VITAL-ECG: a de-bias algorithm embedded in a gender-immune device

Artificial intelligence, sensors technology and sensors networks influence people behavior in everyday life. The diffusion of mobile devices, based on Internet of Things (IoT) paradigms, has created specific solutions for applications, in which physical objects are connected to Internet system. Wearable IoT (WIoT) represents a new IoT area, concerning detection, processing and communication capabilities in the field of healthcare. Vital-ECG is a smart device, related to health monitoring, which complies with gender equality. The wearable device takes the form of a smartwatch, which monitors heart activity and the most important vital parameters: blood oxygen saturation, skin temperature and fatigue level. Electrocardiogram and plethysmogram signals are acquired from Vital-ECG, which is able to track the blood pressure values, through a deep learning implementation. The neural algorithm has been implemented avoiding the “Gender Bias”. The gender balance in machine learning, especially in biomedical application, is a crucial point to prevent algorithms from making a distorted prediction, disadvantaging women.