Assessing generalisability of deep learning-based polyp detection and segmentation methods through a computer vision challenge.

Polyps are well-known cancer precursors identified by colonoscopy. However, variability in their size, appearance, and location makes the detection of polyps challenging. Moreover, colonoscopy surveillance and removal of polyps are highly operator-dependent procedures and occur in a highly complex organ topology. There exists a high missed detection rate and incomplete removal of colonic polyps. To assist in clinical procedures and reduce missed rates, automated methods for detecting and segmenting polyps using machine learning have been achieved in past years. However, the major drawback in most of these methods is their ability to generalise to out-of-sample unseen datasets from different centres, populations, modalities, and acquisition systems. To test this hypothesis rigorously, we, together with expert gastroenterologists, curated a multi-centre and multi-population dataset acquired from six different colonoscopy systems and challenged the computational expert teams to develop robust automated detection and segmentation methods in a crowd-sourcing Endoscopic computer vision challenge. This work put forward rigorous generalisability tests and assesses the usability of devised deep learning methods in dynamic and actual clinical colonoscopy procedures. We analyse the results of four top performing teams for the detection task and five top performing teams for the segmentation task. Our analyses demonstrate that the top-ranking teams concentrated mainly on accuracy over the real-time performance required for clinical applicability. We further dissect the devised methods and provide an experiment-based hypothesis that reveals the need for improved generalisability to tackle diversity present in multi-centre datasets and routine clinical procedures.

Polyp segmentation with consistency training and continuous update of pseudo-label.

Polyp segmentation has accomplished massive triumph over the years in the field of supervised learning. However, obtaining a vast number of labeled datasets is commonly challenging in the medical domain. To solve this problem, we employ semi-supervised methods and suitably take advantage of unlabeled data to improve the performance of polyp image segmentation. First, we propose an encoder-decoder-based method well suited for the polyp with varying shape, size, and scales. Second, we utilize the teacher-student concept of training the model, where the teacher model is the student model's exponential average. Third, to leverage the unlabeled dataset, we enforce a consistency technique and force the teacher model to generate a similar output on the different perturbed versions of the given input. Finally, we propose a method that upgrades the traditional pseudo-label method by learning the model with continuous update of pseudo-label. We show the efficacy of our proposed method on different polyp datasets, and hence attaining better results in semi-supervised settings. Extensive experiments demonstrate that our proposed method can propagate the unlabeled dataset's essential information to improve performance.