Automatic Generation of Synthetic Colonoscopy Videos for Domain Randomization
This addresses the need for high-quality, varied training data to improve generalization in medical imaging for colonoscopy assistance, though it is incremental as it builds on existing domain randomization techniques.
The paper tackled the problem of limited training data for colonoscopy guidance systems by proposing a method to automatically generate synthetic colonoscopy videos with substantial appearance and anatomical variations, enabling the learning of domain-randomized representations that mimic real-world settings.
An increasing number of colonoscopic guidance and assistance systems rely on machine learning algorithms which require a large amount of high-quality training data. In order to ensure high performance, the latter has to resemble a substantial portion of possible configurations. This particularly addresses varying anatomy, mucosa appearance and image sensor characteristics which are likely deteriorated by motion blur and inadequate illumination. The limited amount of readily available training data hampers to account for all of these possible configurations which results in reduced generalization capabilities of machine learning models. We propose an exemplary solution for synthesizing colonoscopy videos with substantial appearance and anatomical variations which enables to learn discriminative domain-randomized representations of the interior colon while mimicking real-world settings.