Purpose/Significance To screen high-potential patents from the perspective of application scenarios, and to promote the clinical translation of medical technologies. Method/Process A classification system for medical technology application scenarios and functional characteristics is constructed. Four patterns, namely, emergency, dominance, expansion and recombination,are defined to assess substitution potential, the substitute ability is evaluated based on a three-dimensional matrix of patent scenario-function-time. Feature extraction is automated through deep learning and large language model (LLM) technologies. An empirical study is conducted by selecting 10 151 patents with neuroregulation technology. Result/Conclusion This technology opportunity discovery framework effectively identifies patterns of technological substitution in neuroregulation, providing a novel approach for the proactive identification of medical technology patents with high application potential.