Denver Tech Center, Colorado, May 27, 2025 — RAIsonance, Inc. today announced results from its latest, highly successful field validation of its market-leading respiratory screening solution, used for the detection of Tuberculosis, COVID-19, and Pneumonia in the sound of a person’s cough. The solution, called AudibleHealth AI, uses proprietary artificial intelligence to detect intricate patterns in cough samples that correlate to specific respiratory diseases, and is currently live in 7 countries.
The AudibleHealth AI solution was designed to meet the WHO’s criteria for a screener which calls for 90% sensitivity (positive accuracy) and 70% specificity (negative accuracy). The solution met or exceeded these requirements for Tuberculosis as per the table below. The AudibleHealth AI solutions designed to screen for COVID-19 and Pneumonia were also evaluated on the participant population, and achieved very strong results:
The AudibleHealth AI solution was designed to meet the WHO’s criteria for a screener which calls for 90% sensitivity (positive accuracy) and 70% specificity (negative accuracy). The solution met or exceeded these requirements for Tuberculosis as per the table below. The AudibleHealth AI solutions designed to screen for COVID-19 and Pneumonia were also evaluated on the participant population, and achieved very strong results:
The study wrapped up in May 2025, having taken place in Ukraine. Completed enrollment was 379 participants and was done in collaboration with the Department of Infectious Diseases and Physiatrics at Kharkiv National Medical University, led by Professor Dmytro Butov MD, PhD, ScD. Study enrollment included 100 healthy participants, 139 Tuberculosis positives, and 140 participants with respiratory symptoms that were ultimately found to be positives for COVID-19, Influenza, Pneumonia, Acute Respiratory Infections, Bronchitis, Emphysema, and Sinusitis. The study was a prospective, multi-site field validation that compared the AudibleHealth AI Clinical Decision Support Software results to Xpert MTB/RIF Ultra PCR testing for the diagnosis of Tuberculosis.
Other respiratory illness positives in the study were confirmed using various diagnostic modalities including PCR, Chest X-Ray, antigen, and/or blood tests. “Our technology has been in development and rigorous testing for over five years and has performed beautifully in all five of our clinical studies. This particular study was our most ambitious yet, and was designed to be an especially challenging test of the overall accuracy and disease distinguishability of our solution – that’s another reason why we’re so thrilled with these results,” said Kitty Kolding, Co-Founder and CEO at RAIsonance.
Key validation study challenges included:
- The study took place in a challenging environment in Ukraine, in the midst of the war that started in 2022 when Russia invaded the sovereign nation.
- The study took place amongst a population with many other acute respiratory illnesses
- The population exhibited an unusually high incidence of cigarette smoking across categories, but was most pronounced in the TB positives; over 78% of the TB positive participants were self-reported, active cigarette smokers.
- 140 of the enrolled patients were diagnosed with acute respiratory illnesses other than Tuberculosis
- Of the 140 participants who were diagnosed with non-TB respiratory infections, 77 (or 55%) were diagnosed with co-infections, meaning that these participants tested positive for two or more respiratory illnesses
- Even with these hurdles, the AudibleHealth AI solutions met or exceeded the identified performance targets.
“To achieve the performance targets we sought in this study, our solution needed to correctly detect TB positives with at least 90% accuracy, but it also had to correctly detect that these many other positives and healthy patients were TB negatives. Similarly, our COVID-19 and Pneumonia solutions had to correctly detect the positives they encountered with at least 90% accuracy, while correctly assessing the other positives as COVID-19 and Pneumonia negatives respectively. Additionally, more than 20% of the population had at least 2 co-infections. This represented a complex and difficult combination of conditions, but our solution performed beautifully,” said Karl Kelley, Chief Medical Officer at RAIsonance.
“Our unique area of expertise is using proprietary, highly advanced, artificial intelligence and machine learning to analyze and interpret the incredibly rich set of data available from cough sounds. To use the solution, an individual coughs into a customized web app on virtually any Android or Apple smart phone for 10 seconds. That recording is then pushed to the company’s cloud-based production environment in Microsoft Azure, with results returned to the clinician or community health worker in approximately 30 seconds, delivering a fast, simple, effective respiratory screening. We’ve also created a very scalable infrastructure enabling us to process over 20,000 screenings per second, so that we can accommodate virtually any volume spikes we encounter anywhere in the world,” explained Mark Fogarty, Chief Technology Officer at RAIsonance.
