Project Description

Developed while a student in Dr. May Wang’s Biomedical and Health Informatics graduate course, this clinical decision support system is a novel approach to assist pathologists diagnose renal cell carcinoma (RCC) subtypes. Differentially expressed genes (DEGs) from whole genome RNA-Seq gene expression profiling and matched histopathological images from the patient (courtesy of the TCGA) are used to create an integrated feature vector for patient stratification and classification. Using machine learning algorithms applied to this integrated, multi-dimensional feature vector, we are able to accurately diagnose RCC subtypes and patient survival time 90% of the time.


Clinical Diagnosis

Fast, reliable predication of clinical cancer subtypes with supervised clustering and training data

Easy Interface

An intuitive front-end design to easily integrate into the clinical diagnosis workflow with powerful back-end machine learning algorithms

True Power

A robust program architecture designed for multimodal, highly dimensional data

Solid Code

Built using Python for cross platform capabilities and intensive data visualization