Consult an Expert
Trademark
Design Registration
Consult an Expert
Trademark
Copyright
Patent
Infringement
Design Registration
More
Consult an Expert
Consult an Expert
Trademark
Design Registration
Login
Analysis of Wrist Pulse Signals to Identify Type II Diabetic Subjects Non-Invasively Using Recurrence Quantification Technique
Extensive patent search conducted by a registered patent agent
Patent search done by experts in under 48hrs
₹999
₹399
Abstract
Information
Inventors
Applicants
Specification
Documents
ORDINARY APPLICATION
Published
Filed on 19 November 2024
Abstract
A wrist pulse signal comprises dynamic evidence about an individual’s health status. The preliminary studies verify the linking of several diseases with wrist pulses. Hence, this research focuses on distinguishing subjects with Type II diabetes and non-diabetes using nonlinear methods on wrist pulse signals. A total of 600 subjects contributed to the study. The wrist pulse recorded at three wrist positions on the radial artery showed many variations for Type II diabetic subjects. At the same time, there were not many variations for non-diabetic subjects. The relation between the three positions of the wrist pulse is obtained to understand the effect of pulse variation. Another experiment is conducted to analyse the impact of hypertension and obesity on Type II diabetic conditions using nonlinear signal processing techniques. A multichannel wrist pulse acquisition system was designed, and data were collected on 600 subjects. Analysis of signals recorded from the multichannel wrist pulse acquisition system for different experimental conditions showed comparable results with expected results. Nonlinear methods were more effective than spectral analysis methods in understanding the variations and dynamics of the wrist pulse signals. The average of the derived parameter from nonlinear signal processing technique is used as input for the classification algorithms, and performance measures were analysed. The design of a low cost, simple multichannel wrist pulse acquisition system is proven effective for pulse diagnosis. The integration of the acquisition system with the nonlinear signal processing techniques may lead to a robust, portable system that can record wrist pulse signals and aid in the non-invasive diagnosis of many pathological conditions of the human body.
Patent Information
Application ID | 202441089746 |
Invention Field | BIO-MEDICAL ENGINEERING |
Date of Application | 19/11/2024 |
Publication Number | 48/2024 |
Inventors
Name | Address | Country | Nationality |
---|---|---|---|
Dr. S. Hema Priyadarshini | Department of Medical Electronics Engineering, Dayananda Sagar College of Engineering, Bangalore-560111 | India | India |
Dr. Anand Prem Rajan | Department of Biomedical Sciences, School of Biosciences and Technology, Vellore Institute of Technology, Vellore | India | India |
Applicants
Name | Address | Country | Nationality |
---|---|---|---|
Dayananda Sagar College of Engineering | Shavige Malleshwara Hills, Kumaraswamy Layout, Bangalore | India | India |
Specification
Description:FIELD OF INVENTION
[001] The present invention relates to the field of biomedical signal processing and diagnostic healthcare. Specifically, it pertains to the non-invasive identification of Type II diabetes using wrist pulse signals.
BACKGROUND AND PRIOR ART
[002] The literature review focussed on the classical theories behind the ancient use of Nadi Pariksha and the modern glimpse of wrist pulse acquisition systems available to analyze signals obtained at the wrist using various signal processing procedures. It can be observed that most of the studies are concentrated on the wrist pulse signal as a single pulse or palpation felt at the wrist. Hence, wrist pulse signals are said to be used effectively to understand the relationship between hypertension, obesity, and type II diabetes which may help detect the onset of these disorders in the body. Only a few studies have tried to elaborate on the effect of using wrist pulse signals to identify diseases or disorders of the human body. Since the traditional method of diagnosis of diseases is related to the Nadi or pulse, studying the effect of utilization of the pulse for differentiating between non-diabetic and type II diabetic will be interesting from the perspective of non-invasive diagnosis.
SUMMARY OF THE INVENTION
[003] The key objective of this work was to use nonlinear techniques of signal processing to analyse the non-diabetic and type II diabetic subjects using signals acquired at the wrist. Signal analysis techniques such as recurrence plots and their derived parameter for analysing the wrist pulse signals have been used. Classification techniques using machine learning algorithms are used to classify the non-diabetic and type II diabetic subjects. It has also been observed and identified that the wrist pulse signals show variations for hypertension and obesity on type II diabetic subjects. The uniqueness of the research work lies specifically in the acquisition of wrist pulse signals and the analysis techniques used.
[004] The relation between the wrist pulse signals' three positions was analyzed for subjects with type II diabetic and non-diabetes (N=600). The differences between the three positions were analyzed using recurrence technique. It was seen that the non-diabetic subjects showed repeated patterns for the three positions indicating higher repeatability and less chaos in the wrist pulse signal. There was an increase in entropy in position 2 of type II diabetic subjects. The results imply that the pancreas' insulin production is less and has high blood sugar.
