankles. Later, numbness almost Hosp. Rep.,
extended to abdomen. Paralysis of bladder. obliterated London,
Hyperæsthesia in right leg. Obstinate cord. Tubercles 1872, 3d S.,
constipation. Bed-sores. Some paralysis of in lungs, xvii. 428-436.
respiratory muscles. Duration, fourteen months. bowels, and
uterus. Bed-
 sore had
opened spinal
canal.

Lower part of
dorsal cord.
38 M. 63. Progressive paresis of left leg for five years. Right Psammoma Charcot,
leg then paretic. Paralysis then in left leg. Rigidity adherent to Arch. de
on extension of right leg. Paroxysms of clonic dura mater. Physiol.,
spasms in right leg. Joint pains, sciatic pains. In Cord softened. Paris, 1869,
left leg, hyperæsthesia, in right leg, anæsthesia. Ascending ii. 291-296.
Plantar reflex retained; other reflexes degeneration in
exaggerated. Diplegic contractions in right leg posterior
from irritation in left. Late symptoms: purulent columns, and
urine, with retention; chest and lumbar pains like descending
bone pain; extension changed to flexion; swelling degeneration of
of legs and ecchymosis; sacral and other lateral columns.
eschars.
In dorsal region
just above
lumbar
enlargement,
anterior left
side.
39 M. 20. Paralysis of lower extremities; tremor; Organized C. B.
exaggerated reflexes, hyperæsthesia of trunk; blood-clot Nancrede,
bed-sores. Œdema of feet. Fever. Pus in urine. exterior to dura Am. Journ.
mater. Cord Med. Sci., O.
compressed S., lxi. 156.
and softened.

Opposite lower
dorsal and
upper lumbar.
40 F. 38. Pain around abdomen, in back, and legs. Hydatid cysts of S. Wilks, Dis.
Paraplegia. Anæsthesia and tingling of feet and vertebræ (?) Nerv. Syst.,
legs. Paralysis of bladder. and spin. canal. p. 265.
 Lower part of
spinal canal
(probably
lumbar region).
41 F. 23. Bronzing of skin for two years; then headache, Tumor, W. H.
giddiness, fever. Choreic movements in left arm, consisting of Broadbent,
then in leg, then general. Bronzing increased. granular matter, Trans. Path.
Vomiting after meals. Duration, two years and two with a few Soc. Lond.,
months. nerve-fibres 1861-62, viii.
and cells, 246.
springing from
centre of cord
backward to
posterior
fissure. Cord
slightly
widened.
Suprarenal
capsules large
and nodulated.

Lumbar
enlargement.
42 10 Twitching and convulsive movements of right leg. Tumor outside Arthur
ms. After removal of exterior tumor the movements of sacrum, and Johnson,
ceased. Child died of peritonitis. also protruding ibid., 1856-
through sacral 57, viii. 28,
opening. 29.
Reported to
have been
behind and
pressing upon
cord (?). Fatty
growth within
membranes.
43 F. 54. Paresis, first of left arm and leg; then paralysis of Hydatid cyst. H. S. Wood,
 these and of right arm and leg. Pain in back and Cyst also in Australian
hips early; then, suddenly, darting pains and liver. Fluid Med. Journ.,
incontinence of urine. Paræsthesia of left arm and beneath 1879, N. S. i.
leg; no anæsthesia. Coma. membranes of 222.
cord and brain.

At first and
second left
sacral foramen,
opposite last
lumbar and
upper three
sacral
vertebræ.
44 M. 46. Fibrillary twitching. Increased patellar reflexes. Glioma. Lachman,
Paræsthesia and hyperæsthesia in legs, Arch. f.
disappearing. Constriction of chest (?). At filum Psychiat.,
Headache. Dysuria for two years. Straining at terminale, Berl., 1882,
stool. Indigestion. Bloody vomiting. Cardiac upper part. xiii. 50-62, 1
palpitation; intracardial murmurs; slow pulse. pl.
Swollen inguinal glands. Variations in
temperature. Bed-sores.
45 M. 38. Pain in legs. Œdema. After two years could not lie A lobulated W. W. Fisher,
down: rested on hands and knees. Paralyzed in tumor from pia Tr. Prov. M.
legs; pain in seat. Anæsthesia in legs, not mater at lower and S. Ass.,
complete in right. Paræsthesia in left. Dysuria and end of spinal 1882, x. 203-
constipation. Before death had incontinence with canal, 208.
hæmaturia, and was able to lie down. surrounded by
nerves of cauda
equina.
Structure not
made out.

