AU2020102907A4

AU2020102907A4 - Novel automated machine learning algorithms based system for sleep staging features analysis

Info

Publication number: AU2020102907A4
Application number: AU2020102907A
Authority: AU
Inventors: Danish Ahamad; Sultan Ahmad; Md.Mobin Akhtar; Mohammed Abdullah Alnahidh; Obaid Alotaibi; Atif Amin; Abu Sarwar Zamani
Original assignee: Individual
Current assignee: Individual
Priority date: 2020-10-21
Filing date: 2020-10-21
Publication date: 2020-12-17
Anticipated expiration: 2028-10-21

Abstract

NOVEL AUTOMATED MACHINE LEARNING ALGORITHMS BASED SYSTEM FOR SLEEP STAGING FEATURES ANALYSIS ABSTRACT With the speeding up of Social Activities, rapid changes in lifestyles, and an increase in the pressure in professional fields, People are suffering from several types of sleep related Disorders. It is witnessed that many of the sleep related disorders are the symptoms of Neurological Disorders in the latter part of Life, and which affect the quality of life in daily Activities. The Crucial Steps involved in diagnosing these disorders are Analysis and Classification of Sleep Scoring. It is very tedious task for clinicians for monitoring entire Sleep durations of the subjects and analysis in the sleep staging in traditional and manual lab environment methods. For the purpose of accurate diagnosis of different Sleep Disorder, We have considered the Novel Automated Machine Learning Algorithms Based System for Sleep Staging Features Analysis is used for automated analysis of sleep epochs, which was collected from the subjects during sleep time. The Present Invention disclosed here in is Novel Automated Machine Learning Algorithms Based System for Sleep Staging Features Analysis comprising of Human Brain (101), Acquisition (102), Channel Data Segmentation (103), Preprocessing (104), Feature Extraction (105), Automatic Feature Selection (106), Classification (107), and Sleep Stages (108). The invention disclosed here provides Automatic Sleep Staging Features Analysis by the Machine Learning Algorithms Based System for Sleep Disordered and Healthy Subjects for sample size of each subject same, was 750 epochs and the length of each epoch is the 30 seconds. 1/4 NOVEL AUTOMATED MACHINE LEARNING ALGORITHMS BASED SYSTEM FOR SLEEP STAGING FEATURES ANALYSIS DRAWINGS 101 102 03 Acquisition Channel Data Preprocessing - Segmentation Human Brain 108 10 Classification Sleep Stages DT 4 M Automatic 4 0Feature Extraction KCNN Feature Selection Random Forest Figure 1: Novel Automated Machine Learning Algorithms Based System for Sleep Staging Features Analysis. 6 Wake 201 4 -N-REMI 202 N-REM2 2 -N-REM3 C' - M 2MB -4 0 5 10 i5 20 25 30 Time(S) Figure 2(a): Sleep Stages behavior of Affected Sleep Problem Subject-15 with a 30 second epoch length.

Description

1/4

NOVEL AUTOMATED MACHINE LEARNING ALGORITHMS BASED SYSTEM FOR SLEEP STAGING FEATURES ANALYSIS DRAWINGS

101 102 03

Acquisition Channel Data Preprocessing Segmentation -

Human Brain

108 10

Classification

Sleep Stages DT 4 M Automatic 4 0Feature Extraction KCNN Feature Selection Random Forest

Figure 1: Novel Automated Machine Learning Algorithms Based System for Sleep Staging Features Analysis.

6 Wake 201 4 -N-REMI 202 N-REM2 2 -N-REM3 C' - M 2MB

-4

0 5 10 i5 20 25 30 Time(S)

Figure 2(a): Sleep Stages behavior of Affected Sleep Problem Subject-15 with a 30 second epoch length.

