A healthcare research project at Amrita Center for Wireless Networks & Applications (AmritaWNA) has been focusing on developing affordable and indigenous sensors, healthcare data analytics techniques, and communication architectures for use in sparsely connected regions in India. One of the early success stories has been the Amrita Spandanam – wearable ECG monitoring device. As a continuation of this project, the Center has developed a communication architecture for healthcare IoT devices, called “Health-Aware Networks”, which is specifically designed for transmission and processing of vitals data over unreliable mobile networks. One of the research papers, “Multi-layer Architectures for Remote Health Monitoring“, describing this work, was recently presented at the prestigious IEEE Healthcom, held at Munich in September 2016.

The research team also came up with an innovative technique called “dynamic log chunking” to reduce the data upload and download from the cloud by up to 20%. This work was presented at the ACM MobiHoc as a poster, “Context Aware Dynamic Log Chunking for Mobile Healthcare Applications“, in August 2016. The resulting IoT data management framework is called “H-Plane”, H-Plane: Intelligent Data Management for Mobile Healthcare Applications, which stands for Healthcare Plane- a log based IoT data storage, processing, and abstraction layer, that helps the application developers be agnostic to the underlying edge and cloud data management techniques.

More recently, the team at Amrita WNA has been working on developing a healthcare data analytics platform and algorithms that can effectively detect severe or fatal heart conditions using a novel technique called RASPRO (Rapid Active Summarization for effective PROgnosis). Using this technique, the body sensor data is quantized and converted in a smartphone into doctor-readable severity levels. This will greatly help in reducing the data transmission from the smartphones, as well as help the practitioners in hospitals to remotely monitor patients much more efficiently.

The RASPRO technique aims to provide the doctors a severity-ordered patient list, so that the most needed patient is attended to first. The initial work was recently presented at ACM UbiComp workshop, IoP Health, held in Heidelberg, in September 2016. Further formalization of this technique has resulted in some early success too, which was presented at a conference hosted at the National Institutes of Health, Bethesda, US – IEEE Wireless Health (Paper: “RASPRO: Rapid Summarization for Effective Prognosis in Wireless Remote Health Monitoring“).

These research outcomes have been due to joint multi-disciplinary efforts between researchers at Amrita WNA, Amrita Center for Cybersecurity and Networks, and doctors at Amrita Institute of Medical Sciences. Currently, the team is also working on expanding these techniques to other medical domains, especially into sleep studies involving sleep apnea patients, as well as critical patients having acute hypotensive episodes.