Improving diagnostic accuracy using multiparameter patient monitoring based on data fusion in the cloud

Accurate clinical decision making in medical monitoring relies on the strategical fusion of multiparameter physiological signals and usually demands a wide variety of complex machine learning approaches and a large set of knowledge data-base. However, those requirements impose great challenges on computing and storage capabilities, which make it impossible to execute on a single portable computing platform. Leveraging emerging cloud computing technologies, we propose to strategically manage the workloads on the mobile medical monitoring device and migrate the highly intricate multipara-meter data fusion and training procedure to the cloud. The mobile device transmits all sensing data acquired from wearable body sensors to the cloud, which now provides a large pool of easily accessible dataset for the training procedures. The well-trained configurations will be sent back to the mobile device and update its existing machine learning based implementations.