[005] The average entropy of the recurrence quantification analysis indicated the relationship between hypertension and obesity for subjects with type II diabetes. These parameters were more for subjects of type II diabetes with obesity since type II diabetic subjects have less insulin production, thus increasing blood glucose levels. For obese subjects, the body needs more insulin to convert excess glucose to glycogen, and hence more glucose particles get accumulated in the bloodstream, which gets reflected in the wrist pulse signals indicating more abnormality and chaos. Similarly, type II diabetic subjects with hypertension show higher non-recurrence and randomness in the system.
[006] The classification algorithm was used to classify non-diabetic from type II diabetic subjects using wrist pulse signals. The derived parameters of the wrist pulse recorded at three positions are used as a predictor. The responses are set as '1' for non-diabetic subjects and '0' for type II diabetic subjects. The performance of the algorithm is visualized through a confusion matrix. It was found that the K-nearest neighbour algorithm showed the highest classification accuracy.
[007] The wrist pulse at three positions on the wrist can be acquired using a system designed using optical sensors and a microcontroller. A simple circuit was used to remove noise. PLX-DAQ converted the data recorded to excel form. Signal processing techniques were applied using MatLab. The components used to design the multichannel wrist pulse acquisition system are less expensive, and the design is less complex. Data were recorded from various subjects for analysis and validation.
[008] The variations in the parameters derived from various signal processing techniques have shown an adequate change among the type II and non-diabetic subjects. Since the variations were also observed considering the various effects on type II diabetic subjects, the classification techniques can aid in the prediction of the conditions significantly. Thus, it may be inferred that nonlinear parameters can be used effectively to analyze different conditions of wrist pulse signals. Analysis of signals from the multichannel wrist pulse acquisition system showed comparable results with the expected results.
BRIEF DESCRIPTIONS OF DRAWINGS
[009] The multichannel wrist pulse acquisition system's schematic diagram is illustrated in Fig.1. The optical sensor transmits and receives the light related to the pulse transmitted in the radial artery. The obtained sensed data is filtered and passed to the microcontroller. The microcontroller receives the analog data at three analog input pins where the sensors are connected and converts the data into digital output, which can be read at the digital output pins. The discrete data is changed to excel using PLX-DAQ converter and used for analysis using various signal processing techniques.
[010] The procedure for pulse signal acquisition from the wrist radial artery at three positions is shown in Fig. 2. The pulse location is felt from the finger, and sensors are placed accordingly for better results. When the three wrist pulse signals are detected, read the data from the three optical sensors to the three analog input pins of the microcontroller. The obtained information is converted into excel form using PLX-DAQ. The excel data stored is used in MatLab, which can be further processed using signal processing techniques.
[011] Average entropy derived from the recurrence quantification analysis (non-linear signal processing technique) is significantly increased in subjects with type II diabetes and decreased for non-diabetic subjects. Increased entropy in subjects with type II diabetes indicates less predictability (p<0.001for P1 and P2) and more chaos in the underlying nonlinear system as in Fig. 3. The results obtained are similar to a study found in literature where the sample entropy was used to classify the normal and diabetic subjects from the wrist pulse signals acquired by MLT100 piezoelectric transducer from AD instruments.
[012] Hypertension and type II diabetes are associated with the cardiovascular system and the pancreas. Blood with high glucose levels travels through the body, and the blood vessels lose their ability to stretch. Hence, the wrist pulse signals acquired in positions one and two represent the Vata and Pitta dominance in the Tridosa, measured as average values of entropy. The average entropy is significantly higher for subjects with type II diabetes and hypertension than for subjects with type II diabetes and non-hypertension, with p<0.001 for position one (P1) and position two (P2); and p<0.01 for position three (P3). Aortic deterioration may cause systolic hypertension. There is an increase in the amplitude of the second systolic pressure wave in the radial artery, which indicates elevated arterial pressure due to the swelling of the brachial artery and its radicles as a result of inside pressure.
[013] Average entropy is more for obese type II diabetic subjects when compared to non-obese subjects with type II diabetes, with p<0.001 for position one and position two and p<0.01 for position three, indicating higher non-recurrence and randomness. For obese subjects, the body needs more insulin to convert excess glucose to glycogen. Hence, more glucose particles accumulate in the bloodstream, reflected in the wrist pulse signals, indicating more abnormality and chaos. The non-obese type II diabetic subjects have reduced insulin section and insulin resistance when compared to obese type II diabetic subjects.
[014] The classifier was trained with 75% of the data and tested with 25% of the data out of 600 data collected. The average values of the recurrence quantification parameters were set as predictors, and 0 and 1 were set as the response of type II diabetes and non-diabetic, respectively. The classifier models were built and tested. The percentage of K Nearest Neighbour's test accuracy was better than other classifications. The performance of the K Nearest Neighbour was measured using the region of the curve.
