Professional Documents
Culture Documents
PDF Digital Orthopedics Guoxian Pei Ebook Full Chapter
PDF Digital Orthopedics Guoxian Pei Ebook Full Chapter
https://textbookfull.com/product/textbook-of-orthopedics-and-
trauma-4-volumes-gs-kulkarni/
https://textbookfull.com/product/the-basic-economic-system-of-
china-changhong-pei/
https://textbookfull.com/product/additive-manufacturing-
developments-in-training-and-education-eujin-pei/
https://textbookfull.com/product/chinese-dream-and-practice-in-
zhejiang-economy-changhong-pei/
ORTHOPEDICS OF THE UPPER AND LOWER LIMB 2nd Edition K.
Mohan Iyer
https://textbookfull.com/product/orthopedics-of-the-upper-and-
lower-limb-2nd-edition-k-mohan-iyer/
https://textbookfull.com/product/pediatric-orthopedics-symptoms-
differential-diagnosis-supplementary-assessment-and-
treatment-1st-edition-jan-douwes-visser/
https://textbookfull.com/product/pediatric-orthopedics-and-
sports-medicine-a-handbook-for-primary-care-physicians-2nd-
edition-amr-abdelgawad/
https://textbookfull.com/product/thermal-energy-storage-analyses-
and-designs-pei-wen-li-and-cho-lik-chan-auth/
https://textbookfull.com/product/brinker-piermattei-and-flos-
handbook-of-small-animal-orthopedics-and-fracture-
repair-5e-fifth-edition-decamp/
Guoxian Pei
Editor
Digital Orthopedics
123
Digital Orthopedics
Guoxian Pei
Editor
Digital Orthopedics
Editor
Guoxian Pei
Orthopedic Hospital
Xijing Hospital, Fourth Military Medical University
Xi An, China
The print edition is not for sale in China Mainland. Customers from China Mainland please order the print book from:
People’s Medical Publishing House.
This Springer imprint is published by the registered company Springer Nature B.V.
The registered company address is: Van Godewijckstraat 30, 3311 GX Dordrecht, The Netherlands
Informative Abstract
The book is the second edition of Digital Orthopaedics. The edition is divided into three parts,
which deal with the basis, research, and applications of digital orthopedics, with a total of 18
chapters. Based on the first edition, the second edition adds “CAD and CAM,” “additive manu-
facturing technology,” “reverse engineering technology,” and other new technologies which
have been developed rapidly over the years and have been actually applied clinically. The big-
gest feature of the edition, as a whole, is its introduction of application research, actual applica-
tion, and latest development of digital orthopedics in clinical medicine over the years. These
include research of 3D visualized application of digital tissue flap and peripheral nerves; appli-
cation of 3D-printed orthopedic models; application of navigation template; application of
computer-aided navigation orthopedic technique in fracture repair, spine surgery, hip surgery,
knee-joint replacement, and bone tumor surgery; and actual application of robots in traumatic
orthopedics, spine surgery, and joint surgical operation. The content of the book is novel and
complete, fully presents the latest developments of digital orthopedic techniques, and is appli-
cable to orthopedists, postgraduates, medical postgraduates, and iconography personnel for
reading and reference.
v
Preface
The development of digital medicine has promoted the rapid development of digital technology
in orthopedics. Over the past 20 years, digital orthopedic techniques have considerably devel-
oped. In 2006, we presented the concept of “digital orthopedics,” wherein we classified, refined,
and promoted the digital technical attributes of the field of orthopedics with great variety and
broad content, achieving a higher and faster digital technology development. Digital orthopae-
dics is a new interdisciplinary subject which combines digital technology with orthopaedics. It
involves anatomy, stereogeometry, biomechanics, materials science, information science, elec-
tronics and mechanical engineering. Digital orthopaedics has a wide range. Any technology
used in orthopaedic research, diagnosis, treatment, rehabilitation and orthopaedic education by
digital means belongs to the scope of digital orthopaedics. The application of digital orthopae-
dic technology will certainly promote the digitalization, individualization, visualization, virtu-
alization, precision and intellectualization of orthopaedic diagnosis and treatment in the future,
and further achieve standardization and standardization. “Digital orthopedics,” as an important
branch and composition of orthopedics, has become one of the subspecialties which developed
most rapidly and has the highest level of new technical content in orthopedics with huge devel-
opment prospect.
In 2008, we compiled and published Chinese version of Digital Orthopaedics. In the past
decade, digital orthopaedics has developed rapidly. New technologies such as 3D printing
technology, VR, AR, MR, robotics technology and artificial intelligence emerge in endlessly,
which promote the individualization and precision of Orthopaedic Surgery Diagnosis and
treatment, and lead and promote the further development of orthopaedics. In order to introduce
the latest theories, knowledge, and technology of digital orthopedics in a timely manner and to
promote its clinical application more effectively, relevant experts were asked to compile the
English edition of Digital Orthopaedics, which is the first monograph of digital orthopeadics
in the world. In order to expand the readership, the Chinese version of the book will be pub-
lished in China at the same time.
The edition is divided into three parts, which deal with the basis, research, and applications
of digital orthopedics, with a total of 18 chapters. Based on the first Chinese edition, this book
adds “CAD and CAM,” “additive manufacturing technology,” “reverse engineering technol-
ogy,” and other technologies which have been developed rapidly over the years and have been
actually applied clinically. The biggest feature of the edition, as a whole, is its introduction of
application research, actual application, and latest development of digital orthopedics in clini-
cal medicine over the years. These include research of 3D visualized application of digital flap,
muscle flap and bone flap, perforator flap, and peripheral nerves; application of 3D-printed
orthopedic models; design and application of surgery navigation template, preoperative plan-
ning, and virtual surgery of computer-aided orthopedics; application of augmented reality
technology in digital orthopedics; application of computer-aided navigation orthopedic tech-
nique in fracture repair, spine surgery, hip surgery, knee-joint replacement, and bone tumor
surgery; and actual application of robots in traumatic orthopedics, spine surgery, and joint
surgical operation. Application of these new technologies expands the range of orthopedic
surgery, promotes its quality, and fully embodies the latest treatment concept and requirement
vii
viii Preface
Xi’an, China
November 2018
Foreword for the English version of Digital
Orthopaedics
Guangzhou, China
November 2018
ix
Foreword
xi
xii Foreword
biomechanical feedback technology simulate the operation process vividly. It can not only
improve the success rate of operation but also train young doctors in surgical skills more quickly
and effectively and improve the efficiency and quality of personnel training.
With the improvement of people’s living standards and the gradual improvement of food and
clothing, the four important elements of people’s lives are evolving into five elements: clothing,
food, housing, transportation, and medicine. People’s demand for medical services is increasing
day by day. This brings us both opportunities and great challenges. At present, there are still many
problems in the field of digital orthopedic medicine in China. Medical and engineering techni-
cians at all levels do not understand each other’s needs, and there is still a shortage of personnel
who can grasp the knowledge and skills of both medical and engineering fields. Although some
research institutes have been set up and some theoretical works have been published, they are still
not mature enough, and many aspects are still in the initial stage. However, the twenty-first cen-
tury is bound to be a new era of great development of digital medical technology, which forms a
huge gap between subjective ability and objective needs.
The publication of the English version of Digital Orthopedics will contribute to the introduc-
tion and popularization of new achievements, knowledge, and communication between doctors
and workers in the field of digital orthopedics. Thanks to the engineering and medical experts for
their hard work in writing this book!
xiii
xiv Brief Introduction of Guoxian Pei
six national invention patents. He was successively recognized as belonging to “The First
Batch of National Young and Middle-Aged Star of Medical Science and Technology” and “The
First Batch of Person Selected for New Century Talents Project” and was acknowledged as
“National Young and Middle-Aged Scientific and Technical Expert with Outstanding
Contributions,” among other titles. In 2000, he was awarded the “Military Scientific and
Technical Star.” In 2001, he was acknowledged as belonging to the first batch of “Academician
Alternatives” by the General Logistics Department. In 2002, he was awarded the “Military
Significant Contribution Award on Professional Skill.” In 2003, he was awarded “Type I Post
Allowance of Excellent Military Talent.” In 2006, he was approved by the Organization
Department of the CPC Central Committee as “senior expert contacted directly by the Central
Government.” In 2014, he led the orthopedic team and won the award of “Three-Star Talent
Innovation Team” issued by the General Logistics Department.
