Non-Invasive Tuberculosis Screening
Docturnal has launched a beta version of its flagship product “TimBre” that screens/detects Tuberculosis (TB) primarily focusing on Pulmonary Category TB.
By recording and analysing a patient’s cough via a mobile app, this technology offers a non-invasive, fast and efficient way to diagnose patients with the disease, helping to dramatically speed up detection, while simultaneously reducing the gestation period and the threat of the infection spreading.
The app is available on Android mobile devices & support for iOS and Windows is in the offing and has been designed to screen a patient’s cough via a special microphone array plugged into the mobile device. To ensure accurate screening, this diagnostic tool has been designed for use by medical practitioners in controlled environments free of extraneous noise.
Using machine learning and acoustic engineering combined with clinical and demographic variables, the app’s algorithms not only register whether a person has TB, but are also able to determine whether the cough signifies other issues like Chronic Obstructive Pulmonary Disease (COPD), Pertussis, Bronchitis or Asthma once the baseline Training Data reaches a significant size. Currently, the focus is on TB & associated variants. To ensure a more accurate diagnosis, the app also captures demographic and socioeconomic information from the patient, including details like their occupation (whether they work with Asbestos or in coal mines for instance) and other background details that might be relevant to their result. Most importantly, the socioeconomic variables may also pave way for determining a patient’s progression into Multi Drug Resistant (MDR), TB.
The product is undergoing several beta deployments in Hyderabad/India, where they are testing its accuracy via a cross validation methodology, where positively screened cases are sent for industry standard TB testing to confirm the accuracy of the app’s diagnosis.
Some media articles around this App:
NASSCOM : http://www.nasscomfoundation.org/get-engaged/nasscom-social-innovation-forum/nsif-2017-finalists/non-invasive-tuberculosis-detection-timbre.html