Quvium has partnered with DSP Concepts to create a wearable device that monitors the coughing sounds of patients with respiratory ailments to alert of an acute attack.

Quvium, a UK startup, has partnered with DSP Concepts, a Silicon Valley firm that sells digital signal-processing technology, to create a wearable device that monitors the coughing sounds of patients with asthma, chronic obstructive pulmonary disease (COPD) or other chronic respiratory ailments, in order to alert caregivers when a patient is likely to have an acute attack that could result in hospitalization. Dubbed CoughAware, the device will soon be validated by physicians and technologists at London’s King’s College, in advance of a clinical trial in the United Kingdom.

“We looked at the progression of respiration problems and found that there are indicators of an [acute attack] beginning six to eight hours before it occurs,” says Steve Schmidt, Quvium’s CEO. He and James Collins, Quvium’s product development lead, began developing an algorithm designed to discriminate coughing indicative of asthma attacks or acute events related to other ailments, including COPD and cystic fibrosis. They then worked with DSP Concepts to integrate this algorithm into other sound-processing algorithms—which could discriminate a cough from, say, a barking dog—and DSP Concepts’ machine-learning software, all of which run on the wearable device.

The device contains a microphone, an ARM Cortex-M7 processor that runs the DSP Concepts software, a subscriber identity module (SIM) card, and a battery that powers the device and is recharged via a USB Micro2 port. It has no moving parts and is roughly the size of a large automobile key fob. To initiate the device and pair it to a patient, he or she coughs three or four times near the device. According to Schmidt, the algorithms collect and analyze the sound of the coughs to personalize it to the user based on its pitch and sound profile, which is unique to each person based on the size and capacity of his or her lungs.

Get the full story at www.iotjournal.com