AUSTRALIA Patents Act 1990

COMPLETE SPECIFICATION INNOVATION PATENT NOVEL AUTOMATED MACHINE LEARNING ALGORITHMS BASED SYSTEM FOR SLEEP STAGING FEATURES ANALYSIS

The following statement is a full description of this invention, including the best method of performing it known to me:

NOVEL AUTOMATED MACHINE LEARNING ALGORITHMS BASED SYSTEM FOR SLEEP STAGING FEATURES ANALYSIS FIELD OF INVENTION

[0001] The present invention relates to the technical field of Biomedical Signal Processing of Biomedical Engineering.

[0002] Particularly, the present invention is related to Novel Automated Machine Learning Algorithms Based System for Sleep Staging Features Analysis of the broader field of Biomedical Signal Processing of Biomedical Engineering.

[0003] More particularly, the present invention is relates to Novel Automated Machine Learning Algorithms Based System for Sleep Staging Features Analysis in which the Sleep Staging Features are automatically analyzed by the Machine Learning Algorithms based System.

BACKGROUND OF INVENTION

[0004] Novel Automated Machine Learning Algorithms Based System for Sleep Staging Features Analysis disclosed here is the most useful where the Analysis and study of human brain behaviour can be very important in different health sector applications such as diagnosis for mental and Neuro-disorder diseases and abnormalities. The activities of human brain analyzed through recordings of EEG signal. The major role of EEG is to assist the physicians for accurate diagnosis of various diseases from recorded behaviour of the subjects which assist to the sleep experts for distinguishing the sleep irregularities and prediction of epileptic seizures etc.

[0005] Among that sleep-related disorders are major global challenges nowadays across global because proper sleep can decide the major balance between our physical and mental health, it directly affects the human quality of life and its consequences so many other diseases are reflected in our body. Therefore sleep study gives more insights into various health risks and issues such as memory impairment, diabetes, and cardiovascular diseases. Henceforth it is more vital for analysis possible symptoms of sleep problems, a sleep staging score is necessary for almost all conditions.

[0006] The procedure sleep scoring is identifying the irregularities that happened with the help of Polysomnography recordings from the patients during sleep hours. Generally, Polysomnography is one of the multi-parameter tests and it helps towards analysis and interpretation of multiple and simultaneous activities happened in the body during sleep through recordings of different electrophysiological signals. During treatment different physiological signals recorded from different parts of the body. Brain behaviour information collected through recordings of electroencephalogram information recorded using electroencephalogram (EEG), similarly the chin, left and right limb activities recorded using the signal from Electromyogram (EMG), the behaviour of eye movements acquired through recording of Electroocoulogram (EOG) signal, and the heart rhythm information collected using Electrocardiogram (ECG) signal, respiratory signals, airflow signals, and oxygen saturation.

[0007] The sleep experts measured all these recordings and analyses through standard sleep manuals named as R&K rules, As per R&K rules, the whole sleep is classified into three stages, awake (W), non-rapid eye movement (NREM) and rapid eye movement (REM).Further NREM sleep stages divided in to N1-N4 sub sleep stages.N1 and N2 considered as light sleep, similarly N3 and N4 as deep sleep. According to R&K rules, each epoch length is 30s data. Furthermore, another sleep standard named as AASM with minor revision to earlier R&K guidelines, According to AASM, the whole sleep was segmented into five sleep stages. The only changes observed between AASM and R&K rule are the total number of sleep stages. In AASM, the N-REM stage3 and N-REM stage4 are combined into one stage called N REM stage3; another sleep stage was the same as it in R&K rules.

[0008] From all the electrophysiological signals, only EEG signals are used mostly with subject to treatment and diagnosis of possible sleep-related diseases because it directly provides the information on brain activities during sleep. Another major advantage with EEG signals that it is characterized by individual sleep stages through different EEG waveforms such as alpha, theta, beta, delta, and gamma. The brain- behaviour recorded from the scalp by attaching the electrodes, which mainly track the changes voltage differences between the different neurons. The placement of electrode managed according to internationally 10-20 EEG electrode placement rules. The concept of 10-20 refers to the placement of the electrode are either 10% or 20% right left or front-back of the scalp. Nowadays for diagnosis of different types of sleep disorder, EEG signal is preferable in the sleep laboratory for monitoring the sleep quality in the real-time environment and sometimes it may also use as a portable manner in the home itself so that one can easily measure the sleep quality home itself.

