What question did this study set out to answer?

This research aims to develop a deep learning model that predicts pathological complete response using only pre-treatment clinical and MRI features.

Abstract PS3-06-27: Deep Learning to Predict Pathological Complete Response in Patients Receiving NAC using Pre-treatment Clinical and Imaging Features

View Full Paper

Ask AI

Bookmark

Authors

Novel ML architecture predicts pCR in breast cancer from pre-treatment data, highlighting clinical implications.

This research aims to develop a deep learning model that predicts pathological complete response using only pre-treatment clinical and MRI features.
Developed a convolutional neural network trained on both pre- and post-treatment MRIs.
Created a second CNN to predict post-treatment radiomic features from pre-treatment MRIs.
Substituted predicted features into the original CNN to avoid the need for post-treatment imaging.
Achieved an AUC of 0.62 for the proposed method using only pre-treatment MRIs.
Improved accuracy by 3% to 0.73 when including clinical data.
Maintained 87% of the accuracy of models using both treatment time-points.

Gifford et al. (2026) studied this question.

View Full Paper

Ask AI

Bookmark