Enhanced detection of the presence and severity of COVID-19 from CT scans using lung segmentation

Robert Turnbull

Mar 2023

摘要

Improving automated analysis of medical imaging will provide clinicians moreoptions in providing care for patients. The 2023 AI-enabled Medical ImageAnalysis Workshop and Covid-19 Diagnosis Competition (AI-MIA-COV19D) providesan opportunity to test and refine machine learning methods for detecting thepresence and severity of COVID-19 in patients from CT scans. This paperpresents version 2 of Cov3d, a deep learning model submitted in the 2022competition. The model has been improved through a preprocessing step whichsegments the lungs in the CT scan and crops the input to this region. Itresults in a validation macro F1 score for predicting the presence of COVID-19in the CT scans at 92.2% which is significantly above the baseline of 74%. Itgives a macro F1 score for predicting the severity of COVID-19 on thevalidation set for task 2 as 67% which is above the baseline of 38%.

展开全部

图表提取

论文十问

参考文献

被引用

提问与回答

暂无人提供速读十问回答

被引用

笔记

问答

领域