[0009] The wake stage is considered a relaxed state where one subject can prepare himself or herself for going to sleep. After sometimes the body going to enter into N REM stage and from here actually starts a sleep cycle, it is a transition phase between wake and sleep. The average duration of this state 5 to 15 minutes in general and the brain normally produces theta waves .In N-REM stage2 is some deeper sleep incomparable to stagel, where the heart rhythm and body temperature reduces .sleep spindles and k-complexes have occurred during this stage of sleep.

[0009a] In N-REM stage3, actual deep sleep starts, and the brain behaves very slow and the same conditions are continuing for N-REM stage4, delta waves are produced from the brain during this stage of sleep. Finally, in the REM stage, small brain waves are seen and blood pressure may suddenly increase, breathing also irregular, and rapid eye movements have occurred in this phase of sleep. This process continues cyclically throughout the whole night from NREM stages to the REM stage and each cycle duration approximately 90 min. One healthy sleep covers normally 3-5 sleep cycles for a subject.

[0010] In general the recordings of EEG signals are highly in nature of complicated forms and the characteristics of these signals continuously fluctuate with regards to amplitude, phase, and frequency. To extract meaningful data from raw EEG signals, we need to analyse the extended periods of the signal. Another aspect of difficulty also observed in terms of EEG recording, generally recorded hours of sleep data is quite difficult to analyze with 5-10s time windows.

[0011] Sometimes it has observed from different sleep studies that the multi-channel

EEG approach raised some limitations towards the evaluation of irregularities due to disturbances found on the subject's health. Therefore most of the sleep researchers used single-channel of EEG for classifying the sleep stages.

[0012] Traditionally, the technicians have analysed and interpreted the sleep stages manually, alternatively, it raised set of limitations with related to manage the huge amount of sleep EEG records and it is also a time-consuming process, highly expensive process and finally, this manual interpretation of sleep records human dependent process. As an outcome, it has required to develop an automated sleep stage classification system to get better classification accuracy.

[0013] Present invention disclosed here uses Sleep Recordings and it was obtained from the ISRUC-Sleep public repository, which was directly prepared by sleep experts at Hospitalar and University of Coimbra (CHUC). This dataset was exclusively prepared for sleep-related analysis research work. In this dataset, three different subgroups of data were contained which was obtained from different medical conditioned subjects, from a first subgroup-I section, as whole 100 subjects one session recording details were contained, similarly in subgroup-I section, data of 8 subjects were contained with two recording session per subjects and finally in subgroup-II, data of 10 healthy subjects were collected one session recording from each subject.

[0013a] All these recordings were recorded from subjects through whole-night that includes six EEG channels, two EOG channels and three EMG channels. The acquired signal recordings were performed with a sampling rate of 200 Hz and these recordings were obtained through a 10-20 international standard electrode placement system. In the present study the major focus to analysis the sleep irregularities happened during sleep through proper sleep stages classifications with the input of single-channel, C3 A2 of EEG signals, C3-A2 channel was selected in most recent sleep studies and several studies achieved higher classification accuracies with the input of C3-A2 channel.

[0013b] The invention disclosed here uses two different subgroups data from ISRUC Sleep dataset, 4 subjects are from ISRUC-Sleep subgroup-I, considered subjects were affected with different sleep problems, whereas other 4 data obtained from ISRUC Sleep subgroup-Ill section, in which the subjects were completely healthy controlled. In this work, we have obtained 8 polysomnography (4 subjects .IPSG +4 subjects.1PSG = 8 PSGs) records. Table 1 and Table 2 presented the different sleep stages epochs from sleep disordered subject and subjects completely healthy controlled.

