Artificial intelligence model detects asymptomatic Covid-19 infections through cellphone-recorded coughs

‘In a paper published recently in the IEEE Journal of Engineering in Medicine and Biology, the [MIT] team reports on an AI model that distinguishes asymptomatic people from healthy individuals through forced-cough recordings, which people voluntarily submitted through web browsers and devices such as cellphones and laptops.

‘The researchers trained the model on tens of thousands of samples of coughs, as well as spoken words. When they fed the model new cough recordings, it accurately identified 98.5 percent of coughs from people who were confirmed to have Covid-19, including 100 percent of coughs from asymptomatics — who reported they did not have symptoms but had tested positive for the virus.’

Read here (MIT News, Oct 29, 2020)

MIT revives project to build makeshift ventilators at US$400-500 using existing hospital supplies: An idea developing countries can consider too

‘Researchers at the Massachusetts Institute of Technology have revived a project from several years ago to develop a makeshift ventilator using “ambu” resuscitation bags, widely available at hospitals around the world. The team intends to share its design instructions free of charge, which will allow others to construct their own unit, using US$400-500 in supplies. The units are not currently FDA-approved, but the team hopes to obtain approval in the future."

This is the reasoning behind the project:

‘We are one of several teams who recognised the challenges faced by Italian physicians, and are working to find a solution to the anticipated global lack of ventilators. In the US alone, the COVID-19 pandemic may cause ventilator shortages on the order of 300,000-700,000 units (CDC Pandemic Response Plans). These could present on a national scale within weeks, and are already being felt in certain areas. An increase in conventional ventilator production is very likely to fall short and with significant associated cost (paywall warning).

‘Almost every bed in a hospital has a manual resuscitator (Ambu-Bag) nearby, available in the event of a rapid response or code where healthcare workers maintain oxygenation by squeezing the bag. Automating this appears to be the simplest strategy that satisfies the need for low-cost mechanical ventilation, with the ability to be rapidly manufactured in large quantities. However, doing this safely is not trivial.

‘Use of a bag-valve mask (BVM) in emergency situations is not a new concept. A portable ventilator utilizing an ambu-bag was introduced in 2010 by a student team in the MIT class 2.75 Medical Device Design (original paper here and news story here), but did not move past the prototype stage. Around the same time, a team from Stanford developed a lower-cost ventilator for emergency stockpiles and the developing world. It looks similar to a modern ICU ventilator (Onebreath), but “production for US hospitals would start [in] about 11 months”, making it “a second wave solution” (MIT Tech Review Article).

‘Last year, the AMBU®️ Bag concept was re-visited by two student teams, one from Rice university (here & here), and another Boston-based team who won MIT Sloan’s Healthcare prize (MIT News: Umbilizer).’

Read here (MIT, April 1, 2020)