Forecast emerging AI-enabled healthcare systems with the text mining framework based on lingo algorithm

This study establishes an integrated text mining framework to forecast developmental trends and structural mismatches in AI-enabled healthcare diagnostic systems, by jointly analyzing scientific literature and patent data from 2018 to 2022. We employ the LINGO clustering algorithm combined with singular value decomposition (SVD), TF-IDF weighting, and expert evaluation to identify technological themes, map lifecycle stages, and construct a systematic technology roadmap. Results reveal that enabling technologies including diagnostic imaging, machine learning, and the Internet of Things (IoT) have reached maturity, while disease-oriented applications such as neurological disorders and chronic diseases remain in early growth phases, demonstrating a clear structural asymmetry between technological maturity and clinical readiness. A notable temporal lag between scientific research output and patent commercialization is observed, with a 1-2 year gap in most disease related domains. Notably, canceroriented AI diagnostics exhibit strong growth potential and high investment value. Meanwhile, ethical, legal, and data security constraints are increasingly prominent and may restrict largescale deployment. Distinct from singlesource analyses, this work innovatively integrates literature and patents within a unified forecasting paradigm, reveals the divergence between foundational technologies and clinical translation, and provides a mechanismlevel interpretation for the science-commercialization gap. This study offers theoretical and practical insights for understanding evolutionary pathways, bridging application gaps, and guiding investment and policy decisions in AI-driven healthcare diagnostics.