Using supercomputers to diagnose brain tumours

A research team at the University of Texas, led by Professor George Biros have been able to accurately characterise gliomas using a super computer

Primary brain tumours include a wide range of tumours and are dependent on the type of cell, and the aggressiveness and stage of the tumour.

The most common primary brain tumours are called gliomas. Accurate diagnoses of these tumours is vital to create a suitable treatment plan.

Currently, specialised radiologists are tasked with accurately characterising gliomas which then results in an appropriate diagnosis. However, recent research has suggested that in the near future, super computers could play a vital role in the diagnostic process.

The research team led by Professor Biros, at the University of Texas at Austin, have created computer algorithms - which are defined set of instructions for the computer to perform tasks - which can scan images to detect tissue patterns that define gliomas.

"Our goal is to take an image and delineate it automatically and identify different types of abnormal tissue - edema, enhancing tumour (areas with very aggressive tumours), and necrotic tissue. It's similar to taking a picture of one's family and doing facial recognition to identify each member, but here you do tissue recognition, and all this has to be done automatically,” says Professor Biros.

The research team demonstrated this technology at the 20th International Conference on Medical Image Computing and Computer Assisted Intervention. At this conference they were given data from 140 patients with brain tumours and they characterised the tumours with 90% accuracy, which is comparable to human radiologists.

While this technology cannot replace radiologists, it could play a supportive role to characterise brain tumours, allowing for an early and accurate diagnosis.