In testing SpiroSmart in clinics in India and Bangladesh, University of Washington computer science & engineering doctoral student Mayank Goel (front row, center) and collaborators learned that many patients did not have access to a smartphone. Credit: University of WashingtonThroughout the world hundreds of millions of people suffer from chronic respiratory diseases. According to the World Health Organization, 3 million people die each year from chronic obstructive pulmonary disease (COPD), and 235 million people suffer from asthma.

“Many of these diseases are chronic and the patients need to manage their condition throughout their lives,” Mayank Goel, a computer science and engineering doctoral student at the University of Washington, told R&D Magazine. “So, it is extremely important to improve the state of diagnosis and management of these diseases.”

Home spirometers, an instrument that measures the amount and speed of airflow to gauge lung function, allows patients to monitor their lungs. However, the cheapest apparatuses still cost around $800, according to Goel.

Five years ago, Goel and colleagues decided to leverage a piece of prominent technology, and utilize it to measure lung function. The smartphone.

“Almost every person has a phone these days and all of these phones have a microphone on them,” said Goel. “For us this was a perfect opportunity to convert those microphones into spirometers.

In 2012, SpiroSmart was introduced. When using this device, patients simply take a deep breath and then exhale, hard and fast. The smartphones microphone senses the exerted sound and pressure, and sends that data to a central server, which employs machine-learning algorithms to convert the data into standard lung-function measurements.

SpiroSmart has collected data from more than 4,000 patients, from people in the Seattle area to those in India and Bangladesh.

“We are in the FDA clearance process right now and plan to bring the technology out as soon as we are through the process,” said Goel. 

“We initially focused on smartphones,” he added, “and once we were confident that our approach works, we moved to the harder problem of enabling similar sensing on non-smartphones; basically any phone in the world.”

Thus SpiroCall, a patient call-in service, was born. At the upcoming Association for Computing Machinery’s CHI 2016 conference, set to be held in San Jose, Calif. this month, the researchers, which include Prof. Shwetak Patel and electrical engineering student Elliot Saba (both of the University of Washington), will present their paper on the new development.

“Unlike SpiroSmart, (SpiroCall) transmits the collected audio using the standard voice telephony channel,” the researchers wrote in the paper. “A server receives the data of degraded audio quality and calculates clinically relevant lung function measures and reports to the participants using audio or text message … SpiroCall combines multiple regression algorithms to provide reliable lung function estimates despite the degraded audio quality over a voice communication channel.”

After testing SpiroCall with 50 participants and against two medical spirometers, the researchers found the results were within 6.2 percent of the same patient’s results from the commercial spirometer.