Deep Bayesian active learning-to-rank with relative annotation for estimation of ulcerative colitis severity.

Automatic image-based severity estimation is an important task in computer-aided diagnosis. Severity estimation by deep learning requires a large amount of training data to achieve a high performance. In general, severity estimation uses training data annotated with discrete (i.e., quantized) severity labels. Annotating discrete labels is often difficult in images with ambiguous severity, and the annotation cost is high. In contrast, relative annotation, in which the severity between a pair of images is compared, can avoid quantizing severity and thus makes it easier. We can estimate relative disease severity using a learning-to-rank framework with relative annotations, but relative annotation has the problem of the enormous number of pairs that can be annotated. Therefore, the selection of appropriate pairs is essential for relative annotation. In this paper, we propose a deep Bayesian active learning-to-rank that automatically selects appropriate pairs for relative annotation. Our method preferentially annotates unlabeled pairs with high learning efficiency from the model uncertainty of the samples. We prove the theoretical basis for adapting Bayesian neural networks to pairwise learning-to-rank and demonstrate the efficiency of our method through experiments on endoscopic images of ulcerative colitis on both private and public datasets. We also show that our method achieves a high performance under conditions of significant class imbalance because it automatically selects samples from the minority classes.

Artificial intelligence quantifying endoscopic severity of ulcerative colitis in gradation scale.

OBJECTIVES: Existing endoscopic scores for ulcerative colitis (UC) objectively categorize disease severity based on the presence or absence of endoscopic findings; therefore, it may not reflect the range of clinical severity within each category. However, inflammatory bowel disease (IBD) expert endoscopists categorize the severity and diagnose the overall impression of the degree of inflammation. This study aimed to develop an artificial intelligence (AI) system that can accurately represent the assessment of the endoscopic severity of UC by IBD expert endoscopists. METHODS: A ranking-convolutional neural network (ranking-CNN) was trained using comparative information on the UC severity of 13,826 pairs of endoscopic images created by IBD expert endoscopists. Using the trained ranking-CNN, the UC Endoscopic Gradation Scale (UCEGS) was used to express severity. Correlation coefficients were calculated to ensure that there were no inconsistencies in assessments of severity made using UCEGS diagnosed by the AI and the Mayo Endoscopic Subscore, and the correlation coefficients of the mean for test images assessed using UCEGS by four IBD expert endoscopists and the AI. RESULTS: Spearman's correlation coefficient between the UCEGS diagnosed by AI and Mayo Endoscopic Subscore was approximately 0.89. The correlation coefficients between IBD expert endoscopists and the AI of the evaluation results were all higher than 0.95 (P < 0.01). CONCLUSIONS: The AI developed here can diagnose UC severity endoscopically similar to IBD expert endoscopists.