Contents
xv
xvi Contents
1 he Establishment of Digital
T dation for China’s “Digital Human” and digital anatomy in
Orthopedics general, in which three-dimensional anatomy is created from
the two-dimensional. The development of digital anatomy has
Since the 1970s, anatomy and clinical practice have greatly rapidly improved traditional teaching, training, and operation
influenced the field of clinical medical research, resulting in designs. The technology of digital medicine has provided a
the development of many new surgical methods. The simula- new development platform for clinical medicine with three-
tion of surgical procedures has been used for clinical training, dimensional anatomy and clinical training simulations. Virtual
teaching, and operation design. Traditional clinical teaching human technology and digital anatomy have expanded into a
is completely based on the two-dimensional plane, which can variety of fields, including bone science. Using digital virtual
be intuitive. However, a three-dimensional dynamic display technology, bone science and clinical research can be organi-
has been used recently, allowing visualization of the vascular cally combined with digital technology, creating a new, multi-
tissue, nerve tissue, and anatomical structures. disciplinary branch of bone science—digital bone science.
“Visual Human Plan” (1989) first promoted the emergence The concept of digital orthopedics is based on digital
and development of digital medical imaging using modern medicine, Digital technology, digital medical equipment,
information technology in medical disciplines. This was a new computer network platforms, and medical professional soft-
frontier that had a far-reaching impact on the development of ware provided the basis for a digital, three-dimensional
science and technology. Digital technology uses dynamic orthopedic anatomy atlas. In addition to the treatment of
three-dimensional images to supplement traditional two- patients, orthopedic data acquisition, storage, transmission,
dimensional medical images. For example, a computed tomog- and processing into digital orthopedic databases have
raphy (CT) image is a three-dimensional reconstruction in allowed us to achieve two other goals: information and
rows, which can accurately show the complex structure of bio- resource sharing. Orthopedic surgeons can consult the digital
logical tissue. CT allows for arbitrary rotation and slicing in database for the latest technical information, then share their
observation and procedures; it also can measure a three-dimen- experiences and achievements to improve the data. Also
sional structure for reconstructions, including the length, area, through the computer network, the technology can be used in
volume, and angles, thus providing precise anatomic parame- teaching to demonstrate techniques and establish an orthope-
ters. Such imaging can be used for clinical diagnosis, auxiliary dic simulation classroom.
surgery, and surgical simulation, among others [1].
Zhong Shizhen, a professor at the Southern Medical
University (formerly the First Military Medical University), 2 The Scope of Digital Orthopedics
launched a “virtual Chinese human” model as part of his
research on the human body. This provided the necessary foun- The rapid development of computer technology, image pro-
cessing technology, and medical physics have improved
medical diagnosis and treatment in all fields of medicine.
Surgeons use digital orthopedics along with modern diag-
G. X. Pei (*)
nostic imaging, surgical techniques, and advanced materials
The First Affiliated Hospital of Air Force
Medical University, Xi’an, China to ensure a patient’s safety and accurately perform minimally
invasive surgery. The computer navigation system is a new
Y. Z. Zhang
The Affiliated Hospital of Inner Mongolia advancement in the field of digital medicine, allowing for
Medical University, Hohhot, China speed, accuracy, and flexibility in minimally invasive and
highly unified surgery. This technology has been gradually application and development in China.” In 2007, the
introduced to spinal surgery from its initial applications in research emphasis on digital humans was transferred into
neurosurgical, orthopedic, and ear-nose-throat procedures. the digital medicine sphere when a exploratory conference
Thus, a new field of computer-aided navigation surgery was (“The First China Digital Medical Seminar”) was held in
created. The successful use of computer-assisted navigation Chongqing. To plan, communicate, and coordinate national
in orthopedics also fully embodies the advanced nature of academic activities, a Chinese digital medical research
digital bone science, and its applications will only become group was established in conjunction with national aca-
more extensive [2]. demic groups. In 2008, the Chinese Academy of
In recent years, rapid prototyping and detection technol- Engineering’s first “Digital Medical Conference” was held
ogy, such as spiral CT and magnetic resonance imaging in Beijing. In May 2011, a digital medicine branch of the
(MRI) with three-dimensional image reconstruction func- Chinese Medical Association was founded in Chongqing.
tionality, have become more widely used in orthopedics, car- Then, in November 2011, the Chinese Medical Association
diovascular surgery, ear-nose-throat surgery, forensic formally established a medical engineering branch of the
medicine, tissue engineering, and oral surgery, among oth- digital orthopedic group in Xi’an.
ers. By combining rapid prototyping technology and reverse
engineering science in bone applications, we can broaden the
research scope in the field of bone science. Reverse engi- 4 he Development Trends in Digital
T
neering is used in mechanical engineering, biology, materi- Orthopedics
als science, engineering, and various field of medicine; its
applications will continue to expand with the development of Orthopedic virtual simulation systems allow for image pro-
biological materials science and medical imaging technol- cessing, virtual reality, electronic communications, and
ogy. At present, the medical applications of reverse engineer- orthopedic surgery planning using rational, quantitative, and
ing have been integrated with computer software, such as individualized interactive systems. The design, simulation,
Geomagic, Imageware, and Rapidform. Its applications in intervention, and evaluation of orthopedic surgery can be
digital orthopedics are mainly in the following areas: (1) achieved using visual, auditory, touch, and other sensory
design and production of implants; (2) design, teaching, and experiences. The orthopedic surgeon can produce immersive
evaluation of complex surgical procedures; (3) creation and interactive visual simulations, including computer graphics,
analysis of a three-dimensional model of human bone with image processing, and pattern recognition. The fields of
body movement mechanics [3]. intelligent technology, sensor technology, language process-
Through rapid prototyping technology and reverse engi- ing, audio technology, network technology and many other
neering, we can build an orthopedic surgery planning system sciences will contribute to the further development and
to design and evaluate orthopedic surgical procedures. This application of digital medicine. Using a virtual simulation
system can aid in the setup of internal fixation devices, the system, the orthopedic physician can effectively communi-
creation of a three-dimensional classification database for cate with patients, inform patients on the basic procedures
fractures, the three-dimensional simulation of the recon- they will undergo, and thus reduce the number of patients
struction and restoration of various fracture types, and the experiencing stress about their surgeries.
choice of internal fixators. Although digital technology and digital medicine provide
a good theoretical basis for the realization of the concepts of
digital bone science, many research institutions have
3 he Present Situation of Digital
T obtained good results using the Visible Human Project data-
Orthopedics set in their digital technology research and development. In
digital anatomy, institutes have developed virtual simula-
Held in Beijing in 2001, the 174th Xiangshan Science tions of adult males in a systemic model (Visible Photographic
Conference was attended by more than 40 scholars from dif- Man, VIP-Man); the Voxel-Man system (developed by the
ferent disciplines, who gathered to discuss the state of digi- University of Hamburg, Germany); an ultrasound, CT, and
tal research in China. The theme of that meeting was Chinese MRI dataset to build individualized numbers (France); a
digitized virtual human technology. In 2003, at the 208th computerized virtual human body system (jointly developed
Xiangshan Science Conference, with support from the by the British PA consulting company with Orme Scientific);
national “863-Project” and the National Natural Science nerve function reconstruction of the brachial plexus (Fudan
Foundation of China, the human body database with Chinese University); virtual human organs for study (China
characteristics was presented. Early versions of virtual sim- University of Science and Technology); and reconstructions
ulation software were also presented. This second confer- of the uterus, liver, kidney, and bones (Southern Medical
ence theme was “digitized virtual human studies for University, Third Military Medical University, Capital
1 The Establishment of Digital Orthopedics 5
Fig. 1.1 Upper limb artery and three-dimensional reconstruction of the latissimus dorsi myocutaneous flap in digital bone anatomy
University of Medical Sciences). Other surgical procedure Digital bone surgery may apply a three-dimensional
design and s imulation research has been based on microcom- fracture or bone disease database model, intuitive analysis
puter technology, including a force feedback abdominal sur- and observation of a fracture, fracture type classification
gery simulator (C.S. Tseng), knee surgery training in a virtual using a database, and digital operation design and simula-
environment (A.D. McCarthy), a force feedback endoscopic tion to improve outcomes and enrich the database (Figs. 1.2
surgery simulator (C. Baur), and a complete atlas of the brain and 1.3) [6–8].
