Technology

There are many digital stethoscopes that enable recording of lung sounds.
Medsensio provides a platform, enabling lung sound analysis. How?

 
Screenshot_2018-08-22 Lung sounds for auscultation training - Medsensio.png

Deep learning

Our innovation brings artificial intelligence to digitalized lung auscultation. Lung sound interpretation is a subjective and manual process, research has shown that medical personnel classify lung sounds differently.

Medsensio uses state-of-the-art deep learning algorithms, trained and validated on the world’s biggest population-representative lung sound dataset in collaboration with the University of Tromsø. Our algorithms achieve near human-level performance.

Data

Medsensio has a unique position with access to the worlds biggest annotated dataset, recorded in a natural clinical environment, which means that it contained environmental noise such as unfiltered heart and bowel sounds and talking. This plays a crucial role in developing the algorithms, which would be useful in a regular clinical setting.  

db_icon_0fkghg.png

building-strategy.png

Expertise

Medsensio has a team with expertise in deep learning, software development, health technologies and has a strategic partnership with UiT and the Tromsø Study. Professor Hasse Melbye at the Department of Community Medicine (ISM) leads the research group in lung sounds. Professor Lars Ailo Bongo leads the reaserach group at the Department of Computer Science, with expertise in Deep Learning and Bioinformatics.

computer-connection-data-API-1181675.jpg

Build with our API!

Want to integrate our lung sound analysis in your products?
Or utilize our lung sound analysis for your research?

Get in touch with us and we can tell you more about our technology!