Virtual crystal violet chromoendoscopy: a digital alternative to dye-based staining.
Аннотация
BACKGROUND AND AIMS: Routine use of crystal violet (CV) chromoendoscopy for pit pattern-based colorectal lesion characterization is limited by safety concerns. We developed and clinically validated an artificial intelligence (AI)-based virtual CV (vCV) chromoendoscopy technique that reproduces CV-equivalent images without dye.
METHODS
Using an unsupervised image-to-image translation framework (UVCGAN v2), we trained a model to transform endoscopic images into vCV images. This retrospective study included 71 colorectal lesions with paired CV and narrow-band imaging (NBI) scans. Six endoscopists independently evaluated randomized magnifying endoscopy images of vCV, CV, and NBI.
RESULTS
Interobserver agreement for pit pattern interpretation with vCV was similar to that with conventional CV, with minimal differences. Quantitative image analysis demonstrated morphologic fidelity between vCV and CV. The diagnostic accuracy of vCV for histologic prediction was similar to CV and numerically higher than NBI.
CONCLUSIONS
AI-based vCV chromoendoscopy can replicate the diagnostic interpretability of conventional CV staining without dye application.
Подготовка и перевод публикации
Обработка статьи · этапы, ошибки и управление
Ошибки и ограничения
История · было → стало
Состояние обновляется каждые 5 секунд, пока блок открыт. Процент показывается только там, где известен объём работы. Пауза останавливает продолжение этапа; текущий запрос к источнику может завершиться.
Тип публикации указан по данным источника. Уровень доказательности автоматически не присваивается.