According to the World Health Organisation, in 2015, 2.8 million people developed Tuberculosis in India. The organisation also notes that two deaths occurring every three minutes in India is due to TB. However, with proper care and treatment, TB is a disease that can be completely cured. And early detection is vital in ascertaining results.
And that’s why an app called TimBre for Tuberculosis could play a vital role moving forward.
While it notes that it is a screening tool and not a diagnostic one, TimBre can provide clarity to anyone with access to a mobile phone with an internet connection. The user is expected to cough into the phone three times and also input certain other details about themselves. Then the app will analyse the data and report the results through an SMS and an email.
How the app works:
The developer of the app Rahul Pathri, in a video posted on YouTube, explains, “We are currently leveraging mobility, machine learning analytics, deep learning and spectral analysis to come up with a non-invasive solution where a suspect coughs into a mobile phone or a smart phone with an internet connection; inputs clinical demographic and socio-economic variables.”
Sound features are then extracted and the data is processed through proprietary algorithms to understand the patterns behind a pathological cough and that is combined with clinical demographic variables to get a result.
The team consists of a spectral analyst, a data scientist and a pathologist to ensure the veracity of its results, but the developers urge people to go to the doctor to run proper diagnostics when symptoms persist despite a negative result from the app.
According to Rahul, curbing the incidence of TB in India is a personal mission for many in the team due to their own personal experiences. He says, “Most recently some of our team members or family members have been affected by TB and that’s the reason we think we understand the various nuances.”
The explainer video can be viewed below
The app can be downloaded here