A Neural Network-based Algorithm for Assessing the Cleanliness of Small Bowel during Capsule Endoscopy
Résumé
Background and Aims. Cleanliness scores in small bowel (SB) capsule
endoscopy (CE) have poor reproducibility. The aim of this study was to
evaluate a neural network (NN)-based algorithm for automated
assessment of the SB cleanliness during CE.
Methods: First, 600 normal third-generation SBCE still frames were
categorized as “adequate” or “inadequate” in terms of cleanliness by
three expert readers, according to a 10-point scale and served as a
training database. Then, 156 third-generation SBCE recordings were
categorized in a consensual manner as “adequate” or “inadequate” in
terms of cleanliness. This testing database was split into two
independent 78-video subsets for the tuning and evaluation of the
algorithm.
Results: Using a threshold of 79% adequate still frames per video to
achieve the best performance, the algorithm yielded a sensitivity of
90.3%, a specificity of 83.3%, and an accuracy of 89.7%. The
reproducibility was perfect. The mean calculation time per video was 3 ±
1 minutes. Conclusion: This NN-based algorithm allowing automatic assessment of
SB cleanliness during CE was highly sensitive and paves the way for
automated, standardized SBCE reports.