The Modality Gap in Foot and Ankle Artificial Intelligence: A Scoping Review Highlighting the Absence of Language-Based Diagnostic Reasoning Models

Artificial intelligence (AI) research in orthopaedics has predominantly focused on image-based computer vision tasks. However, clinical diagnosis in foot and ankle surgery frequently relies on language-based information, including patient-reported symptoms, functional complaints, and contextual clinical reasoning rather than imaging alone. This scoping review aims to systematically map the current landscape of AI research in foot and ankle surgery across data modalities (vision, signal/sensor, and language) and across clinical task complexity (perception/detection versus diagnostic reasoning and decision support). Using an evidence gap map, this review will quantify the imbalance between image-based AI and language-based reasoning models and identify critical evidence voids, particularly regarding the use of natural language processing (NLP) and large language models (LLMs) for diagnostic reasoning.