Projects

Joint Solution for PET Image Segmentation and Denoising

In this project I implemented a research paper on Joint Solution for PET Image Segmentation and Denoising on ITK. First, Generalized Anscombe Transformation (GAT) is used for Guassianizing the multiplicative noise in the PET images. Then segmentation is performed on the Generalised Anscombe Transformed PET images by affinity propogation based clustering algorithm. At the end of the segmentation process, Inverse GAT is performed on the segmented PET images to achieve denoising.

Fitness Rubber band Exercise Monitor

Fitness rubber band or the resistive band is widely used in comprehensive full-body workouts which are used to exercise most of the major muscle groups in the body. Monitoring these parameters such as force and jerk applied on the rubber band during exercise can be useful in evaluating whether the patient has performed the prescribed exercises correctly. To get these measurements, we attached the rubber band to a small metal frame where two strain gauges were attached to its surface. We designed the necessary electronic circuitry to get strain measurements from the two strain gauges and communicated these measurements to the computer via WiFi. We determined the relationship between the strain and force by measuring the strain under different force conditions. The force and the jerk values were plotted real-time using Matlab.