for surgical planning (Harvard University). In China, the Digital bone surgery combines diagnosis tools, advanced
design and implementation of virtual surgery has been suc- equipment, and the surgeon in an organic union to ensure a
cessfully applied in bone, liver, neurosurgery and other patient’s safety while providing accurate and minimally
areas, such as virtual vertebral spine bone resection (Southern invasive surgery. This is a kind of human–computer interac-
Medical University) and functional simulation and clinical tion system, which reasonably and quantitatively uses multi-
applications of the eye (Xiamen University). Digital technol- variate data and navigation systems for surgical planning,
ogy is new to the field of orthopedics, however,. intervention, and evaluation to promote the development of
Digital orthopedics has three aspects: digital bone anat- minimally invasive surgery. Virtual reality technology and
omy, surgery, and virtual simulation systems. Three- virtual teaching are perhaps the most important applications
dimensional digital bone anatomy is the basis of virtual of orthopedic simulation training. Use of the technology
teaching in orthopedics, allowing students to interactively allows the physician to become immersed in the virtual
browse a model of the human body. With digital bone anat- scene, using several senses to learn about a surgical
omy, users can freely explore the body without causing any procedure.
external disturbances to anatomical structures, learn about Currently, multilayer spiral CT three-dimensional recon-
anatomy more quickly, and fully visualize individual ana- structions provide good images, but the reconstruction results
tomical characteristics (Fig. 1.1) [4, 5]. can only be displayed on a computer; they cannot be exported
6 G. X. Pei and Y. Z. Zhang
Fig. 1.2 Design and implementation of ankle fracture surgery in digital orthopedic surgery
Fig. 1.3 Design and implementation of spinal corrective surgery in digital orthopedic surgery
to other interfaces. The use of three-dimensional reconstruc- that has flexibility, a sense of tension, and authenticity (e.g.,
tion software in Initial Graphics Exchange Specification blood, sound). The simulation can provide a realistic surgical
(IGS) and stereolithography (STL) formats allow the model environment with multiple angles and a full view of the
to be created and then imported into other computer-aided implant position, location of the injury, and adjacent anatomy,
design and reverse engineering software for further segmen- which can greatly improve the accuracy of the procedure and
tation and manipulation to design a virtual surgery. Through effectively avoid any complications (Fig. 1.4) [9].
the extraction of CT data, different digital operation models Surgical simulation training can improve traditional
can be created, which in turn can provide a surgical model teaching methods beyond the operating table. This type of
1 The Establishment of Digital Orthopedics 7
1.3 MRI rectly work on the human eyes and produce visual percep-
tion. Visual perception is the main approach for human being
The principle of formation of MRI is to put human body in the getting information from nature. According to statistics,
magnetic field, then obtain electromagnetic signal by using the among the information people obtained, visual information
magnetic resonance of human tissue, and combine cross-sec- accounts for 60%, auditory information about 20%, and the
tional images through a computer. Every sampling point of rest information only about 20%. This shows that visual
MRI has three sample values, representing the proton density information is very crucial for human beings. At the same
of tissue, T1 relaxation time, and T2 relaxation time, respec- time, image is the main way for people to get visual informa-
tively, so it is a vector volume data. The imaging of MRI is tion and the most important, richest, and largest amount of
based on the amount of protons of different tissues, the differ- information source that people can experience. Generally
ence between T1 relaxation time and T2 relaxation time and speaking, the object is three dimensional in space, but the
other numerous parameters. Adopting different scanning image captured from the object is two dimensional. And the
sequences can better observe the lesion of bone, muscle, liga- tomography cross-sectional images also can be considered
ment, cartilage, and nervous tissues. MRI has a higher resolu- as two dimensional in space.
tion than CT in imaging soft tissue. However, nowadays MR
imaging still cannot be isotropic. The thickness of the section Simulation Image
is generally 7 mm, so in the application of digital orthopedics, It includes optical image, camera image and television
MRI is usually applied in combination with CT. image, etc. For instance, the image people saw under a
MR’s spatial resolution of sections is 1 mm, while the microscope is a simulation of optical image. The processing
thickness of the section is 7 mm [1]. rate of simulation image is quick, but it has shortage in accu-
racy and flexibility, and it is hard to be found and judged.
is outputted. Frequently used image processing methods are The image processing tools of MATLAB support many
image enhancement, image restoration, image coding, image kinds of image data format, including DICOM format of
compression, and so on. There are a lot algorithms and soft- medical image. At the same time, it offers a lot of functions
ware of image processing that can be easily understood and to apply to image processing. By using these functions, peo-
grasped by medical professionals. Here, two kinds of com- ple are able to analyze image data, obtain detailed informa-
mercial software will be introduced: Adobe Photoshop and tion of image, design corresponding filtering algorithm, and
MATLAB. For medical researchers, these two have their filter the noise contained in the image data. Many functions
own advantages and disadvantages that can compensate each of mathematical morphology are used to process grayscale
other, which is a relatively better choice for medical image image or binary image and realize quick edge detection,
processing [3]. image denoising, skeleton extraction, watershed segmenta-
tion, and so on. And all of them can process multi-dimen-
2.2.1 Adobe Photoshop sional image data.
Photoshop is the world famous image processing and design Image processing tools provide many high-level image
software. Its functions can be divided into four parts: image processing functions, including operations like permutation,
editing, image composition, color correction and modula- transformation, sharpening, etc. Also, image cutting and
tion, and special effects. Image editing can do amplifying, size changing can be achieved by employing these
contracting, rotating, tilting, mirroring, changing perspec- functions.
tive, and so on. It also can duplicate, remove speckles, repair MATLAB itself is a powerful data visualization tool,
and modify damaged image, etc. Image composition is to which can present analyzed data through various forms, for
compound several images into an intact image conveyed a example, gray level histogram, contour line, montage mix-
clear and definite meaning by applying layer operational ture, pixel analysis, layer transformation and texture map-
tools. Color correction and modulation is to adjust the bright- ping, etc. Using visualized image not only can assess the
ness and to correct color bias. They can also switch between properties of graphic images but also can analyze the color
different colors in order to meet the need of application of distribution and other conditions of image.
image in different fields. Special effects are mainly done
through a synthetic application of filter, channel, layer, and
other tools of Photoshop. Creative idea and words with spe- 3 Image Standard
cial effect and some traditional painting techniques like oil
painting, embossment, plaster oil painting, and sketch all can 3.1 Pixel, Resolution, and Pixel Size
be made by it.
Photoshop has batch processing function and uses vari- Pixel is the most basic unit forming an image. It can be
ables and event-based scripting, which makes it able to pro- seen as an extremely tiny square color blocks. An image is
cess repetitive tasks. It also integrates many mature algorithm usually made up by many pixels, which are arranged in
of image processing, so it has advantages like great interac- rows and columns. When the user uses the scale tool to
tivity, fast operational speed, and easy to grasp. Now zoom in to a certain degree, he/she would see the result
Photoshop is widely used in many aspects of scientific and resembling mosaic and each block is a pixel. Pixel is an
medical image processing. However, due to its lack of accu- individual element of digital image matrix. Its position
rate and flexible scientific computing ability, it is difficult to decides the objects’ position in the image, and its size is
conduct image registration and segmentation. the grayscale value of grayscale image and RGB value of
color image.
