Smart Medical Equipment
Heroic-Faith Medical Science Co., Ltd.
heroic-faith.com
Highlights
- Brought the auscultation field forward 200 years with the world’s first AI-powered continuous lung sound monitor
- Protected hundreds of healthcare providers in China and Taiwan during the COVID-19 pandemic by providing remote respiratory monitoring
- Saves lives with faster detection of breathing irregularities than traditional methods

Founded by experts in medicine and electronics, Heroic-Faith Medical Science Co. Ltd. aims to save lives through technological innovation. The firm’s inaugural respiratory auscultation monitor proved especially invaluable in the COVID-19 pandemic, helping physicians stay safe while continuously monitoring patient lung functionality from a distance.
Auscultation limitations
Ever since its invention in 1816, the stethoscope has been a staple of the medical profession. The customary tube hanging around a physician’s neck has seen few iterations over its 200-year history, maintaining its reputation as a symbol of medicine and an essential component to any medical toolkit.
However, the stethoscope is not without its drawbacks. Auscultation, or listening to the sounds of the body, provides key information about a patient’s condition hidden to the eye, but a doctor certainly cannot perform surgery or stand throughout the night with a stethoscope pressed to a patient’s chest. This can be a particular obstacle in intensive care units (ICUs), where certain types of lung diseases are notoriously difficult to diagnose and especially fatal.

While a patient is undergoing surgery without intubation, their vital signs are usually monitored through electrocardiography (ECG), blood pressure and peripheral pulse oximeter. If a patient suddenly stops breathing, it takes about a minute for it to be reflected in their blood oxygen level — a fatally long time for the very young and very old.
To track respiration itself, the US Food and Drug Administration requires the use of capnography, or the monitoring of carbon dioxide in respiratory gases, as objective evidence of lung functionality during or after surgery or intensive care, but it does not always accurately depict breathing rate.
These limitations were put on full display during the COVID-19 pandemic. While facing a virus that mostly presents in the lungs, physicians and researchers were stymied by a lack of consistent physical access to isolated patients, relying mostly on electronic stethoscopes that allow physicians to remain at a safe distance, but are still not continuous or independently functioning.

High-tech update for a dormant field
Recognizing the need for a 21st century update to one of the oldest diagnostic tools in medicine, Frank Wu and Dr. Hsu Fu-shun in 2018 received angel investment to fulfill their vision for a continuous lung auscultation monitor backed by artificial intelligence (AI).
“With an AI continuous listening system, medical personnel can transcend time and space to understand a patient’s condition,” Hsu said.
Heroic-Faith gathered input from physicians, who recommended a noninvasive, lightweight design that can be affixed externally without disrupting other equipment in an ICU or operating room.
Once the sensors are adhered to a patient’s chest, the device can immediately begin recording respiratory sounds, which are visualized on a spectrogram, or can be listened to directly on a computer or phone. This type of monitoring is particularly valuable in non-intubated anaesthesia and ICUs, where something as simple as listening to breathing can save precious seconds in reaction time, and can save physicians from manually collecting and aggregating auscultation information.
To train the device, Heroic-Faith invited physicians with at least five years of clinical respiratory experience to annotate the wealth of data in this world-first breathing sound database. These notations are then fed into an AI, which through deep learning continuously perfects an algorithm that can detect differences between normal and adventitious breathing sounds, such as potentially serious stridor and wheeze sounds, as well as the characteristic crackle that points to pulmonary edema. In the event of apnea or respiratory deterioration, medical personnel are immediately notified and can provide timely assistance.
At the onset of the COVID-19 pandemic, Heroic-Faith found itself to be in the right place with the right product at the right time. Chinese and Taiwanese physicians communicate regularly on new developments and even complete virtual rounds, meaning that Hsu already had regular communication with doctors in Wuhan by February 2020, and by March had respiratory monitors in the hands of doctors across the city.
“Without affecting the quality of medical care, remote respiratory monitoring can reduce the risk of cross-infection between medical personnel and patients,” Hsu said, giving doctors a piece of assurance in trying times.

A new direction in clinical monitoring
In Taiwan, Heroic-Faith’s monitors have already received institutional review board approval and are being used in hospitals across the country, collecting more data through its more than 400,000 uses daily to perfect its unpinning algorithm.
The firm has also started to expand to the US, where it maintains a close relationship with the Stanford University Sleep Surgery Clinic, expects FDA approval by 2022 and was even showcased in the 2020 Consumer Electronics Show in Las Vegas.
“In the future, I hope that respiratory measurements — like blood pressure — can become a primary indicator of physiological functionality,” Hsu said. “I also believe that the creation of a remote lung monitoring system will greatly improve Taiwan’s public healthcare system and serve as a global model for remote auscultation and respiratory care.”
SELLING POINTS
- World’s first continuous AI respiratory auscultation monitor allows physicians to monitor breathing of multiple patients simultaneously
- Remote monitoring protects medical personnel from exposure
- Deep-learning AI trained on world’s first and largest breathing sound database automatically differentiates between respiratory functions
- Lung sounds provide faster and more accurate data than other types of monitoring in ICUs and surgery rooms, such as peripheral pulse oximeter and capnography


