Evaluating Transferability for Covid 3D Localization Using CT SARS-CoV-2 segmentation models

TL;DR

This paper evaluates the transferability of deep learning models, specifically U-Net architectures, for segmenting Covid-19 infected regions in CT scans, demonstrating improved accuracy across different datasets.

Contribution

It investigates the effectiveness of transfer learning with pre-trained U-Net models for Covid-19 segmentation in CT images, highlighting its potential when limited training data is available.

Findings

Transfer learning improves segmentation accuracy.

Pre-trained U-Net models generalize well across datasets.

Enhanced detection of Covid-19 infected regions.

Abstract

Recent studies indicate that detecting radiographic patterns on CT scans can yield high sensitivity and specificity for Covid-19 localization. In this paper, we investigate the appropriateness of deep learning models transferability, for semantic segmentation of pneumonia-infected areas in CT images. Transfer learning allows for the fast initialization/reutilization of detection models, given that large volumes of training data are not available. Our work explores the efficacy of using pre-trained U-Net architectures, on a specific CT data set, for identifying Covid-19 side-effects over images from different datasets. Experimental results indicate improvement in the segmentation accuracy of identifying Covid-19 infected regions.