AI-Assisted Diagnosis for Covid-19 CXR Screening: From Data Collection to Clinical Validation

TL;DR

This paper introduces the Co.R.S.A. project, which develops and clinically validates an AI system for Covid-19 diagnosis from chest X-ray images, including dataset release, model development, and real-world radiologist testing.

Contribution

It provides a publicly available dataset, a novel deep learning detection pipeline, and clinical validation demonstrating real-world effectiveness.

Findings

The AI system achieves reliable Covid-19 detection results.

Radiologists improved accuracy and efficiency using the AI tool.

The project offers a comprehensive pipeline from data collection to clinical validation.

Abstract

In this paper, we present the major results from the Covid Radiographic imaging System based on AI (Co.R.S.A.) project, which took place in Italy. This project aims to develop a state-of-the-art AI-based system for diagnosing Covid-19 pneumonia from Chest X-ray (CXR) images. The contributions of this work are manyfold: the release of the public CORDA dataset, a deep learning pipeline for Covid-19 detection, and the clinical validation of the developed solution by expert radiologists. The proposed detection model is based on a two-step approach that, paired with state-of-the-art debiasing, provides reliable results. Most importantly, our investigation includes the actual usage of the diagnosis aid tool by radiologists, allowing us to assess the real benefits in terms of accuracy and time efficiency. Project homepage: https://corsa.di.unito.it/