Transferability limitations for Covid 3D Localization Using SARS-CoV-2 segmentation models in 4D CT images
This addresses the problem of unreliable model transfer for COVID-19 diagnosis in medical imaging, though it appears incremental as it focuses on limitations rather than new solutions.
The paper investigates transferability limitations in deep learning models for segmenting pneumonia-infected areas in COVID-19 CT images, finding that retraining models multiple times on large datasets actually decreases segmentation accuracy.
In this paper, we investigate the transferability limitations when using deep learning models, for semantic segmentation of pneumonia-infected areas in CT images. The proposed approach adopts a 4 channel input; 3 channels based on Hounsfield scale, plus one channel (binary) denoting the lung area. We used 3 different, publicly available, CT datasets. If the lung area mask was not available, a deep learning model generates a proxy image. Experimental results suggesting that transferability should be used carefully, when creating Covid segmentation models; retraining the model more than one times in large sets of data results in a decrease in segmentation accuracy.