“Cough sounds – not to be confused with vocal biomarkers – carry enormous amounts of data that RAIsonance has been studying for years. For each cough collected, we analyze more than 5000 distinct data points. And because every respiratory illness or condition creates different impacts in the respiratory system such as inflammation, airway constriction, changes in mucus quantity and characteristics, etc., our data scientists have discovered an efficient, reliable way to train our models to detect these harmonic patterns in the data that correlate to specific respiratory diseases. That’s how we’re able to not just detect a specific disease but also differentiate between conditions like TB and COVID and Pneumonia,” said Kris Hopkins, Chief Product Officer at RAIsonance.
The AudibleHealth AI solution is designed to give providers and practitioners a fast, portable, inexpensive, non-invasive way to assess their patients during one-to-one interactions as well as to identify likely positive cases at scale. It requires no equipment other than a smartphone, no supplies, generates no waste, requires no shipping or warehousing, and works via SMS text message or WhatsApp all over the world. Because the solution works within a mobile browser, no app downloads or software maintenance is required. As a screener, AudibleHealth AI gives providers fast, critical insights to inform care decisions, so they can proceed with appropriate confirmatory testing as needed.
The company’s AudibleHealth AI platform is designed to house more than 20 different screening solutions. At present the company offers solutions for identifying Tuberculosis, COVID-19, Pneumonia and, for the insurance industry, whether or not a person smokes cigarettes. Solutions for Measles, Influenza, and RSV are in development, and there are another 15 additional solutions on the product roadmap including Pertussis, Emphysema, Asthma, Lung Cancer, Congestive Heart Failure, and many more. Practitioners can choose to have their patient screened for one or many different conditions to provide them with as much data as possible as they assess their patients. The AudibleHealth AI screening solutions for TB, COVID-19 and Pneumonia are now available in seven countries with another 10 markets on the roadmap for 2025.
“To bring our solution to market, we’ve activated distributor partnerships with companies all over the world in telehealth, pharmacies, colleges, digital health platforms, and of course clinics, physician practices, and hospitals. We provide all our customers with our web-based solution management system called TestHub – free of charge – to ensure that they can easily send out screening requests and monitor results from anywhere. It’s exciting to work with so many talented, knowledgeable partners in so many markets to introduce this solution to patients they serve,” said Kraig Monteferrante, SVP of Global Partnerships for RAIsonance.
Cough sequence on phone
Multi-test Resuls
About RAIsonance, Inc.
RAIsonance is a family of companies headquartered in the Denver Tech Center, in Colorado, dedicated to Artificial Intelligence/Machine Learning-based healthcare solutions. RAIsonance companies operate a suite of respiratory wellness products including MyAdvocate for individuals with chronic respiratory conditions, and FCV Sentinel for groups, teams, military, and enterprise applications. It also operates a disease screening platform called AudibleHealth AI designed to screen for specific respiratory illnesses including COVID-19, Pneumonia, Tuberculosis. To learn more about RAIsonance Inc., please visit RAIsonance.ai. or FCVWellness.com.
Media Contact
Anne E. Humpich
Email: Press@RAIsonance.ai
RAIsonance is a family of companies headquartered in the Denver Tech Center, in Colorado, dedicated to Artificial Intelligence/Machine Learning-based healthcare solutions. RAIsonance companies operate a suite of respiratory wellness products including MyAdvocate for individuals with chronic respiratory conditions, and FCV Sentinel for groups, teams, military, and enterprise applications. It also operates a disease screening platform called AudibleHealth AI designed to screen for specific respiratory illnesses including COVID-19, Pneumonia, Tuberculosis. To learn more about RAIsonance Inc., please visit RAIsonance.ai. or FCVWellness.com.
Media Contact
Anne E. Humpich
Email: Press@RAIsonance.ai