2.2.2 MATLAB Image Processing The more pixels in a unit area will result in the higher
MATLAB is the abbreviation of Matrix Laboratory. It is a resolution, so the effect of the image is better. The unit of
scientific computing commercial software using matrix form image resolution is pixel/inch, namely, the amount of pixel
to process data and now is widely applied to analysis, simu- in an inch. If the image’s resolution is 72 pixel/in., it means
lation and design of scientific computation, control system, each inch includes 72 pixels. The higher image resolution
information processing, and other fields. With a specialized means there are more pixels in each inch, therefore more
tool developed by the experts of each domain, our work details and smoother color transition of image.
would have a higher starting point. And MATLAB command What is needed to be noted is that usually image resolution
is very close to the symbol and formula in mathematics, refers to the amount of pixels in a unit area of photographic
readable and easy to grasp. paper after printing the image. And also the pixel size of the
MATLAB provides an intuitive and reliable integrated medical cross-sectional images is the proportion of raw dis-
developmental tool for image processing engineers, scien- tance between two dots on the image (whose unit is the num-
tists, and researchers. These tools are widely used in aero- ber of pixel) and actual physical distance of an object (whose
space, remote sensing and measuring, biotechnology, medical unit is cm). The references of these two distances are different:
imaging, scientific image processing, and other fields. the previous one is printed photographic paper, and the latter
12 G. Pei and S. Xiuyun
is the actual object. Their applications are also different: raw 3.4 Image Format
distance is applied to photograph and photo printing industry,
while the physical distance is used in medical imaging [4]. Bitmap, also called raster graphics, is the image displayed by
pixel matrix. The color information of each pixel is presented
by RGB combination or grayscale value.
3.2 Color Depth BMP file is a method of data exchange and storage devel-
oped by Microsoft Company. Every version of Windows all
Color depth indicates the number of color in an image, namely, supports files in BMP format. Windows provides a fast and
the binary digits storing matrix elements of digital image. convenient method to store and compress BMP file. The
One binary number can only save 0 or 1 two numerical shortage of BMP format is that it would take up a relatively
values, so it is also called binary image or logical image. large storage space and the file size is too big.
Usually image segmentation is resulted in a single bit image. TTFF (Tagged Image File Format) is a commonly used
Meanwhile, the morphological operation of image is based image format for desktop publishing system developed by
on a single bit image. Sixteen bit is 2^16 = 65,536, which is Aldus and Microsoft Company. TIFF is a non-distorted com-
able to save 65,536 grayscale value. It is the bit of storage of pressed format (2–3 times compression ratio top). This com-
image of DICOM format. If these grayscales are all dis- pression is the compression of the file itself, namely,
played in an image, then the human eyes cannot distinguish recording some repeated information in the file through a
it. Thus, the DICOM format image displaying 16 bit can only special method. The file can be completely restored main-
choose a small scope (window width) of a certain position taining the color and level of the original image. The strength
(window level) to display the grayscale. All the pixels lower of TIFF is the high quality of image, but it would occupy a
than window width is shown as black, and the pixels higher larger space.
than window width is shown as white. JPEG is the abbreviation of Joint Photographic Experts
Group. It is the most common used image file format devel-
oped by a Federation of software development organiza-
3.3 Color Models of Image File tions. It is a lossy compression format that compresses the
image into a very small storage space, so the repeated and
The frequently used color models are RGB, CMYK, and Lab. unimportant materials will be lost. Thus it is easy to cause
RGB color model is also named RGB color space. It rep- damage of image data. However, JPEG compression tech-
resents three colors: R for red, G for green, and B for blue. nique is advanced. It uses lossy compression to remove the
Red, green, and blue are known as three primary colors of redundant image data. As a result, it achieves an extremely
light. Most visible spectrum in nature can be shown by mix- high compression ratio and shows a vivid image at the same
ing light of these three kinds of color with different propor- time. In another word, it can receive a relatively high image
tions and intension. quality by using the least disk space. Besides, JPEG is a very
CMYK color model is a subtractive color model, which is flexible format which has the function to adjust the image
the fundamental difference between CMYK and RGB. We quality, allowing compressing the file with different com-
use this subtractive color model not only when we observe pression ratio and supporting multiple compression levels.
objects but also when it is applied to printing on the paper. The compression ratio is usually ranged from 10:1 to 40:1.
CMYK refers to the four colors: C for cyan, M for magenta, The bigger the compression ratio is, the lower quality will
Y for yellow, and K for key (black). be. Inversely, the lower the compression ratio means higher
Lab color model is the color space of human vision. It quality. JEPG mainly compresses information of high fre-
based on the only principle of vision—the uniform color quency which can better reserve the color information. It is
space that is the same amount of movement causing same suitable for application in the internet with advantages like
sense of color changes to the eyes. Lab color space is device saving transmission time and supporting 24-bit true colors. It
independent, producing colors matching with various devices is also commonly applied to continuous tone image.
like colors of display and printer. Lab color model makes up
for the shortcomings of RGB and CMYK. Moreover, it can
realize color change of various devices as intermediate color. 3.5 DICOM Format
In lab color model, L refers to lightness, a refers to the color
changing from green to red, and b for the color changing from As of the early 1990s, with the rapid development of com-
yellow to blue. L is definitely a positive value. When a or b is puter technology, communications technology, and network
positive, it appears as red or yellow, respectively. When a or technology, image analysis, image processing, and PACS
b is negative, it appears as green or blue, respectively. (Picture Archiving and Communication Systems) have been
2 Basis of Digital Medical Imaging 13
playing an increasingly important role in clinical diagnos- which nowadays has reached more than 10 or even 20 mega
tics, telemedicine, and medical education. One of the techni- pixels. Building a fine-pitch data layers are the key technolo-
cal problems to be solved in PACS is the standardization of gies and the resolution matches. The key technology to build
image data formats of various digital imaging devices and datasets thin distance lies in the matching of layer distance
data transmission. DICOM 3.0 (Digital Imaging and with resolution.
Communications in Medicine 3.0), an important network CDH-M1 and CDH-F1, the two datasets of Chinese digi-
standard and communication protocol for PACS to become tal human have a cutting layer distance of 0.2 mm. And the
an open system, is a new standard of digital imaging and resolution is 3024 × 2016, the cross-sectional area of the
communications, in accordance with which the production tomogram to be collected is 600 mm × 400 mm, and mini-
of different types of digital imaging devices can be realized mum side length of distinguishable pictures is 0.2 (mm).
as one communicates with different manufacturers through CDH C-F1 dataset has a cutting layer distance of 0.1 mm, the
PACS. resolution of 4256 × 2848, the cross-sectional area of tomo-
DICOM format image files are files stored in accordance gram to be collected of 280 mm × 280 mm, and side length
with DICOM standard. A DICOM file generally consists of for the minimum unit of distinguishable pictures of 0.1
DICOM File Meta Information and DICOM data set. (mm).
