ENHANCING SURGICAL EFFICIENCY THROUGH ARTIFICIAL INTELLIGENCE IN THE SURGICAL MANAGEMENT OF ADHESIVE OTITIS MEDIA

Abstract

Adhesive otitis media is a chronic middle-ear disorder characterized by severe tympanic membrane retraction, progressive adhesion of the tympanic membrane to middle-ear structures, conductive or mixed hearing loss, and a risk of ossicular damage, recurrent atelectasis, and secondary cholesteatoma. Contemporary surgical management includes cartilage tympanoplasty, endoscopic transcanal ear surgery, selective ossicular reconstruction, and, in selected cases, adjunctive Eustachian tube balloon dilatation. Recent clinical studies suggest that endoscopic surgery provides excellent visualization of retraction limits and retrotympanic spaces, while cartilage tympanoplasty combined with Eustachian tube balloon dilatation may further improve hearing outcome and quality of life in selected patients. Direct evidence for artificial intelligence in adhesive otitis media surgery remains limited. However, adjacent otologic AI literature has shown promising performance in tympanic membrane image classification, temporal bone CT interpretation for chronic middle-ear disease, and hearing-related prediction tasks. A 2024 review in otology noted that AI is already being used for eardrum-image diagnosis, temporal bone CT analysis, and postoperative hearing-outcome prediction after tympanoplasty, while a 2024 explainable 3D convolutional neural network achieved strong performance in CT-based evaluation of chronic otitis media. Image-based systematic reviews have also reported average diagnostic accuracies around 90.8% for combined segmentation and classification tasks in middle-ear disease. The aim of this modeled study was to evaluate whether an AI-assisted workflow could improve preoperative planning, intraoperative efficiency, and 12-month postoperative outcomes in 135 patients with adhesive otitis media treated at the TDTU ENT Department. The modeled protocol combined standard otologic assessment, endoscopic staging, temporal bone CT, AI-assisted tympanic membrane image analysis, AI-supported CT interpretation, risk-based surgical decision support, and postoperative outcome modeling. Synthetic results suggested that AI support improved diagnostic concordance, reduced time to final operative planning, shortened operating time, reduced intraoperative plan changes, and improved hearing and graft-related outcomes at 12 months.