Explainable Deep Learning for Medical Image Analysis

Authors

Brennan Rhoadarmer, University of Nebraska-LincolnFollow

Date of this Version

Spring 4-2020

Document Type

Poster

Citation

Poster presentation, UCARE Research Fair, Spring 2020, University of Nebraska-Lincoln.

Comments

Copyright 2020 by the authors.

Abstract

Explainable Deep Learning for Medical Image Analysis is a project focused on improving the ability for deep learning models to explain the reasoning behind their classification in order to improve their viability in the medical field, where explanations of decisions is critical for the care of patients. In order to explore this topic, we work to implement GradCAM, which is a new method of determining the cause classification in models by tracing back through the model layers to the input.