DICOM File Meta Information contains information
about the data set. It can be understood as recording all the
useful information of a certain DICOM format image. For 4.2 Color Management
example, in a DICOM format CT image, the DICOM File
Meta Information records the patient’s name, image size, Color, the inherent properties of objects, is also susceptible
layer thickness, layer distance, pixel resolution, as well as to the environment. The image we get through digital camera
other clinical and image-related information. or scanner should be revealed on the display and transmitted
As for the description of the image, DICOM takes advan- to the printer to be printed out. Different color modes lead to
tage of bitmaps, in which, as elaborated above, grayscale different color ranges; and different devices with the same
images are usually stored in 16-bit. DICOM allows three color mode can bring out different color ranges. The way to
matrices to represent the three components, respectively, or ensure the faithfulness to the original color during shooting
one matrix for the representation of the entire image. The and transmission process and achieve accurate reproduction
former can be used to store color images, such as the recon- is called color management.
structed three-dimensional color image we see in the PACS Color management, simply put, is the use of computer
system and the latter to store 16-bit grayscale images. technology for orderly management of color system, namely,
the color must be consistent through all the stages, from ini-
tially the image signal importing (such as shooting, scan-
4 uality Control of Digital Human
Q ning, etc.) to the middle stages (image processing with image
Images in China editing software), and finally to the output (printing). The
core of color management is ICC profile. ICC profile is a
4.1 elationship Between Cutting Layer
R color description file of a digital device, which signifies the
Space and Pixel correspondence between the expression of this particular
color device and standard CIE Lab color space. ICC profiles
Thanks to the advancement of layer cutting machine and provide description files mainly for devices concerning the
digital camera resolution, the layer distance of equally following three aspects: input (scanners, digital cameras),
spaced digital human datasets has been substantially display (various displays), and output (printers or a variety
decreased. In America, the earlier male datasets of VHP was of color output devices) and require a scientific and reason-
l mm, while the later female datasets distance reached able match among them to reach a correct color reproduction
0.33 mm; the Korean datasets VKH and Chinese datasets of the image.
CDH (of both gender, CDH-M1, CDH-F1) have an modified Standard color card is the card printed with a number of
layer distance of 0.2 mm; Chinese Digital Human Famel- standard colors. On the card are various gradient colors that
Child No. 1 (CDH C-F1) has an even smaller distance of all follow strict standards and have the color descriptions in
0.1 mm. But the key technology of achieving slice datasets the color card Text Description File (TDF). The more color
with smaller spaces is not about layer distance, because with on the card, the better the correction. The commonly used
the help of CDH-VCH-FA00 gantry milling machine spe- card is IT8 color cards. By synchronously shooting IT8 color
cially developed for DCH, the experimental layer distance or special CDH color on the tomogram and then testing
can arrive at 0.02 mm, nor is that about camera resolution color-coded images with colorimeter and spectrum analyzer,
14 G. Pei and S. Xiuyun
the color management profile of tomogram data for Chinese 5 Image Registration
digital human datasets can be established for the completed
property profile, thereby ensuring the tomographic image of 5.1 asic Theory about Medical Image
B
Chinese digital human to be consistent in color throughout Registration
the subsequent processing.
In the acquisition, transfer, and record keeping process of the
image, due to various factors, such as the effects of atmo-
4.3 Data Storage and Backup spheric turbulence, diffraction of the optical system in the
equipment, the nonlinearity of the sensor, aberration of the
Data backup is to store data in a suitable medium through a optical system, the relative motion between the imaging
particular way for the purpose of ensuring that data will not device and the object, the nonlinearity of photosensitive film,
be missing under any circumstances and will always be grain noise of the film, the nonlinearity of television camera
available. The core of data backup is to restore the data—a scanning, and others, the image will inevitably result in dis-
backup that cannot be restored makes no sense. Data backup tortion. Typically, the quality declining caused by these fac-
of CDH is divided into two categories, namely, real-time tors is called image degradation.
backup while data collection and copy backup while data The typical examples of image degradation are blur, dis-
using. That involves data backup media, data backup hard- tortion and additional noise. Because of the degradation of
ware and data backup software. the image, the image displayed in the receiving end is no
Real-time backup data is also known as data synchroniza- longer the original image under transmission; thus the qual-
tion backup. It is a means to prevent data loss during the ity of the image drops dramatically. Therefore, to show bet-
collection process which might be caused by hardware fail- ter the image, it is a must to work on the degraded image and
ure or human errors. The real-time data backup of CDH, restore the original one. And such process is referred to as
through dedicated real-time synchronization backup soft- image restoration.
ware, provides the backup machine with a synchronization Image restoration technique is a very important technol-
tool with folder directories specified by the gathering ogy in image processing. Similar to other basic image pro-
machine of ready access. Thus the directory on the two com- cessing technology such as image enhancement, it also aims
puters can be synchronized and mirrored through a small at the improvement of visual quality. But the difference is
local area network to achieve data backup purposes. that image restoration process is an estimation process dur-
To prevent the dataset from falling short in the actual use ing which degraded images are restored according to the
and protect the established data set from failing which might specified degradation model to get the original picture that
be caused by system crashes, hardware failures, virus attacks, has not been degraded. In short, image restoration is to
misuse, and natural damage, we need a backup copy of the improve the degraded image and to achieve amelioration in
data set, that is, the hardware backup. The backup CDH orig- the visual image.
inal data set currently uses hard disk as a storage medium for Since the image degradation can be caused by various
backup copy and conducts full backup regularly to achieve factors, and the properties of each are not identical, there is
backup copies of the original data set. For safety reasons, currently no uniform method for restoration. Many research-
promptly after a successful construction of CDH original ers have worked out different restoration methods by study-
data set, we produce two copies of the original data set and ing the physical environment of application and thus employ
keep the two original sets and two copied ones in four differ- different degradation models, processing techniques, and
ent sites. estimation criteria accordingly. Degradation and restoration
In addition, by depositing timely the CDH original data are actually a pair of mutually positive and negative issues.
set on websites of partner organizations, such as the Institute Degradation is a well-posed problem, and restoration is an
of Computing Sciences of CSA (Chinese Academy of ill-posed problem. The estimation of degradation factors is
Sciences), the Shanghai Center for Biotechnology the first step of image restoration.
Information (CAS grid) Institute, Huazhong University of Image distortion and registration is a special kind of deg-
Science and Technology (Ministry of Education Grid), etc., radation and restoration. In doing medical image analysis,
whose network could backup the dataset, the function of several images of one patient are often analyzed together for
automatic unattended backup, cross-platform data manage- comprehensive information about the patient, which contrib-
ment, real-time subset backup, database management, disas- utes to the improvement of medical diagnosis and treatment.
ter recovery, error alarm, and others can be realized. It is also While doing quantitative analysis of several different images,
an effective way to ensure the security, reliability, and effi- we must firstly align the images strictly, and this is what we
ciency of the entire data backup. call the image registration.
2 Basis of Digital Medical Imaging 15
that the tomogram is placed in a three-dimensional coordi- in three-dimensional space (rotation of axis X, Y, Z and plane
nate system for the observation of the three-dimensional translation of X, Y, Z), causing tomographic translation,
coordinates of P, a point on the contour of an unknown rotation, and scaling (square corresponds to square), shear
object. The coordinate values of P are X-axis and Y-axis (x, (square corresponds to parallelogram), and tilt (square cor-
y), which can be determined by its location in the horizontal responds to arbitrary quadrilateral; straight line corresponds
tomogram; P in the Z-axis has a coordinate value (z) that can to straight line), which are collectively referred to as projec-
be determined by the product of a cut distance (d) and tomo- tive distortion. Thus registration of continuous tomographic
gram order (n) (z = d * n). Therefore, the continuous tomo- images can be reduced to projective transformation based on
gram point P contains accurate three-dimensional coordinate coordinates of positioning rods.
information that extends from the point to the surface and Firstly, process the original tomographic images to get the
finally to the body. Thus it can be easily concluded: the three- images of the four positioning rods on every layer, calculate
dimensional structural information of an unknown object in MATLAB the centroid coordinate values of the four posi-
can be observed in a continuous tomogram. tioning rods which are used as the coordinate values of the
It should be noted that the continuous tomographic image positioning rods, and take the average of coordinate values
registration is a registration done in a special case where of all the tomograms as the reference coordinates. Secondly,
each of the tomographic image comes theoretically from dif- determine the dimensional projective transformation param-
ferent objects. If taking the conventional medical registration eters based on the coordinate values of positioning rods on
method, we will commit a serious mistake due to different each layer, carry out projective transformations of tomo-
premises. For example, the continuous tomographic images graphic images, and eliminate projective distortions. Then
of a tilt cylinder prove to be an oval. If reconstructed three- get images of the first positioning rod from corrected images
dimensionally by the conventional registration method, the reprocessed in Photoshop, calculate their coordinate values
images will form an elliptical cylinder. in MATLAB, and, based on the coordinate values of the
Accordingly, during the shooting and storage of tomo- positioning rod, cut the tomographic images into ones of the
graphic images, because of different parameters of image same size.
forming devices, as well as varied relative positions and
angles between the image forming device and tomogram,
there is often a positional deviation between the image lay- 6 Image Segmentation
ers. So while embedding and shooting specimens, we need to
add location marks to facilitate the registration of the origi- 6.1 he Overview of Medical Image
T
nal image, in order to ensure that each piece retains exactly Segmentation
the two-dimensional information of the original tomogram.
