C2G-Net: Exploiting Morphological Properties for Image Classification

TL;DR

C2G-Net is a novel image classification pipeline that leverages morphological properties of images with many similar objects, reducing training time and enhancing interpretability without sacrificing accuracy.

Contribution

The paper introduces C2G-Net, combining a segmentation-based image compression with a lightweight CNN for improved efficiency and interpretability in biological image classification.

Findings

Achieved similar accuracy to conventional CNNs on colon cancer images.

Reduced training time by 85%.

Produced more interpretable models.

Abstract

In this paper we propose C2G-Net, a pipeline for image classification that exploits the morphological properties of images containing a large number of similar objects like biological cells. C2G-Net consists of two components: (1) Cell2Grid, an image compression algorithm that identifies objects using segmentation and arranges them on a grid, and (2) DeepLNiNo, a CNN architecture with less than 10,000 trainable parameters aimed at facilitating model interpretability. To test the performance of C2G-Net we used multiplex immunohistochemistry images for predicting relapse risk in colon cancer. Compared to conventional CNN architectures trained on raw images, C2G-Net achieved similar prediction accuracy while training time was reduced by 85% and its model was is easier to interpret.