Automated software-assisted diagnosis of esophageal squamous cell neoplasia using high-resolution microendoscopy

Background and Aims: High-resolution microendoscopy (HRME) is an optical biopsy technology that provides subcellular imaging of esophageal mucosa but requires expert interpretation of these histopathology-like images. We compared endoscopists with an automated software algorithm for detection of esophageal squamous cell neoplasia (ESCN) and evaluated the endoscopists' accuracy with and without input from the software algorithm. Methods: Thirteen endoscopists (6 experts, 7 novices) were trained and tested on 218 post-hoc HRME images from 130 consecutive patients undergoing ESCN screening/surveillance. The automated software algorithm interpreted all images as neoplastic (high-grade dysplasia, ESCN) or non-neoplastic. All endoscopists provided their interpretation (neoplastic or non-neoplastic) and confidence level (high or low) without and with knowledge of the software overlay highlighting abnormal nuclei and software interpretation. The criterion standard was histopathology consensus diagnosis by 2 pathologists. Results: The endoscopists had a higher mean sensitivity (84.3%, standard deviation [SD] 8.0% vs 76.3%, P Z.004), lower specificity (75.0%, SD 5.2% vs 85.3%, P <.001) but no significant difference in accuracy (81.1%, SD 5.2% vs 79.4%, PZ.26) of ESCN detection compared with the automated software algorithm. With knowledge of the software algorithm, the specificity of the endoscopists increased significantly (75.0% to 80.1%, P Z.002) but not the sensitivity (84.3% to 84.8%, PZ.75) or accuracy (81.1% to 83.1%, PZ.13). The increase in specificity was among novices (P Z.008) but not experts (P Z.11). Conclusions: The software algorithm had lower sensitivity but higher specificity for ESCN detection than endoscopists. Using computer-assisted diagnosis, the endoscopists maintained high sensitivity while increasing their specificity and accuracy compared with their initial diagnosis. Automated HRME interpretation would facilitate widespread usage in resource-poor areas where this portable, low-cost technology is needed.

Automated summary

After training and testing, endoscopists showed higher sensitivity but lower specificity compared to the automated software algorithm for ESCN detection. With knowledge of the software algorithm, endoscopists improved their specificity significantly while maintaining high sensitivity and accuracy, especially among novices. The use of computer-assisted diagnosis could facilitate widespread adoption of HRME technology in resource-poor areas.