Of course, the advanced quick-scanning image acquisition Victor Spitzer, the pioneer of the research of Digital Virtual
equipment of CT and MRI has made it unnecessary to con- Human, once said that there was just one challenge in the
sider the alignment between tomograms. research, which is segmentation. Image segmentation is the
foundation and the bottleneck of the succeeding image pro-
cessing including 3D reconstruction. On the one hand, accurate
5.4 egistration of Continuous
R image segmentation takes a lot of time and effort; on the other
Tomographic Image of Chinese Digital hand, the precision level of image segmentation will affect the
Human results of the succeeding medical basic research and clinical
application. Thus it is of particular significance to improve the
During the image acquisition process of Chinese Digital efficiency and precision level of image segmentation.
Human (CDH), the parameter setting of digital cameras Image segmentation is to differentiate areas of special
remains unchanged. Since the shape of human tomograms implications from each other. Both the data sets of Chinese
change at different layers, factors of image degradation cannot Digital Visual Human and those of clinical serial sections are
be accurately estimated by the organ’s changing shape pre- volume data sets containing the 3D information of structure.
sented in the tomograms. To facilitate the locational registra- The segmentation of volume data sets is similar to that of 2D
tion of photos collected, we have added four positioning rods images; namely, it’s to segment voxels of special implica-
while doing embedment. The positioning rods preset in the tions from volume data sets, which is like the process of
cutting specimens remain at the same spatial position on each carving a statue out of a huge stone. There are two ways to
actual layer, so changes of positioning rods in the continuous segment voxel data; one is to separately segment every 2D
tomographic image may represent distortion of each image. slice and the other is to directly segment 3D volume data
With respect to human tomogram plane, a digital camera sets. Thus, image segmentation can also be simply under-
can realize small pan and rotation in six degrees of freedom stood as a process of choice. For every pixel constituting 2D
2 Basis of Digital Medical Imaging 17
images and every voxel volume data sets, the result of choice Thus, at the moment any single kind of section image
is either removed or preserved. Thus the results of image cannot meet the precision demands of all the medical
segmentation can be preserved with binary image. researches and clinical applications and any single kind of
computer image segmentation algorithm is difficult to satis-
6.1.1 he Features of Medical Image
T factorily segment medical images. Although medical image
Segmentation segmentation is a hot research topic in the majority of com-
One of the reasons why medical image segmentation is very puter graphics and it’s continuously making progress, at the
challenging is that whatever imaging mode is adopted, the moment medical researchers cannot be replaced due to the
real information will to some degree be lost or distorted in reasons mentioned above. The purpose of medical image
the process of image acquisition, so that the results of sec- segmentation is to meet the precision demands to the maxi-
tion-image segmentation can only approximately but cannot mum of the medical researches and clinical applications.
completely reflect the real anatomical boundaries. Thus the author would like to give some suggestions to the
Medical researchers and surgeons hope that normal or process of segmenting the medical image data.
pathological anatomical structures reconstructed by com- Firstly, raw data of high quality is to be prepared. As a
puter can truly represent the layer of structure shown in saying goes, if you have no hand, you cannot make a fist.
human anatomy and surgeries and even reproduce fine struc- Therefore, to start, we should fully understand the research
tures seen under a microscope. However, all imaging devices programs, as different research objectives require different
produce some noises, and each kind of imaging mode has precision degrees of segmentation. For example, anatomic
some limitations in reflecting the real anatomical structures. teaching asks for segmenting adjacent tissues as much as
For example, the optical photograph of the anatomical speci- possible, while the reconstruction precision degree of the
men of frozen section shows the reflectivity to visible light brain visual operation project is higher than that of the ortho-
of different tissues. But it cannot clearly distinguish adipose pedics visual operation project. Then, we should fully under-
tissue and nervous tissue that are similar in color as well as stand the imaging features of all kinds of imaging devices,
connective tissues like adjacent muscle compartments, peri- parameter conditions that affect imaging precision and how
osteum, muscle tendons, and joint capsules. CT tomography to choose suitable imaging mode according to the research
reflects the attenuation rates to X-rays of different tissues, objective, in order to go ahead with the succeeding image
but it cannot clearly distinguish soft tissues that are similar in segmentation. For example, to perfuse the vessels of human
density. MRI tomography reflects the density of hydrogen body specimen in advance can make easy blood vessel seg-
protons contained in tissues and the time for relaxation of T1 mentation; regular dyeing of tissue pathological slices or the
and T2, but more often than not, the reflected pathological immunohistochemical staining makes easy image segmenta-
boundaries of the tissues are amplified compared with the tion of calibration structure; high scanning voltage of CT can
real ranges. improve the accuracy of the skeleton segmentation. Finally,
The second reason is that whatever computer segmenta- we should frequently communicate with the section offices
tion algorithm is resorted to, the accuracy of the computer and researchers concerned in order to save resources and
automatic image segmentation can hardly be up to the ana- obtain the raw data of high quality.
tomical level or the level of medical imaging experts’ reading Secondly, we should be familiar with the structure to be
photographs when it comes to specific normal or pathological segmented. Reading slices based on the images require us
anatomical structures. In other words, the computerized algo- to bear them in mind in advance. To start, before we seg-
rithm cannot reach the level of visual thinking. For example, ment anatomical structures, we need to review all the relat-
we see the changing clouds in the sky and imagine them ing anatomical knowledge and literature, to learn their
sometimes as a flock of sheep and sometimes as running anatomical features and variations, to carefully observe
horses. The reason why out of the clouds we can see the anatomical specimen, and to conduct entity anatomy if
images of sheep or horses is that the images are already in our possible. Then, by every means, we need to carefully
mind. And this is the process of visual thinking. On the con- observe the volume database to be segmented and to track
trary, what the computer can recognize are just clouds. In the changes of anatomical structures layer by layer, in
medical image segmentation, there are two things needed to order to briefly segment the volume database in our mind
be done by the experts. One is to choose the computer auto- before we start the work in practice. We should notice that
matic segmentation algorithm, and the other is to revise the the geometrical shapes of the anatomical structures are dif-
results of the computer automatic segmentation. Thus, the ferent, so that the fracture surfaces easily to be observed
process of medical image segmentation must involve the are different. For example, the cord-like cruciate ligaments
efforts of the medical experts, and the accuracy degree of the of the knee joint are easier to be observed and segmented
segmentation results is closely related to the experience of the on the sagittal section than on the coronal plane and the
operators, as the results it cannot be repetitive. cross section.
Another random document with
no related content on Scribd:
reduced outer end of the arm, and a linch-pin is driven through the
arm beyond it.
An improvement on this kind of axle is when the collar at the upper
end or shoulder is made solid by welding, and a screw nut with a
linch-pin through it is substituted for the collar and linch-pin. These
nuts are commonly made six-sided, with a mortise or slot for the
linch-pin through each side, in order to afford greater facility for
adjustment. In all other particulars this axle is the same as the last,
except that it is occasionally case-hardened to prevent wear and
friction.
