Improved Inference via Deep Input Transfer
This addresses the challenge of boosting segmentation accuracy for medical imaging applications, offering an incremental but novel approach compared to traditional methods.
The paper tackles the problem of improving image segmentation performance by proposing a new paradigm that modifies the network's input using gradients from a trained model, rather than altering the network itself. On three medical image datasets, the method achieved Dice score improvements of 5.8%, 0.5%, and 4.8%.
Although numerous improvements have been made in the field of image segmentation using convolutional neural networks, the majority of these improvements rely on training with larger datasets, model architecture modifications, novel loss functions, and better optimizers. In this paper, we propose a new segmentation performance boosting paradigm that relies on optimally modifying the network's input instead of the network itself. In particular, we leverage the gradients of a trained segmentation network with respect to the input to transfer it to a space where the segmentation accuracy improves. We test the proposed method on three publicly available medical image segmentation datasets: the ISIC 2017 Skin Lesion Segmentation dataset, the Shenzhen Chest X-Ray dataset, and the CVC-ColonDB dataset, for which our method achieves improvements of 5.8%, 0.5%, and 4.8% in the average Dice scores, respectively.