You are on page 1of 19

Obstructive Sleep Apnea Detection Using

Machine Learning Techniques


PRESENTED BY

EDWINDEZUSA M (Reg. No. 410317104012)


RAJKUMAR R (Reg. No. 410317104029)
REVATHI K (Reg. No. 410317104032)
YUKENDIRAN N (Reg. No. 410317104040)

Under the Guidance of,


Mr. A ANANDAN M.E.,
Assistant Professor,
Department of Computer Science and Engineering,
ARM College of Engineering and Technology,
Chennai- 603 209.
ABSTRACT

 Obstructive sleep apnea (OSA) is a prevalent but severely undiagnosed sleep disorder
that impacts the natural breathing cycle during sleep with decreased breathing periods or
no airflow at all.
 New sleep apnea classification techniques are nowadays being developed by
bioengineers for most comfortable and timely detection.
 This project focuses on logistic regression and k-means Clustering algorithms to get
better accuracy.
INTRODUCTION

Obstructive sleep apnea (SOA) is the most common type of sleep apena and is
caused by complete or partial obstruction of upper airway.
Recent studies suggest that about 10% of the population are suffering from
obstructive sleep apnea.
 Most case by Old age , decreased muscle tone ,brain injury .

3
OBJECTIVE

 Over the past few years, various OSA detection methods were
developed to extract information about frequency domain, time
domain, and other nonlinear features Obstructive sleep Apnea (OSA)
is the most common type of sleep Apnea
 Apnea datasets from physionet.org were used for model train and test
datasets using machine learning algorithms.

4
MACHINE LEARNING

 Machine learning is a growing technology which enabels which computers tolearn


automatically from past data.
 Meachine learning uses various algorithm forbuilding mathematical models and
making predictions using historical data or information.
 Machine learning using a artificial intelligence that is mainly concerned with the
development of algorithm.

5
EXISTING SYSTEM

The Existing system uses same techniques, but different algorithms to predict
the Obstructive Sleep Apnea (OSA) .
The algorithms used in the existing systems are Support Vector Machine (SVM)
and KNN
DISADVANTAGES

Less number of Data set collection


Feature handled is little complex
Consuming huge time
This study introduces an efficient approach of monitoring the patient for
OSA based on EEG channel using discriminating features and an
experiment of machine-learning algorithms like logistic regression and k-
means clustering.
Logistic Regression is to find a relationship between features and
probability of particular outcome .
The k-means clustering algorithm attempts to split a given anonymous
data set (a set containing no information as to class identity) into a fixed
number (k) of clusters.

8
 Logistic regression is one of the most popular machine learning
algorithm
 which comes under the supervised learning technique
 logistic regression predicts the output of a categorical dependent
Variable technique
 it can be either yes or no,0 or 1, true or false and instead of giving
the exact value as 0 or 1

9
LINEAR REGRESSION ALGORITHM

 Linear regression is a machine lerning algorithm


based on supervised learning.
 It is performs a regression task.
 It is mostly used for finding out the relationship
between variables and forecasting.

10
K-MEANS CLUSTERING ALGORITHM

K-means clusting is an unsupervised learning algorithm that is


used to solve the clustering problems in machine learning or data
science.
In this topic, we will learn whith is k-means clustering algorithm
How the algorithm works along with the python implementation
of k-means clustering
ADVANTAGES

Easily handled feature extraction


Easy detection of the sleep Apnea disease
High accuracy
High performance
SYSTEM MODULES

 Module 3: Feature selection and extraction


 Module 1: Dataset collection
 Module 2: Pre-processing
 Module 4: Detection and prediction

13
SYSTEM ARCHITECTURE

MODELING/
TRAIN
LEARNING
DATA

FEATUR DETECTION
PRE EXTRACTION
TEST DATA AND
PROCESSING SELECTIONS PREDICTION
SOFTWARE REQUIREMENTS

 Operating System : Windows


Simulation Tool : Opencv python
Documentation : Ms-Office
16
CONCLUSION

This is manuscript has reviewed distinct set of alternate models


Machine leaning models that were proposed which can help in improving the
accuracy of the diagnostics
If a conglomerated models of multi-tiered approach for detection if designed
too can be an effective model for development

17
REFERENCES

D. S. Avalur, “Human breath detection using a microphone,” Master’s thesis,


Faculty of Mathematics and Natural Sciences, University of Groningen, Aug.
2018.
L. Jiang et al., “Automatic sleep monitoring system for home healthcare,” in
IEEE-EMBS International Conference on Biomedical and Health
Informatics, Jan. 2017.

18
Thank You ..,

19

You might also like