[0013c] The experimental Data considered in the invention disclosed here is from two subjects as duplicated in the Table 1 and Table 2.

TABLE 1

Experimental Data Information: Sleep Disorder Affected Subject

Sleep Stages Number of Segments

Wake 847 NI Stage 446 N2 Stage 881 N3 Stage 434 REM 392

TABLE2

Experimental Data Information: Healthy Controlled Subject

Sleep Stages Number of Segments Wake 648 NI Stage 349 N2 Stage 942 N3 Stage 674 REM 387

[0014] The analysis is made manually to classify the sleep staging features in the inventions disclosed previously. Now the invention disclosed here is Novel Automated Machine Learning Algorithms Based System for Sleep Staging Features Analysis provides automatic analysis for the two categories of the Sleep Subjects.

SUMMARY OF INVENTION

[0015] Novel Automated Machine Learning Algorithms Based System for Sleep Staging Features Analysis is invented here provides Automatic Sleep Staging Features Analysis by the Machine Learning Algorithms Based System for Sleep Disordered and Healthy Subjects for sample size of each subject same, was 750 epochs and the length ofeachepochis the 30 seconds.

[0016] Sleep Stages behavior of Affected Sleep Problem Subject-15, Subject-16, Subject-19, and Subject-23 with a 30 second epoch length is analyzed by the Machine Learning Algorithms Based System.

[0016a] Sleep Stages behavior of Healthy Subject-1, Subject-2, Subject-5, and Subject-8 with a 30 second epoch length is analyzed by the Machine Learning Algorithms Based System.

[0016b] Sleep Stages behavior of Affected Sleep Problem Subject and Healthy Subject are considered as two categories in the invention disclosed here in to Analyze Sleep Staging Features automatically by the Machine Learning Algorithms based System.

[0016c] The Sleep Stages behavior of Affected Sleep Problem and Healthy Subject contains five stages namely Wake (201), N-REM1 (Non-Rapid Eye Movement 1) (202), N-REM2 (Non-Rapid Eye Movement 2) (203), N-REM3 (Non-Rapid Eye Movement 3) (204), and REM (Rapid Eye Movement) (205). Each subject is having these five stages comes finally under three classes namely Wake (201), Non-Rapid Eye Movement, and Rapid Eye Movement.

[0017] Totally twelve statistical features are extracted, in different combinations of five, nine and twelve feature set combinations are also extracted to understand the performance of the disclosed system.

[0018] The Present Invention disclosed here in is Novel Automated Machine Learning Algorithms Based System for Sleep Staging Features Analysis comprising of Human

Brain (101), Acquisition (102), Channel Data Segmentation (103), Preprocessing (104), Feature Extraction (105), Automatic Feature Selection (106), Classification (107), and Sleep Stages (108).

[0018a] From the Human Brain, EEG signals are acquired in two subjects called Sleep Stages behavior of Affected subject and Healthy Sleep subject. These acquired signals may suffer from artifacts and noise and are removed by principal component analysis and Butterworth filter after C3-A2 Channel Data segmentation.

[0018b] The input EEG signals are segmented into 30 seconds(6000 sample points) in the invention disclosed here, a statistical approach to extract the twelve time-domain features to characterize the signal properties from each segment of the input records is considered.

[0018c] The three Machine Learning Classification Algorithm such as Decision Tree (DT), K-Nearest Neighbor (KNN), and Random Forest (RF) are used in the invention disclosed to classify the sleep staging features and analyzing the features.

[0018d] The three Machine Learning Classification Algorithm such as Decision Tree (DT), K-Nearest Neighbor (KNN), and Random Forest (RF) are used to classify Sleep Stages behavior of Affected subject and Healthy Sleep subject.

