MIT researchers use mHealth tools to identify COVID-19 in cough

November 3, 2020 -Researchers at the Massachusetts Institute of Technology are developing an mHealth tool that can detect evidence of COVID-19 in a cough.

In a recently published study IEEE Journal of Engineering in Medicine and BiologyThe team of researchers said,[Usingtensofthousandsofsamplesofpeoplewhoarecoughingortalkingthereisa985%chancethatpeoplewhotestpositiveforcoronaviruscanpinpointcoronavirussymptomsIwasabletocreateanAItoolIfasymptomatic[peoplecoughingandtalkingandwereabletocreateanAItoolthatcouldaccuratelyidentifycoronavirussymptoms985percent0047timeamongthosewho’dtestedpositiveforthevirusandineveryasymptomaticcase[咳や会話をしている人々の数万のサンプルを使用し、コロナウイルスの検査で陽性を示した人の985％の確率で、コロナウイルスの症状を正確に特定できるAIツールを作成することができました。無症候性の場合。[peoplecoughingandtalkingandwereabletocreateanAItoolthatcouldaccuratelyidentifycoronavirussymptoms985percentofthetimeamongthosewho’dtestedpositiveforthevirusandineveryasymptomaticcase

“Both speaking and coughing sounds are affected by the vocal cords and surrounding organs,” said Brian Subirana, a research scientist and co-author of the study at MIT’s Auto-ID Institute. Said in a press release Published by the university. “This means that when you speak, part of your speech is like a cough, and vice versa. Also, something that can be easily derived from a fluent speech, AI is a person. It also means that you can easily pick up from your cough, such as your gender, native language, and even your emotional state. In fact, the way you cough has emotions embedded in it. “

The idea of ​​detecting a cough disease is not new. 2014, ResApp Health Developed the mHealth app At the University of Queensland, Australia, it may help clinicians identify dyspnea such as pneumonia, bronchitis, croup, and asthma in children and adults who cough on their smartphones.

As the world is on the verge of a pandemic, healthcare systems and digital health companies like Fitbit Development of telemedicine and mHealth platform Early signs of the virus can be detected through wearable-tracked vital signs, activity, and sleep patterns.

Amy McDonough, General Manager and Senior Vice President, Fitbit Health Solutions, said the company’s recent partnership with Northwell Health and the United States Army on Connected Health: business. “We believe Fitbit can reliably detect these signals, and prior to this, to warn people that they may get sick before unknowingly spreading it to others. It gives us a great opportunity to help. “

The Mayo Clinic is also researching ways to identify illnesses using voice-based medical biomarkers. Recently, the healthcare system announced a partnership with Vocalis Health to develop new tools for monitoring patients with respiratory problems such as pulmonary hypertension.

“We have seen the clinical benefits of voice analysis for patient screening throughout the COVID-19 pandemic. This collaboration provides an opportunity to continue expanding research, starting with pulmonary hypertension,” said Israel CEO. Co-founder Tal Wenderow said. Bass Vocalis HealthSaid recently Israeli era.. “Voice analysis can help physicians make more informed decisions about patients in a non-invasive and cost-effective way. This technology will help telemedicine and remote in the very near future. We believe it has the potential to have a significant clinical impact on patient monitoring. “

At MIT, Subirana and his colleagues have developed an AI tool to measure the strength of a person’s vocal cords, focusing on a tool that uses a person with Alzheimer’s disease as a model to detect apparent emotional states by voice. I did. On top of that, they hierarchized a database of various coughs.

The resulting AI tool identifies four biomarkers unique to COVID-19. Vocal cord strength, emotions, lung and breathing performance, and muscle deterioration.

“I think this shows that even asymptomatic, using Covid changes the way sound is produced,” says Subirana.

Researchers are currently working on creating an mHealth app that will allow care providers and patients to use the tool on smartphones and tablets, perhaps ultimately on smart speakers and other digital health platforms in the home. We are also planning partnerships with large healthcare systems to enhance and extend our database to identify other conditions.

“If the pre-screening tools are always on in the background and constantly improving, the pandemic can be a thing of the past,” they concluded in the study.