In travelling, these axles require to be fresh greased every two or
three days, and the trouble thus caused is very considerable,
besides the risk of omission, in which case the axle is likely to be
entirely spoiled.
The commonest kind of oil axle is called the “mail,” because the
peculiar mode of fastening was first used in the mail coaches. The
arm is not conical, but cylindrical, in the improved kind. At the
shoulder of this axle a solid disc collar is welded on for the box to
work against. Behind this shoulder collar revolves a circular flange-
plate of wrought iron, pierced with three holes corresponding with
holes in the wheel from front to back, through which long screw-bolts
are driven, and their nuts screwed sufficiently tight against the
circular flange-plate to allow easy motion. The wheel, when in
motion, thus works round the shoulder collar, while the flange-plate
secures it against coming off. This is not neat or accurate, but it is
simple and secure, and no nut or linch-pin is required to the axle in
front, while the front of the nave can be entirely covered in. When
screwed up for work, a washer of thick leather is placed between the
shoulder collar and the box, and another between the shoulder collar
and the circular disc, which extends over the whole surface of the
back of the nave. The box of this axle is of cast iron. The front is
closed with a plate of metal, between which and the end of the axle-
arm a space is left of about 1 inch as a reservoir for oil, which is
poured in through a tube passing through the nave of the wheel and
closed by a screw pin. At the back of the box there is a circular
reservoir for oil, ¾ inch in depth and ½ inch wide. When the wheel is
in motion the revolving of the box keeps the lubricating material in
circulation between the two reservoirs; any portion getting below the
arm at the shoulder gradually works its way out and is wasted. The
oil in the back reservoir does not waste by leakage so rapidly as that
in the front; but when the leather washer becomes saturated with
water the oil is liable, by reason of its lightness, to float on the water
in or about the washer, and thus get wasted.
This axle requires frequent examination when very much in use;
but as it is neat in appearance, and under ordinary circumstances
tolerably safe in working, and is not very expensive, it is much used.
Both axle-box and axle-arm are case-hardened.
The other kind of axle used by carriage-builders is that known as
“Collinge’s Patent.” The original intention of the inventor was to make
it a cylindrical arm, with the box running round it against a coned
shoulder, and secured by a coned nut in front; but, as it was found in
practice that a leather washer was necessary at the shoulder to
prevent jarring, this part of the plan was abandoned.
The commonest form of this axle now in use consists of a
cylindrical arm with a broad shoulder collar. The box is of cast iron,
and the back of it is similar to that of the mail axle before described.
The front of it has a rebate cut in the box to receive a small conical
collar and the screw of an oil cap. The arm of the axle is turned down
in the lathe to two-thirds of the total thickness from the point where
the rebate of the box begins. A flat side is filed on this reduced
portion, and along it is made to slide a small collar of gun metal, with
a conical face in the interior to fit against the coned interior of the
rebate in the box. Against this collar, technically called the “collet,” a
nut of gun metal is screwed, and against that again a second nut of
smaller size, with a reversed thread, is tightly fixed. These two nuts,
thus screwed in different directions, become as firm as though they
were part of the axle itself, and no action of the wheel can loosen
them, because the collet, which does not turn, removes all friction
from them. But, as a further security, the end of the axle-arm projects
beyond the farthest nut, and is drilled to receive a spring linch-pin.
Over all a hollow cap of gun metal is screwed into the end of the box.
This contains a supply of oil for lubricating purposes.
When the wheel is in motion the oil is pumped upwards from the
cap and passes along the arm to the back reservoir, constantly
revolving round the cap with the wheel. If the cap be too full of oil—
that is, if the summit of the column of oil in the cap be at a horizontal
level above the leakage point at the shoulder—it will pump away
rapidly, and be wasted till it comes to the level of the leak, where it
will be economically used. It is essential to the perfection of an oil
action that the oil should not be permanently above the level of the
leak, but that small portions should be continually washing up into
that position by the action of the wheel in turning.
In order to insure their greater durability and freedom from friction
these axles and their boxes are always case-hardened, i.e. their
rubbing surfaces are converted into steel to a trifling depth by the
process of cementation with animal charcoal for about two hours,
when they are plunged into water. The boxes are ground on to the
arms with oil and emery, either end being applied alternately, until a
true fit between the two is accomplished.
The mode in which oil acts as a lessener of friction is by its being
composed of an infinite number of movable globules, over which the
fixed surfaces of the arm and box roll without causing that friction
and wearing away which would be the result of the two iron surfaces
worked together without any lubricant. This saving in the wear and
tear of the axle-arm is accomplished by the destruction of the oil.
From this we deduce that the greater the mass of oil or grease used
the longer will the axle run, and in order to facilitate this as much as
possible there should be so much space left between the bearing
surfaces of the arm and the box as will allow of a film of oil to be
between them.
A highly polished surface is desirable in an axle and box, as the
bearing is more perfect and true. A rough surface is a surface of
sharp angles, which will pierce through the oil and cause friction by
contact.
To guard against the axle running dry, the arm is reduced in
thickness at the centre for about an inch to allow a lodgment for the
oil, and in the process of working this constitutes a circular pump,
which draws up the oil from the front cap and distributes it over the
area of the arms. But this, of course, will soon run dry, so that the
best remedy to prevent the oil being exhausted and the sticking of
the axle-arm in the box is careful attention.
A danger arising from careless fitting is the introduction of grit into
the box. This grit is composed of small grains of silex, which is very
much harder than iron or steel; the consequence is that it cuts and
scores the bearing surfaces in all directions, and keys them firmly
together, so that it is sometimes necessary to break the box to
pieces in order to get it off the arm.
A patent was taken out to remedy these defects by casting three
longitudinal triangular grooves in each box. The advantages gained
by this are, that if grit gets in it finds its way to the bottom of the
grooves and does not interfere with the action of the wheel, and,
moreover, the grooves keep up a constant surface of oil in contact
with the arm, instead of trusting to the mere capillary attraction. This
does not interfere with the bearing surface in any marked degree.
In order that the axle shall be perfect the following considerations
are necessary:—
That there be sufficient bearing surface for the arm to rest on.
That the box be of a convenient shape for insertion in the wheel.
That as large a body of oil as possible be kept in actual contact
with the arm by washing up as the wheel revolves.
That the column of oil may be in no case above the horizontal
level of the leakage point while the wheel is at rest.
Setting Axles.
Setting axles is giving them the bend and slope required, in order
to fall in with the principles of the dished wheel. It is chiefly applied to
the axle-arm, and this is the most important part, setting the beds
being mere caprice.
The great object to be obtained is, to give the arm the right pitch
every way, to make the carriage run easy and as light as possible,
even in the absence of a plumb spoke. All carriages do not look best,
when running, with the bottom spoke plumb or vertical. In some of
the heavier coaches or carriages more slope or “pitch” has to be
given to the arm to carry the wheel away from the body, so as to
bring them to some specified track, in order to suit some particular
customer, so that we must be governed by circumstances.
There is a patent “axle-set,” but it is not of much assistance, for
half the smiths know nothing about it, and if they did it would not be
generally used, as the advantages derived from its use are not equal
to the trouble of using it. Besides, the wheels are not always dished
exactly alike, and it would require adjusting to each variety of wheel;
and again, the wheels are not always (though they ought to be)
ready; and when the smith knows the sort of vehicle he is working
upon he can give his axles the required pitch, within half a degree or
so, and the patent axle-set is, unfortunately, not capable of being
adjusted to an idea.
Fig. 21.
Fig. 21 shows a contrivance for setting the axles when cold, and
consists of an iron bar a, 2 feet 1 inch long, and about 2 inches
square at the fulcrum b. A hole is punched through the end to allow
the screw c to go through; this hole to be oval, to allow the screw to
move either way. At the end of this screw is an eye of sufficient size
to go on to the axle-arm. In setting the axle the eye is slipped on to
about the centre of the arm; the clevis, d, is placed on the bar a,
near the end; the fulcrum, b, is placed at the shoulder, either on top
or underneath, according as the axle may be required to set in or
out. When the fulcrum is laid on top, a strip of harness leather should
be placed on the axle bed, and on that, an iron e, of the shape of the
axle bed, and on the end of this the fulcrum is placed; then by
turning the screw the axle may be bent or set to any required pitch.