[015] The parameters such as sensitivity, F score, specificity, precision and recall are measured from the confusion matrix derived values of true positive, false positive, true negative and false negative for each of the classifiers derived from the confusion matrix. KNN works best when there are fewer features than SVM, which requires a larger number of features. Therefore, the K-Nearest Neighbour classification technique can be used as a classifier of non-diabetic and type II diabetic subjects using wrist pulse signals.
DETAILED DESCRIPTION OF THE INVENTION
[016] The ancient research in Ayurveda, Tibetan medicine, Mongolian medicine, Siddha medicine, Chinese medicine and Unnai are all based on pulse diagnosis, which focuses on sensing the palpations of the pulse at the radial artery in the wrist using the fingertips of the practitioners. In Ayurveda, the three pulse positions, such as Vata, Pitta and Kapha, reflect the health condition of the body's physiological systems. This non-invasive diagnostic approach requires modern techniques and methodologies to analyse wrist pulse signals. This research mainly dealt with processing pulse signals acquired from the wrist to differentiate subjects without diabetes and type II diabetes using nonlinear signal processing technique. The pre-processing for the acquired wrist pulse signals has been carried out for the removal of noise artifacts. The experimental results observed that the wavelet transform technique is much suited for high-frequency noise removal in the wrist pulse signals, giving comparatively better results with less computation.
[017] The variations in the wrist pulse signals recorded for three positions were visually observed in the recurrence plot, and the correlation between the three positions was well understood by the parameters such as recurrence rate, entropy, diagonal line length, divergence and ratio of determinism and recurrence rate derived from the recurrence plot using recurrence quantification analysis techniques. The entropy shows more substantiable changes and hence considered for the analysis and representation. These parameters were highly correlated between the positions with p<0.001 for non-diabetic subjects compared to diabetic subjects. The effect of hypertension and obesity on type II diabetic subjects was analysed from the wrist pulse signals. The results were validated using the conventional power spectrum and derived energy ratio and reflection index parameters. The nonlinear method was more advantageous in understanding the dynamics of the wrist pulse signals since the processing is faster and differentiates chaos in the complex system efficiently and accurately.
[018] A multichannel wrist pulse acquisition system was intended to record the pulse signals at the wrist could identify the differences between the healthy and unhealthy conditions of the subjects. Classification techniques such as ensemble bagged trees, finite decision tree, logistic regression, KNN, and SVM were used to classify the type II diabetic and non-diabetic subjects based on the nonlinear parameters. The K-nearest neighbour technique had a higher accuracy than the other classification techniques. The wrist pulse signals analysis carried out in this work provides a good understanding of variations in pulse signals at the wrist for unhealthy and healthy conditions. The nonlinear signal processing method using recurrence quantification technique can be effectively used to observe and detect pulse patterns in specific unhealthy conditions; hence can be used to promote a non-invasive ancient approach for analysing wrist pulse signals. , C , Claims:[019] 1. The device developed able to Record the wrist pulse signals using three channel acquisition system. Able to identify the type II diabetic and non-diabetic subjects using the wrist pulse signals non-invasively using nonlinear signal processing technique.
[020] 2. Able to Identify the effect of hypertension on type II diabetic subjects using average entropy derived from recurrence quantification technique.
[021] 3. Able to identify the effect of obesity on type II diabetic subjects using average entropy derived from recurrence quantification technique.
[022] 4. Claims (2) to (4) are analysis used for identification of the effect and prominence of the pulse variation. The subject is identified to be diabetic or non-diabetic non-invasively by using pulse signals obtained from the wrist. The type II diabetic subjects can have effect of hypertension and obesity associated which is shown to be effectively recognised by considering the wrist pulse signals.
Documents
Name | Date |
---|---|
202441089746-COMPLETE SPECIFICATION [19-11-2024(online)].pdf | 19/11/2024 |
202441089746-DRAWINGS [19-11-2024(online)].pdf | 19/11/2024 |
202441089746-FORM 1 [19-11-2024(online)].pdf | 19/11/2024 |
202441089746-FORM 18 [19-11-2024(online)].pdf | 19/11/2024 |
202441089746-FORM-9 [19-11-2024(online)].pdf | 19/11/2024 |
202441089746-REQUEST FOR EXAMINATION (FORM-18) [19-11-2024(online)].pdf | 19/11/2024 |
Talk To Experts
Calculators
Downloads
By continuing past this page, you agree to our Terms of Service,, Cookie Policy, Privacy Policy and Refund Policy © - Uber9 Business Process Services Private Limited. All rights reserved.
Uber9 Business Process Services Private Limited, CIN - U74900TN2014PTC098414, GSTIN - 33AABCU7650C1ZM, Registered Office Address - F-97, Newry Shreya Apartments Anna Nagar East, Chennai, Tamil Nadu 600102, India.
Please note that we are a facilitating platform enabling access to reliable professionals. We are not a law firm and do not provide legal services ourselves. The information on this website is for the purpose of knowledge only and should not be relied upon as legal advice or opinion.