Quality assessment of the inflammatory bowel disease biopsy diagnosis using digital images

Introduction. Morphological diagnosis of inflammatory bowel diseases (IBDs) based on endoscopic biopsies remains a rather challenging task and does not always allow the pathologist to make the final conclusion. Difficulties arise due to the absence of a classical morphological picture in biopsy specimens as well as the lack of clinical and laboratory data. These issues may lead to an incorrect interpretation of the detected pathological changes and an erroneous conclusion. The aim of the study was to assess the quality of biopsy diagnosis of IBDs using Internet diagnostic tools, namely, the digitalized histological slides. Materials and methods. We created an Internet platform, which contained 100 scanned images of histologi-cal slides from 70 patients with suspected IBD. Based on our own practical experience and guidelines of international European and British societies of pathology, we arranged a survey. The diagnosis was made on the combined clinical and morphological approach, considering also a complex of clinical, laboratory, endoscopic, and morphological data. The first stage of the survey was anonymous and included a list of 25 questions. Ten pathologists, who agreed to take part in the study, completed the questionnaire. After statistical processing, the results were assessed using a Fleiss' kappa criterion. Results. The comparative analysis of morphological study showed that the participants were able to diagnose accurately only 71% of ulcerative colitis cases and 63% of Crohn’s disease cases. At the same time, the agreement coefficient demonstrated satisfactory values: the Fleiss' kappa was 0.34 and 0.25, respectively. In some cases, the participants found it difficult to perform a differential diagnosis between IBD and other forms of colitis and were not always capable of distinguishing ulcerative colitis from Crohn’s disease. A few of the individual criteria, on which the diagnosis is based, rarely reached agreement above a kappa of 0.5. The participants’ agreement was considered more often as “weak” and “satisfactory”. This is largely due to the fact that we did not discuss the exact criteria for the proposed parameters and that these parameters were divided into four degrees of severity from 0 to 3. Conclusion. The results showed the current discrepancies in the assessment of biopsies in patients with suspected IBD. The developed Internet platform provides additional significant opportunities to improve the quality of morphological diagnosis in ileal and colon biopsies by means of refining the criteria and their assessment in the final diagnosis statement. Keywords: inflammatory bowel disease, colitis, digital pathology