Utilizing HPC

The bladder vision project is using a High Performance Computing Cluster known as SHAMU which provides a 100 times increase in computation speed.

Computer Vision

Using an Olympus Cystoscope we are able to capture video footage inside the bladder for post-in-vivo tests.

Reconstruction and Analysis

Utilizing Algorithms such as scale invariant feature transform (SIFT) and structure from motion (SFM), we are better able to characterize abnormalities. 

Bladder Vision

Using computer vision to detect carcinogenic anomalies in bladders

A software pipeline solution to help provide higher accuracy for the targeted bladder cancer treatment external beam radiotherapy (EBRT). This software pipeline can create 3D reconstructions of a bladder or inner surface of an organ using 2D images collected from a flexible cystoscope. Through the use of computer vision image preprocessing techniques and machine learning algorithms scale invariant feature transform (SIFT) and structure from motion (SFM) to identify key points and spatially stich them for reconstruction.