TherapyAssist takes sophisticated but ‘simple’ approach to sleep therapy

 - 
Friday, October 16, 2020

PITTSBURGH – sovaSage has launched TherapyAssist, a software that its co-founder says optimizes the sleep therapy process, including finding patients the right masks, using artificial intelligence and computer vision. 

“It’s well acknowledged that the right mask fit is, not the only indicator, but the No. 1 indicator for determining whether or not therapy succeeds or fails,” said William Kaigler, who has a 30-year career in the HME industry, which includes working at Philips Respironics and founding medSage, which sold to Philips in 2011. “It’s also the most laborious part of the process.” 

Put simply, TherapyAssist sends a patient a text with a link that opens the TherapyAssist app, prompting them to answer a short series of questions and take, essentially, a selfie using their smart phone. 

Then the software’s computer vision technology maps out a patient’s anatomical characteristics and its AI mines their answers to questions, as well as data from the clinical outcomes of previous patients that sovaSage has been collecting for more than two years, Kaigler says.  

“What the algorithm does – it does not make that decision for the RT – but it presents a set of options prioritized based on what the algorithm recommends,” said Kaigler, who also teaches at Carnegie Mellon University. 

Kaigler believes TherapyAssist’s ability to recommend masks from any manufacturer, as well as its technology, sets it apart in what’s becoming an increasingly competitive field of digital tools for sleep therapy. 

“It’s simple,” he said. “We don’t need 3D cameras; we don’t need a ruler in the picture; we don’t even need people to hold their phone a certain distance.” 

TherapyAssist also goes beyond mask selection, Kaigler says, taking a more encompassing approach to therapy. 

“There’s also a monitoring tool,” he said. “One example is the question, do you feel drowsy after driving for an hour? Instead of asking that at some random time, why not ask that right after they’ve taken that drive. We have context sensitive monitoring. And all that monitoring feeds back into the algorithm.” 

While the COVID-19 pandemic has put a new emphasis on digital tools, Kaigler was at work on TherapyAssist three years ago, driven by the goal of increasing the number of patients RTs can see in a day, reducing masks exchanges and improving patient compliance. 

“We have customers who will use our software and still meet with the patient,” he said. “But now instead of going through the process of selecting and sizing a mask, that’s already done, and they can move on to other things.”