[0019] For Category-I subject that is Sleep Stages behavior of Affected subject, three Machine Learning Classification Algorithm such as Decision Tree (DT), K-Nearest Neighbor (KNN), and Random Forest (RF) performing better accuracy of 93.42%, 94.4%, and 96.70% respectively.

[0020] For Category-I subject that is Healthy Sleep subject, three Machine Learning Classification Algorithm such as Decision Tree (DT), K-Nearest Neighbor (KNN), and Random Forest (RF) performing better accuracy of 93.42%, 94.4%, and 96.70% respectively.

[0021] Sleep stage classification is according to AASM sleep rules-based a single channel of EEG signals in the invention disclosed.

[0022] For both Category-I subject and Category-II subject, three Machine Learning Classification Algorithm such as Decision Tree (DT), K-Nearest Neighbor (KNN), and

Random Forest (RF) performing better accuracy of well above 90% proves the ability of the present invention disclosed.

BRIEF DESCRIPTION OF DRAWINGS

[0023] The Accompanying Drawings are included to provide further understanding of the invention disclosed here, and are incorporated in and constitute a part this specification. The drawing illustrates exemplary embodiments of the present disclosure and, together with the description, serves to explain the principles of the present disclosure. The Drawings are for illustration only, which thus not a limitation of the present disclosure.

[0024] Referring to Figure 1, illustrates Novel Automated Machine Learning Algorithms Based System for Sleep Staging Features Analysis, in accordance with an exemplary embodiment of the present disclosure.

[0025] Referring to Figure 2(a), illustrates Sleep Stages behavior of Affected Sleep Problem Subject-15 with a 30 second epoch length, in accordance with another exemplary embodiment of the present disclosure.

[0025a] Referring to Figure 2(b), illustrates Sleep Stages behavior of Affected Sleep Problem Subject-16 with a 30 second epoch length, in accordance with another exemplary embodiment of the present disclosure.

[0025b] Referring to Figure 2(c), illustrates Sleep Stages behavior of Affected Sleep Problem Subject-19 with a 30 second epoch length, in accordance with another exemplary embodiment of the present disclosure.

[0025c] Referring to Figure 2(d), illustrates Sleep Stages behavior of Affected Sleep Problem Subject-23 with a 30 second epoch length, in accordance with another exemplary embodiment of the present disclosure.

[0026] Referring to Figure 3(a), illustrates Sleep Stages behavior of Healthy Subject-1 with a 30 second epoch length, in accordance with another exemplary embodiment of the present disclosure.

[0026a] Referring to Figure 3(b), illustrates Sleep Stages behavior of Healthy Subject-2 with a 30 second epoch length, in accordance with another exemplary embodiment of the present disclosure.

[0026b] Referring to Figure 3(c), illustrates Sleep Stages behavior of Healthy Subject-5 with a 30 second epoch length, in accordance with another exemplary embodiment of the present disclosure.

[0026c] Referring to Figure 3(d), illustrates Sleep Stages behavior of Healthy Subject-8 with a 30 second epoch length, in accordance with another exemplary embodiment of the present disclosure.

[0027] Referring to Figure 4, illustrates Structural framework of statistical feature extraction, in accordance with another exemplary embodiment of the present disclosure.

DETAIL DESCRIPTION OF INVENTION

[0028] Referring to Figure 1, Novel Automated Machine Learning Algorithms Based System for Sleep Staging Features Analysis comprising of Human Brain (101), Acquisition (102), Channel Data Segmentation (103), Preprocessing (104), Feature Extraction (105), Automatic Feature Selection (106), Classification (107), and Sleep Stages (108). The invention disclosed here provides Automatic Sleep Staging Features Analysis by the Machine Learning Algorithms Based System for Sleep Disordered and Healthy Subjects for sample size of each subject same, was 750 epochs and the length of each epoch is the 30 seconds..

[0029] The Human Brain (101) is so sophisticated organ controls all the human activities including Sleep, can be understand by the recording of Electroencephalogram (EEG).

