Deep Learning Models for Calculation of Cardiothoracic Ratio from Chest Radiographs for Assisted Diagnosis of Cardiomegaly
This work addresses the need for objective, automated diagnosis of cardiomegaly in medical imaging to reduce human subjectivity, though it is incremental as it applies existing deep learning methods to a specific medical task.
The authors tackled the problem of automating cardiothoracic ratio calculation from chest radiographs to assist in diagnosing cardiomegaly, achieving high sensitivity (up to 0.96) and specificity (up to 0.97) with low mean absolute error (as low as 0.018) on test datasets.
We propose an automated method based on deep learning to compute the cardiothoracic ratio and detect the presence of cardiomegaly from chest radiographs. We develop two separate models to demarcate the heart and chest regions in an X-ray image using bounding boxes and use their outputs to calculate the cardiothoracic ratio. We obtain a sensitivity of 0.96 at a specificity of 0.81 with a mean absolute error of 0.0209 on a held-out test dataset and a sensitivity of 0.84 at a specificity of 0.97 with a mean absolute error of 0.018 on an independent dataset from a different hospital. We also compare three different segmentation model architectures for the proposed method and observe that Attention U-Net yields better results than SE-Resnext U-Net and EfficientNet U-Net. By providing a numeric measurement of the cardiothoracic ratio, we hope to mitigate human subjectivity arising out of visual assessment in the detection of cardiomegaly.