Combining Image- and Geometric-based Deep Learning for Shape Regression: A Comparison to Pixel-level Methods for Segmentation in Chest X-Ray
This work addresses the challenge of generating anatomically plausible predictions in medical imaging, particularly for corrupted data, with an incremental improvement in robustness for chest X-ray segmentation.
The authors tackled the problem of anatomical segmentation in chest X-rays by proposing a hybrid method combining a CNN backbone with a geometric neural network for shape regression, achieving comparable segmentation quality to state-of-the-art methods (e.g., 4±1.9 mm error vs. 3.9±2.9 mm) while being more stable against image distortions, outperforming them at 30% corruption levels.
When solving a segmentation task, shaped-base methods can be beneficial compared to pixelwise classification due to geometric understanding of the target object as shape, preventing the generation of anatomical implausible predictions in particular for corrupted data. In this work, we propose a novel hybrid method that combines a lightweight CNN backbone with a geometric neural network (Point Transformer) for shape regression. Using the same CNN encoder, the Point Transformer reaches segmentation quality on per with current state-of-the-art convolutional decoders ($4\pm1.9$ vs $3.9\pm2.9$ error in mm and $85\pm13$ vs $88\pm10$ Dice), but crucially, is more stable w.r.t image distortion, starting to outperform them at a corruption level of 30%. Furthermore, we include the nnU-Net as an upper baseline, which has $3.7\times$ more trainable parameters than our proposed method.