[0030] The Acquisition (102) is the process of obtaining the Electroencephalogram (EEG) signals from the Human Brain (101). The EEG signals recorded here are the dataset for the invention disclosed here.

[0030a] For Category-I subject that is Sleep Stages behavior of Affected subject, and for Category-I subject that is Healthy Sleep subject, EEG signals are recorded and are the classified into the different subjects like Subject-15, Subject-16, Subject-19,and Subject-23 in case of category-I subject, Subject-1, Subject-2, Subject-5, and Subject-8 in case of category-II.

[0031] The Channel Data Segmentation (103) used here is C3-A2 Channel Data Segmentation. The C3-A2 Channel Data Segmentation of the dataset of two categories having five sleep stages is listed in the Table 1 and Table 2. The input EEG signals are segmented into 30-seconds (6000 sample points).

[0032] The Preprocessing (104) is for removing the different types of artifacts and noise present in the recorded EEG signals from subjects during sleep hours. The general form of recorded signals is

Rs=Es+Ns Equation 1

Where Es represents the original EEG signal and Ns represent the noise signal.

[0033] The Extraction (105) uses a statistical approach to extract the several time domain features to characterize the signal properties from each segment of the input records. The statistical features are Mean, Variance, Maximum, Minimum, Range, Median, Mode, Standard deviation, Zero crossing, rate, Skewness, Kurtosis, and Third quartile.

[0033a] The invention disclosed here in uses three different experiments based on three different sets of feature vectors from all the 12 features in the process of classification and result from two sets are considered with the best weightage 9 features and 5 features respectively.

[0034] The Automatic Feature Selection (106) uses ReliefF feature selection algorithm for screening the relevant features which directly help to recognize the irregularities of sleep patterns during sleep. ReliefF is one of the supervised features weighting selection algorithm. It helps to analyze the extent to which features are most useful for discriminating between different stages and measures the effectiveness. The major advantage of this algorithm is to deal with unknown and redundant data.

[0034a] ReliefF feature selection algorithm for screening the best five and nine features, which helps to systematically screen the sleep characteristics. It is observed that the five-set features yield the best performance for both categories of subjects for identifying the sleep irregularities.

[0035] Classification (107) is with three Machine Learning Classification Algorithm such as Decision Tree (DT), K-Nearest Neighbor (KNN), and Random Forest (RF) are used to classify Sleep Stages behavior of Affected subject and Healthy Sleep subject.

[0035a] For Category-I subject that is Sleep Stages behavior of Affected subject, three Machine Learning Classification Algorithm such as Decision Tree (DT), K-Nearest Neighbor (KNN), and Random Forest (RF) performing better accuracy of 93.42%, 94.4%, and 96.70% respectively.

[0035b] For Category-I subject that is Healthy Sleep subject, three Machine Learning Classification Algorithm such as Decision Tree (DT), K-Nearest Neighbor (KNN), and Random Forest (RF) performing better accuracy of 93.42%, 94.4%, and 96.70% respectively.

[0035c] For both Category-I subject and Category-II subject, three Machine Learning Classification Algorithm such as Decision Tree (DT), K-Nearest Neighbor (KNN), and Random Forest (RF) performing better accuracy of well above 90% proves the ability of the present invention disclosed.

[0036] Referring to Figure 4, Structural framework of statistical feature extraction comprising of Single Channel C3-A2 Data of ECG Signal (301), Sleep Disordered Data (302), Subject (303), Feature Extraction (304), Healthy Controlled Data (305), Subject (306), and Feature Extraction (307).

[0037] The Single Channel C3-A2 Data of ECG Signal (301) performs Channel Data Segmentation (103). The C3-A2 Channel Data Segmentation of the dataset of two categories having five sleep stages is listed in the Table 1 and Table 2. The input EEG signals are segmented into 30-seconds (6000 sample points).

