PrepNet: A Convolutional Auto-Encoder to Homogenize CT Scans for Cross-Dataset Medical Image Analysis

TL;DR

PrepNet is a convolutional auto-encoder designed to homogenize CT scans across different datasets, improving the generalization of COVID-19 diagnosis models by reducing variability caused by different imaging technologies.

Contribution

This paper introduces PrepNet, a novel auto-encoder architecture that enhances cross-dataset generalization in medical image analysis by mitigating scan variability.

Findings

Improves cross-dataset generalization by up to 11.84 percentage points.

Capable of extracting discriminative features for better diagnosis.

Maintains competitive within-dataset performance despite variability reduction.

Abstract

With the spread of COVID-19 over the world, the need arose for fast and precise automatic triage mechanisms to decelerate the spread of the disease by reducing human efforts e.g. for image-based diagnosis. Although the literature has shown promising efforts in this direction, reported results do not consider the variability of CT scans acquired under varying circumstances, thus rendering resulting models unfit for use on data acquired using e.g. different scanner technologies. While COVID-19 diagnosis can now be done efficiently using PCR tests, this use case exemplifies the need for a methodology to overcome data variability issues in order to make medical image analysis models more widely applicable. In this paper, we explicitly address the variability issue using the example of COVID-19 diagnosis and propose a novel generative approach that aims at erasing the differences induced by e.g. the imaging technology while simultaneously introducing minimal changes to the CT scans through leveraging the idea of deep auto-encoders. The proposed prepossessing architecture (PrepNet) (i) is jointly trained on multiple CT scan datasets and (ii) is capable of extracting improved discriminative features for improved diagnosis. Experimental results on three public datasets (SARS-COVID-2, UCSD COVID-CT, MosMed) show that our model improves cross-dataset generalization by up to $11.84$ percentage points despite a minor drop in within dataset performance.