Rich Fletcher, an MIT research scientist, and Dan Chamberlain, a graduate student at MIT, in collaboration with the Chest Research Foundation (CRF) recently received the 3rd prize in Vodafone’s annual Wireless Innovation Project awards program worth $100,000. Each year, three awards are given to the top wireless innovations that can be used as a worldwide social benefit.
Fletcher’s team, which is based at MIT D-Lab and at the Tata Center for Technology and Design, received the award for advancing a low-cost mobile platform to diagnose pulmonary diseases. About one-fifth of the world’s population is affected by some kind of pulmonary disease such as asthma, tuberculosis, pneumonia, lung cancer and chronic obstructive pulmonary disease (COPD). COPD is responsible more than 14 percent of deaths globally.
Pulmonary diseases are a major burden for developing countries where the quality of the air is often poor. In addition, the majority of doctors are not trained for pulmonary disease diagnosis and often resort to homeopathy or traditional medicine, since testing equipment and modern therapies are too expensive or difficult to procure.
“We were surprised to learn that there are only approximately one dozen comprehensive pulmonary testing labs in the entire country of India. Clearly, most of India’s population is diagnosed using very simple tools, which can result in misdiagnosis or underdiagnosis,” said Fletcher. “To address this problem, various research groups around the world have developed mobile versions of a spirometer. However, a spirometer can only diagnose certain types of pulmonary disease and requires too much physical effort from the patient. As a result, spirometers are only used by about 20 percent of general practitioner doctors in India, and cannot be used with sick children or elderly patients.”
Fletcher, Chamberlain and the Chest Research Foundation developed a solution that consists of using a mobile phone to capture data and predict the odds of a patient having a specific pulmonary disease. Chamberlain explained: “There are two layers of algorithms used in the project: The first set of machine learning algorithms are used to automatically recognize and classify the different types of abnormal lung sounds, and a second layer of algorithms are then used to combine the lung sounds with other data to predict the likelihood of specific diseases.”
The major challenge was for engineers to capture some of the knowledge and the capabilities of an experienced pulmonologist in order to embed it into a mobile phone app that could easily and effectively aid physicians.
“There is a wealth of innovation happening at MIT in the area of technology and design for low-resource settings. The Tata Center has helped to create a community of researchers across MIT connected by this common purpose, and we are happy to see this work receive international recognition,” said Rob Stoner from the Tata Center.