By Shiza Charania (shizacharania.com)

Overview:

<aside> 💡 This research plays a part of the probabilistic brain tumour mapping project I worked on

</aside>

Two compelling research avenues have recently emerged in neurooncology:

1. probabilistic maps that quantitatively depict where brain tumors tend to arise and
2. the field of omics that capture the biological characteristics that comprise a sample of an organism, such as its genomic features

This model is part of a probabilistic brain tumor mapping project, which works to find correlations between brain tumor grades (severities) and their location in the brain. Firstly, you have brain tumour segmentations (already manually segmented) that are overlaid on a heat map to understand the frequency of tumors arising in a certain region (as shown below)

In addition to this, by finding the grade of all the tumors that are segmented, you would be able to correlate the frequency of different tumor grades arising in the brain. For instance, a correlation could be that there are a lot of grade 4 (very severe) tumors in the temporal lobe.

The project as will give more insight to clinicians in neuroscience about the origination of tumors/where they tend to arise, leading to better techniques for targeting and treatment.

Code:

https://github.com/shizacharania/Brain-Tumour-Grading

My Novel Architecture and What Makes it Novel: