Data Compression in Gesture-Based Human Machine Interface for Continuous Digital Health

Abstract

Continuing our previous work on gesture tracking in the environment of arthroplasty, and
because of the complexity and potential complications, there is a need for a new approach for
any application of hand gestures in a continuous digital health ecosystem, that could be easily
monitored and executed. With this approach, we tend to capture hand gestures in some
predefined space. The first step is to find a mathematical model that will fit some well-known
kernel shapes. If a motion is not derived from the kernel shapes, the full search for motion
vectors with pairing and computation of vectors could be performance exhaustive, so we
propose the combination of motion based on motion prediction on kernel shapes. To predict
motion parameters from specific features in specific medical data-acquiring models, we must
know the visualization requirements of the continuous digital health system that will synthesize
the motion. It is also necessary to relate these parameters to the actions that will enable data
compression in storing all the motions in hand gestures of medical specialists controlling and
manipulating medical data. This motion prediction is focusing on a specific object visualization
framework that we use as an input for a gesture-based human-machine interface applicable to
any continuous digital health system, which we described in this paper.