Rapid Endoscopic Diagnosis of Benign Ulcerative Colorectal Diseases With an Artificial Intelligence Contextual Framework.

BACKGROUND & AIMS: Benign ulcerative colorectal diseases (UCDs) such as ulcerative colitis, Crohn's disease, ischemic colitis, and intestinal tuberculosis share similar phenotypes with different etiologies and treatment strategies. To accurately diagnose closely related diseases like UCDs, we hypothesize that contextual learning is critical in enhancing the ability of the artificial intelligence models to differentiate the subtle differences in lesions amidst the vastly divergent spatial contexts. METHODS: White-light colonoscopy datasets of patients with confirmed UCDs and healthy controls were retrospectively collected. We developed a Multiclass Contextual Classification (MCC) model that can differentiate among the mentioned UCDs and healthy controls by incorporating the tissue object contexts surrounding the individual lesion region in a scene and spatial information from other endoscopic frames (video-level) into a unified framework. Internal and external datasets were used to validate the model's performance. RESULTS: Training datasets included 762 patients, and the internal and external testing cohorts included 257 patients and 293 patients, respectively. Our MCC model provided a rapid reference diagnosis on internal test sets with a high averaged area under the receiver operating characteristic curve (image-level: 0.950 and video-level: 0.973) and balanced accuracy (image-level: 76.1% and video-level: 80.8%), which was superior to junior endoscopists (accuracy: 71.8%, P < .0001) and similar to experts (accuracy: 79.7%, P = .732). The MCC model achieved an area under the receiver operating characteristic curve of 0.988 and balanced accuracy of 85.8% using external testing datasets. CONCLUSIONS: These results enable this model to fit in the routine endoscopic workflow, and the contextual framework to be adopted for diagnosing other closely related diseases.

The Expression and Clinical Significance of Ezrin and MMP-9 in Colorectal Cancer Tissue.

Objective: Colorectal cancer is a malignant tumor with high mortality, but is hard to detect at its early stage. Recent studies highlighted the crucial roles of Ezrin protein and MMP-9 in the development and malignancy of colorectal cancer, but Ezrin protein and MMP-9 in early diagnosis of colorectal cancer require further investigation. Therefore, we aimed to investigate their roles in the occurrence and metastasis of colorectal cancer, and to analyze their clinical significance in diagnosing and treating colorectal cancer. Method: The diagnosis of collected colorectal cancer tissue and adjacent tissue samples from colorectal cancer patients confirmed by clinical symptoms was performed using Hematoxylin Eosin staining. The expression levels of Ezrin and MMP-9 in 50 colorectal cancer tissue and 50 cases adjacent colorectal cancer tissue were detected by the immuno-histochemical MaxVision method. The relationship between the positive expression rate of Ezrin and MMP-9 in colorectal cancer tissue and clinical pathological factors was analyzed, and the correlation between Ezrin and MMP-9 was examined. Results: The positive expression rate of Ezrin in colorectal cancer tissue (78%) was significantly higher compared to adjacent non-cancerous tissues (6.0%) (P < .05). There was no significant correlation of gender/age and Ezrin/MMP-9 expressions (P > .05). The expression level of Ezrin exhibited statistically significant differences in the pathological factors including tumor diameter, depth of invasion, degree of differentiation, presence or absence of lymph node metastasis, and distant metastasis (P < .05). Additionally, the positive expression rate of MMP-9 in colorectal cancer tissue (76%) was markedly elevated compared to adjacent tissues (8.0%) (P < .05). The expression level of MMP-9 showed statistically significant differences in the pathological factors including tumor diameter, depth of invasion, degree of differentiation, presence or absence of lymph node metastasis, and distant metastasis (P < .05). In addition, the expression of Ezrin and MMP-9 in colorectal cancer tissue showed a significant positive correlation (r=0.637, P < .01). Conclusion: Ezrin and MMP-9 may synergistically participate in the occurrence, invasion, and metastasis of colorectal cancer. The combined assessment of Ezrin and MMP-9 expression levels in colorectal cancer patients holds significant potential for clinical diagnosis and personalized therapeutic applications.