XVision, First Place Winner of the 2018 Microsoft Big Idea Challenge

XVision, a project started by four students from the Politehnica University of Timisoara, is the First Place Winner of the 2018 Big Idea Challenge, hosted by Microsoft as part of the Imagine Cup annual global competition. This year, over 100 teams from all over the world took the challenge. Each team had to pitch their idea as well as a project plan showing how their projects can have an impact over the world.

The team behind XVision is composed of Cristian Avramescu, Bogdan Bercean, Stefan Iarca and Andrei Tenescu, students at the Computers study program of the Faculty of Automation and Computing. The team is coordinated by Dr. Eng. Sebastian Fuicu.

XVision is a system designed to automatically detect anomalies and diseases encountered anywhere in the human body. With radiologist-level accuracy, XVision analyzes common medical X-ray images with the help of Azure technologies such as Azure Machine Learning Studio. The application lets the medical team using the X-ray scanner to upload the generated image and provides a diagnostic, along with a heat map of the original scan in which the anomalous areas are highlighted. XVision uses a Convolutional Neural Network to classify the body part that the scan presents, then two DenseNet Neural Networks for the anomaly detection and its classification.

This product can provide a much needed solution for people in areas that lack access to radiology diagnostics while also acting as an assistant tool for the medical experts.