Scope and Objectives

Obstructive sleep apnea (OSA) is caused by complete or partial obstruction of the upper airway and affects adult population seriously. It is characterized by repetitive episodes of shallow or paused breathing during sleep. The very cessation (apnoea) or reduction (hypopnea) of airflow can lead to morbidities like Hypertension, Diabetes, Heart Failure, Coronary Artery Disease and Stroke. The current mode of confirmation of OSA is through Polysomnography (PSG) that requires hospitalization and a data collection process found inconvenient by subjects. This project proposes development of an affordable, convenient, home-care option where breathing sound will be recorded by an array of microphones placement for USA detection. Simultaneous measurement of the pulse oximetry and non-contact image data for the respiratory movement will make the system perform better to evaluate apneic and hypopneic episodes. Besides monitoring and recording, appropriate alarm can be generated by the proposed system for clinical decision support.

Deliverables

Projected outcome is development of an inexpensive, easy to operate, and automated homecare system to detect Obstructive Sleep Apnea (OSA). The proposed system will be capable of recording breath sound and storage of recorded audio signal. The system will consist of an auditory unit and a display unit as well. The signal processing engine will classify mild, moderate and severe OSA groups by discriminating the breathing, snoring and non-snoring sound signals using machine learning approaches.

Scientific Output

We do not want to put the intermediate findings in the public domain. It will be put in due course after securing the IP / peer review. The findings were put before the expert reviewers in the last review meeting.

Results and outcome till date

We have been able to develop a 4-channel data recording system and tested it for overnight data acquisition. We have filed a provisional patent for the same.

Societal benefit and impact anticipated

Obstructive Sleep Apnea (OSA), suffered by about 7.5% population in India, can lead to (a) micro-fragmentation of sleep, (b) an over-activity of the sympathetic nervous system, (c) reactive pulmonary vasoconstriction, and (d) others including systemic inflammation and endothelial dysfunction. These physiological changes can cause an array of symptoms and lead to various morbidities including hypertension, glucose intolerance, cardiovascular diseases, stroke, and others and can also result in sudden cardiac death. The current mode of investigation of OSA requires costly sleep study in hospital care. For this, a large number of sensors e.g. EOG, EEG, ECG, pulse oximetry, respiratory movement analyzers, thermos-sensors are connected to subject which affects the sleep of the subject. The attempts to develop an affordable, convenient, home-care based solution to detect OSA that remains under-diagnosed, specially for our elderly and marginalized population, to improve their quality of life.

Next steps

We are trying to add an auxiliary channel to the data acquisition system after which complete patent specification will be done. This will follow hospital data acquisition and development of signal processing algorithms.

Publications and reports

Patents

One provisional patent applied.

Scholars and Project Staff

Two project staffs joined on Jan 19, 2018 and February 06, 2018, respectively. Earlier, one project staff who joined on August 04, 2017, resigned on December 22, 2018 to join MTech Program in an IIT starting January, 2018.

Challenges faced

It required four round of advertisements to get two project staffs working in the project. This effectively reduced the man hour that could have been provided in the project by this time to a great extent.

Other information

Rs 10 Lakh is booked from the available fund to start hospital data acquisition. A request is made to IMPRINT Central Team on temporary use of the Equipment Grant for another IMPRINT project supported by same Ministry.