Fig. 22.
Fig. 23.
The figure shows the two ways of doing this, one with the bar or
lever on top and the other with the lever below.
Figs. 22 and 23 show two improved forms of axles.
Fig. 24.
Fig. 24 shows another variety of the axle-set. It consists of a bar
hooked on to the axletree in two places. The bar is fastened by the
clamp m, and fulcrum block f. The eyebolt, l, is hooked over the end
of the spindle or arm, and the adjustment of the latter is
accomplished by the screw, s, and the nuts j, k.
SPRINGS.
Springs in locomotive vehicles are the elastic substances interposed
between the wheels and the load or passengers in order to intercept
the concussion caused by running over an uneven road, or in
meeting with any slight obstacle.
A great variety of substances have been used for this purpose,
such as leather, strips of hide, catgut, hempen cord, &c.; but these
have now been totally superseded by metal springs, so that what is
technically understood by the word “spring” is a plate or plates of
tempered steel properly shaped to play in any required mode.
It is very probable that the earliest steel springs were composed of
only one plate of metal. This was very defective in its action; and
unless it was restrained somewhat in the manner of the bow by the
string, it was liable to break on being subjected to a sharp
concussion.
There is no hard and fast rule by which the spring-maker can be
guided so as to proportion the strength and elasticity of his springs to
the load they are required to bear; and even were such a rule in
existence it would be practically useless, because the qualities of
spring steel differ so much that what is known in mathematics as a
“constant” could hardly be maintained. The only guide to the maker
in this respect is observation of the working of certain springs under
given loads, such springs being made of a certain quality of steel,
and any peculiar features that appear should be carefully noted
down for future reference and application.
Springs are of two kinds, single and double; i.e. springs tapering in
one direction from end to end, and those which taper in two opposite
directions from a common centre, as in the ordinary elliptic spring.
The process of making a spring is conducted in the following
manner:—
The longest or back plate being cut to the proper length, is
hammered down slightly at the extremities, and then curled round a
mandrel the size of the suspension bolt. The side of the plate which
is to fit against the others is then hollowed out by hammering; this is
called “middling.” The next plate is then cut rather shorter than the
first; the ends are tapered down so as not to disturb the harmony of
the curve. This plate is middled on both sides. A slit is then cut at
each end about ¾ of an inch in length and ⅜ inch wide, in which a
rivet head slides to connect it with the first plate, so that in whatever
direction the force acts these two plates sustain each other. At a little
distance from this rivet a stud is formed upon the under surface by a
punch, which forces out a protuberance which slides in a slit in the
next plate. The next plate goes through precisely the same
operations, except that it is 3 or 4 inches shorter at each end, and so
on with as many plates as the spring is to consist of. The last plate,
like the first, is of course only middled on one side.
The plates of which the spring is to be composed having thus
been prepared, have next to undergo the process of “hardening” and
“tempering.” This is a very important branch of the business, and will
bear a detailed description. There is no kind of tempering which
requires so much care in manipulation as that of springs. It is
necessary that the plates be carefully forged, not over-heated, and
not hammered too cold; one is equally detrimental with the other. To
guard against a plate warping in tempering, it is requisite that both
sides of the forging shall be equally well wrought upon with the
hammer; if not, the plates will warp and twist by reason of the
compression on one side being greater than on the other.[1]
The forge should be perfectly clean, and a good clean charcoal
fire should be used. Or if coal be used it must be burned to coke in
order to get rid of the sulphur, which would destroy the “life” of the
steel. Carefully insert the steel in the fire, and slowly heat it evenly
throughout its entire length; when the colour shows a light red,
plunge it into lukewarm water—cold water chills the outer surface too
rapidly—and let it lie in the water a short time. Animal oil is better
than water; either whale or lard oil is the best, or lard can be used
with advantage. The advantage of using oil is that it does not chill the
steel so suddenly, and there is less liability to crack it. This process
is called “hardening.”
Remove the hardened spring-plate from the water or oil and
prepare to temper it. To do this make a brisk fire with plenty of live
coals; smear the hardened plate with tallow, and hold it over the
coals, but do not urge the draught of the fire with the bellows while
so doing; let the fire heat the steel very gradually and evenly. If the
plate is a long one, move it slowly over the fire so as to receive the
heat equally. In a few moments the tallow will melt, then take fire,
and blaze for some time; while the blaze continues incline the plate,
or carefully incline or elevate either extremity, so that the blaze will
circulate from end to end and completely envelop it. When the flame
has died out, smear again with tallow and blaze it off as before. If the
spring is to undergo hard work the plates may be blazed off a third
time. Then let them cool themselves off upon a corner of the forge;
though they are often cooled by immersion in water, still it is not so
safe as letting them cool by themselves.
After tempering the spring-plates are “set,” which consists in any
warps or bumps received in the foregoing processes being put
straight by blows from a hammer. Care should be taken to have the
plates slightly warm while doing this to avoid fracturing or breaking
the plates.
The plates are now filed on all parts exposed to view, i.e. the
edges and points of the middle plates, the top and edges of the back
plate, and the top and edges of the shortest plate. They are then put
together and a rivet put through the spring at the point of greatest
thickness, and this holds, with the help of the studs before
mentioned, the plates together.
It is evident from the above description of a common mode of
making springs, that the operation is not quite so perfect as it might
be. The plates, instead of being merely tapered at the ends, ought to
be done so from the rivet to the points. And another thing, it would
surely make a better job of it if the plates were to bear their whole
width one on the other; in the middled plates they only get a bearing
on the edges, and the rain and dust will inevitably work into the
hollows in the plates, and it will soon form a magazine of rust, and
we all know what an affinity exists between iron and oxygen and the
result of it; as far as carriage springs are concerned, it very soon
destroys their elasticity and renders them useless and dangerous.
To prevent oxidation some makers paint the inner faces of the
springs, and this is in a measure successful, but the play of the
spring-plates one upon the other is sure to rub off some portions of
the paint, and we are just as badly off as ever. A far better plan
would be to cleanse the surfaces by means of acid, and then tin
them all over, and this would not be very expensive, and certainly
protect the plates of the spring longer than anything else.
The spiral springs, used to give elasticity to the seats, &c., are
tempered by heating them in a close vessel with bone dust or animal
charcoal, and, when thoroughly heated, cooled in a bath of oil. They
are tempered by putting them into an iron pan with tallow or oil, and
shaking them about over a brisk fire. The tallow will soon blaze, and
keeping them on the move will cause them to heat evenly. The steel
springs for fire-arms are tempered in this way, and are literally “fried
in oil.” If a long slender spring is needed with a low temper, it can be
made by simply beating the soft forging on a smooth anvil with a
smooth-faced hammer.
In setting up old springs where they are inclined to settle, first take
the longest plate (having separated all the plates) and bring it into
shape; then heat it for about 2 feet in the centre to a cherry red, and
cool it off in cold water as quick as possible. This will give the steel
such a degree of hardness that it will be liable to break if dropped on
the floor. To draw the temper hold it over the blaze, carrying
backward and forward through the fire until it is so hot that it will
sparkle when the hammer is drawn across it, and then cool off.
Another mode is to harden the steel, as before stated, and draw
the temper with oil or tallow—tallow is the best. Take a candle, carry
the spring as before through the fire, and occasionally draw the
candle over the length hardened, until the tallow will burn off in a
blaze, and then cool. Each plate is served in the same way.
Varieties of Springs.
The names given to springs are numerous, but the simple forms
are few, the greater part of the varieties being combinations of the
simple forms.