Adversarial Networks for Prostate Cancer Detection
This work addresses challenges in prostate cancer detection from MRI for medical imaging applications, representing an incremental improvement over existing methods.
The paper tackled the problem of data-hungry deep neural networks and ambiguous medical imaging annotations by using adversarial training for segmentation networks, achieving superior detection sensitivity and dice-score for aggressive prostate cancer on MRI images from 152 patients, with gains especially notable in small dataset limits.
The large number of trainable parameters of deep neural networks renders them inherently data hungry. This characteristic heavily challenges the medical imaging community and to make things even worse, many imaging modalities are ambiguous in nature leading to rater-dependant annotations that current loss formulations fail to capture. We propose employing adversarial training for segmentation networks in order to alleviate aforementioned problems. We learn to segment aggressive prostate cancer utilizing challenging MRI images of 152 patients and show that the proposed scheme is superior over the de facto standard in terms of the detection sensitivity and the dice-score for aggressive prostate cancer. The achieved relative gains are shown to be particularly pronounced in the small dataset limit.