New strategies for endoscopic recognition of Barrett neoplasia

One of the most prominent clinical challenges in Barrett surveillance is the endoscopic recognition of early neoplasia. In this thesis, three different approaches to increase the visualization and recognition of early Barrett neoplasia are investigated.
In part 1 (chapters 1-2) of this thesis, two new optical chromoscopy techniques are interrogated in two different clinical settings. Chapter 1 evaluates the additive value of blue light imaging (BLI) for the visualization of Barrett neoplasia when used by expert endoscopists, in particular for delineation of early BE neoplasia prior to endoscopic resection, which is generally performed in tertiary (i.e. expert) setting. Chapter 2 evaluates the additive value of optical chromoscopy techniques BLI and linked color imaging (LCI) when used by non-expert endoscopists.
Part 2 (chapter 3) describes the development and validation of an online, interactive training tool that aims to enhance endoscopists’ ability to detect and delineate Barrett neoplasia by the use of high-quality HD-WLE video materials.
Part 3 (chapters 4-8) focuses on the use of machine learning techniques for the primary detection and classification of early Barrett neoplasia. These techniques, often referred to as computer aided detection (CADe) or computer aided diagnosis (CADx) can assist endoscopists during endoscopic surveillance, by indicating areas of interest that are missed by the endoscopist or discriminate between neoplastic and non-neoplastic tissue.
In chapter 9 the main findings of this thesis are reviewed and recommendations for further research are discussed.