CT-xCOV: a CT-scan based Explainable Framework for COVid-19 diagnosis

TL;DR

This paper introduces CT-xCOV, an explainable deep learning framework for COVID-19 diagnosis from CT scans, combining lung segmentation, classification, and explanation techniques with validated performance metrics.

Contribution

It presents an end-to-end explainable framework integrating lung segmentation, multiple CNN models, and XAI techniques, with a novel ground-truth-based evaluation method for explanations.

Findings

U-Net achieved 98% Dice coefficient in lung segmentation.

Standard CNN achieved 98.4% accuracy in COVID-19 detection.

Grad-Cam provided the most accurate visual explanations.

Abstract

In this work, CT-xCOV, an explainable framework for COVID-19 diagnosis using Deep Learning (DL) on CT-scans is developed. CT-xCOV adopts an end-to-end approach from lung segmentation to COVID-19 detection and explanations of the detection model's prediction. For lung segmentation, we used the well-known U-Net model. For COVID-19 detection, we compared three different CNN architectures: a standard CNN, ResNet50, and DenseNet121. After the detection, visual and textual explanations are provided. For visual explanations, we applied three different XAI techniques, namely, Grad-Cam, Integrated Gradient (IG), and LIME. Textual explanations are added by computing the percentage of infection by lungs. To assess the performance of the used XAI techniques, we propose a ground-truth-based evaluation method, measuring the similarity between the visualization outputs and the ground-truth infections. The performed experiments show that the applied DL models achieved good results. The U-Net segmentation model achieved a high Dice coefficient (98%). The performance of our proposed classification model (standard CNN) was validated using 5-fold cross-validation (acc of 98.40% and f1-score 98.23%). Lastly, the results of the comparison of XAI techniques show that Grad-Cam gives the best explanations compared to LIME and IG, by achieving a Dice coefficient of 55%, on COVID-19 positive scans, compared to 29% and 24% obtained by IG and LIME respectively. The code and the dataset used in this paper are available in the GitHub repository [1].