Medsensio - Cloud services for real-time detection and classification of abnormal lung sounds

More than 300 million worldwide suffer from chronical respiratory diseases such as asthma and COPD, and these numbers are continuously growing. However, the practices of discovering symptoms called auscultation still rely on stethoscopes, which have not improved significantly since last century. Auscultation can be difficult and inexperienced doctors may interpret symptoms differently. In addition, people want to have a better overview of their health without requiring a doctor, which might be difficult and costly. We in Medsensio want to help people with lungs issues in several ways: First, we want to help medicine students, nursing students and young inexperienced practitioners to train their skills and get a deep experience in auscultation, so that later patients can get better medical help. Second, we want to help doctors and nurses, by integrating our solution in smart stethoscopes, which will automatically discover and identify abnormal lung sounds. Last, we want to give patients the ability to monitor their health at home, without the need of a doctor. We want to develop a device that can discover early signs of a worsening of the disease, so that actions to reduce the severity of the disease can be taken. This in turn can reduce the amount of hospital visits and sick leave. All of this we can achieve by building training tool (1) and software integration with smart stethoscopes (2) and medical devices (3), based on our core technology, which can automatically analyze lung sounds. Our technology is a deep learning based algorithm, which has been trained on 18 000 sound files. With Forskningsrådet financing our startup Medsensio will test our existing prototype, collect feedback from potential customers and users, develop the end-product (1) and foundation for product (2), and establish a network of partners, who will help us to bring our products to the market.