[0038] The Sleep Disordered Data (302) is obtained from the EEG Signal of Sleep affected person's human Brain.

[0039] The Subject (303) is generated based on the features and subjects are Subject-15, Subject-16, Subject-19, and Subject-23 in case of category-I subject.

[0040] The Feature Extraction (304) uses a statistical approach to extract the several time-domain features to characterize the signal properties from each segment of the input records. The statistical features are Mean, Variance, Maximum, Minimum, Range, Median, Mode, Standard deviation, Zero crossing, rate, Skewness, Kurtosis, and Third quartile. The Features are considered as three different sets of feature vectors from all the 12 features in the process of classification and result from two sets are considered with the best weightage 9 features and 5 features respectively.

[0041] The Healthy Controlled Data (305) is obtained from the EEG Signal of Sleep Healthy person's human Brain.

[0042] The Subject (306) is generated based on the features and subjects are Subject-1, Subject-2, Subject-5, and Subject-8 in case of category-II.

[0043] The Feature Extraction (307) uses a statistical approach to extract the several time-domain features to characterize the signal properties from each segment of the input records. The statistical features are Mean, Variance, Maximum, Minimum, Range, Median, Mode, Standard deviation, Zero crossing, rate, Skewness, Kurtosis, and Third quartile. The Features are considered as three different sets of feature vectors from all the 12 features in the process of classification and result from two sets are considered with the best weightage 9 features and 5 features respectively.

[0044] The present invention disclosed here in can guide clinicians to diagnosis accurately and take appropriate decisions for the treatment of the different types of sleep-related disorders.

Claims

NOVEL AUTOMATED MACHINE LEARNING ALGORITHMS BASED SYSTEM FOR SLEEP STAGING FEATURES ANALYSIS CLAIMS We claim:

1. Novel Automated Machine Learning Algorithms Based System for Sleep Staging Features Analysis comprising of Human Brain (101), Acquisition (102), Channel Data Segmentation (103), Preprocessing (104), Feature Extraction (105), Automatic Feature Selection (106), Classification (107), and Sleep Stages (108) provides Automatic Sleep Staging Features Analysis by the Machine Learning Algorithms Based System for Sleep Disordered and Healthy Subjects for sample size of each subject same, was 750 epochs and the length of each epoch is the 30 seconds.

2. Novel Automated Machine Learning Algorithms Based System for Sleep Staging Features Analysis as claimed in claim 1, wherein it uses Category-I subject that is Sleep Stages behavior of Affected subject, and Category-II subject that is Healthy Sleep subject, Category-I and Category-I by EEG signals are recorded and are classified into the different subjects like Subject-15, Subject-16, Subject-19,and Subject-23 in case of category-I subject, Subject-1, Subject-2, Subject-5, and Subject-8 in case of category-II.

3. Novel Automated Machine Learning Algorithms Based System for Sleep Staging Features Analysis as claimed in claim 1, wherein the input EEG signals are segmented into 30-seconds (6000 sample points), preprocessed by Principle Component Analysis and Band pass Butterworth Filter.

4. Novel Automated Machine Learning Algorithms Based System for Sleep Staging Features Analysis as claimed in claim 1, wherein it use twelve features namely Mean, Variance, Maximum, Minimum, Range, Median, Mode, Standard deviation, Zero crossing, rate, Skewness, Kurtosis, and Third quartile, Features are taken as set of five, nine and twelve to analyze the performance of the system as claimed in claim 1.

5. Novel Automated Machine Learning Algorithms Based System for Sleep Staging Features Analysis as claimed in claim 1, wherein it uses three Machine Learning Classification Algorithm such as Decision Tree (DT), K-Nearest Neighbor (KNN), and Random Forest (RF) provides accuracy of 93.42%, 94.4%, and 96.70% respectively for Sleep Stages behavior of Affected subject and accuracy of 93.42%, 94.4%, and 96.70% respectively for Healthy Sleep subject.