Medical Image Registration on Tumor Growth with Time Series

Medical Image Registration on Tumor Growth with Time Series

In medical image word, doctors usually want to compare images taken from different time periods of the same patient. However, due to the misalignment of patient’s posture, comparison between the MRI scans is tedious. We applied 3 different methods to align the images from different scans and compared the performance of these methods.

tumer

Automatic brain aneurysm detection

This project aims to detect cerebral aneurysms with the help of machine learning method. Traditionally doctors usually have to assess each individual medical image (MRI/MRA) to decide whether there is an aneurysm in the patient’s brain. Because there can be hundreds of images, the work is time consuming and tiring. With the help of automatic detection system, the computer can point out the possible locations of the aneurysm and the doctor and read the specific image and location without going through all the images.

Researcher: Kuocheng Wang