At cauda
equina.
46 —— This case had symptoms of posterior spinal Myo-lipoma W. R.
sclerosis, which possibly had no relation to attached to Gowers, Tr.
 growth, according to reporter. conus Path. Soc.
medullaris. Lond., 1875-
Crescentic, 76, xxvii. 19-
clasping cord 22.
from anterior to
posterior
fissure. Nerve-
roots of cauda
equina
imbedded in it.
Contained
striated
muscular fibres.
47 Und. Spina bifida (?); hydrocephalus; convulsions, Congenital W. F. Jenks,
1 yr. bloody stools; partial paraplegia. (Above sacral neuroma M.D., Trans.
symptoms came on after closing of sacral amyilinicum. Path. Soc.
opening by surgical operation.) Philada.
(1871-73),
1874, iv. 190-
192.
48 M. 30. Pain in back; abdominal girdle sensation. Pain in Aneurism, Wilks, Dis.
legs; paraplegia; nearly complete anæsthesia; eroding vert. Nerv. Syst.
paralysis of bladder; bed-sores. and
compressing
cord. Location
not given.
49 M. 54. Paralysis of both legs, of sphincter ani, and of Gumma from Delafield, N.
bladder; urine alkaline, with pus and blood. inner layer of Y. Med. Rec.,
Partial anæsthesia. Pyonephritis. dura mater and 1875, x. 131.
involving pia
mater. Location
not given.
50 Still- —— Tumor, size of Virchow,
born. head of child Monatschr. f.
two years old, Geburtsk.,
projected Berl., 1857,
between legs ix. 259-262.
 from spinal
column. Nerves
of cauda equina
over anterior
part. Some
bone in tumor
(dermoid
cyst?).

INFANTILE SPINAL PARALYSIS.

BY MARY P. JACOBI, M.D.

SYNONYMS.—Essential paralysis of childhood (Rilliet and Barthez);

Myogenic paralysis (Bouchut); Acute fatty atrophic paralysis
(Duchenne); Atrophic paralysis (Ferrier); Acute anterior poliomyelitis
(Kussmaul, Erb, Seguin); Regressive paralysis (Barlow);
Tephromyelitis (Charcot).

DEFINITION.—Of all the titles which have been given to the disease it
is our purpose to describe, two alone may be considered
irreproachable. In the present state of our knowledge it is
unnecessary to argue that this disease is not essential—i.e. destitute
of characteristic anatomical lesions. Neither can the theory of its
myogenic origin be maintained; nor even is fatty degeneration
invariably present in the paralyzed muscles. Finally, the disease
cannot longer be regarded as peculiar to childhood,1 since cases in
adults have been in these last years quite numerously reported2—
four with autopsies demonstrating the identity of the lesion. But there
are two definitions in our list of synonyms which embrace between
them the most striking characteristics of the disease, yet contain no
error of fact. Atrophic paralysis describes at once the two most
salient symptoms; acute anterior poliomyelitis defines at once the
seat and nature of the lesion, classes it with the systematic
diseases3 of the spinal cord, and notes the peculiarity in the mode of
invasion by which it is so remarkably distinguished from nearly all the
organic diseases of this centre.
1 W. H. Barlow, On Regressive Paralysis, 1828. See Brain, April, 1879.

2 In Dec., 1873, I quoted 14 cases of adult spinal paralysis, as follows: Duchenne, 4

cases; Charcot and Petitfils, 3; Moritz Meyer, 2; Bernhardt (Archiv Psych., 1873), 1;
Cumming (Dublin Quart. Journ., 1869), 1; Lucas Championnière (by Hallopeau,
Archives gén., 1861), autopsy, 1; Gombault (Archives de Psych., 1873), 1; personal,
1.

In 1874, Seguin published a summary of all the foregoing cases except the last, and
added 6 personal observations, also 3 from Duchenne and 1 from Hammond. In the
enlarged edition of his essay in 1877, Séguin increased the list to 45—by new
personal cases, 3; cases related by Frey (Berlin. Wochens., 1874), 4; cases by Erb
(Arch. f. Psych. u. Nervenkrank., v.), 4; case by Cornil and Lépine (Gaz. méd., 1875),
autopsy, 1; case by Soulier (Lyon méd., 1875), 1; case by D. H. Lincoln (Boston Med.
and Surg. Journ., 1875), 1; case by Lemoine (Lyon méd., 1875), 1; case by George
M. Beard, 1; case by Leyden (Klinik Ruckenmarks Krankheiten) Bd. iv. 1; case by
Hammond (6th ed. Treatise), 4; case by Courty (Gaz. méd., 1876), 1; case by
Dejerine (Arch. de Phys., 1876), 1.

To these may be added—case by Goltdammer (Berl. klin. Wochen., 1876), 1; case by

Webber (Trans. Amer. Neurol. Ass. for 1875, vol. i.), autopsy, 1; case by Klose (Diss.