| Popis: |
Takanobu Hirosawa,1,* Tomoharu Suzuki,2,* Tastuya Shiraishi,3,* Arisa Hayashi,1,* Yoichi Fujii,4,* Taku Harada,1,4,* Taro Shimizu1,* 1Department of Diagnostic and Generalist Medicine, Dokkyo Medical University, Tochigi, Japan; 2Department of Hospital Medicine, Urasoe General Hospital, Okinawa, Japan; 3Higashinihonbashinaika clinic, Tokyo, Japan / Ubie, inc, Tokyo, Japan; 4General Medicine, Nerima Hikarigaoka Hospital, Tokyo, Japan*These authors contributed equally to this workCorrespondence: Takanobu Hirosawa, Department of Diagnostic and Generalist Medicine, Dokkyo Medical University, 880 Kitakobayashi, Mibu-cho, Simotsuga-gun Tochigi, Tochigi, 321-0293, Japan, Tel +81-282-86-1111, Fax +81-282-86-4775, Email hirosawa@dokkyomed.ac.jpPurpose: Artificial intelligence (AI) holds great potential for revolutionizing health care by providing clinicians with data-driven insights that support more accurate and efficient clinical decisions. However, applying AI in clinical settings is often challenging due to the complexity and vastness of medical information. This perspective article explores how AI development methodologies can be adapted to support clinicians in their decision-making processes, emphasizing the importance of a hybrid approach that combines AI capabilities with clinicians’ expertise.Patients and Methods: We developed a conceptual framework designed to integrate AI-driven hybrid intelligence into clinical practice to enhance decision-making. This framework focuses on adapting key AI concepts, such as backpropagation, quantization, and avoiding overfitting, to help clinicians better interpret complex medical data and improve diagnosis and treatment planning.Results: Several AI methodologies were adapted to enhance clinical decision-making. First, backpropagation allows clinicians to refine initial assessments by revisiting them as new data emerges, improving diagnostic accuracy over time. Second, quantization helps break down complex medical problems into manageable components, enabling clinicians to prioritize critical elements of care. Finally, avoiding overfitting encourages clinicians to balance rare diagnoses with more common explanations, reducing the risk of diagnostic errors and unnecessary complexity.Conclusion: The integration of AI-driven hybrid intelligence has the potential to enhance clinical decision-making. By adapting AI methodologies, clinicians can enhance their ability to analyze data, prioritize treatments, and make more accurate diagnoses while preserving the essential human aspect of health care. This framework highlights the importance of combining AI’s strengths with clinicians’ expertise for more effective and balanced decision-making in clinical practice. This perspective highlights the value of hybrid intelligence in achieving more balanced, effective, and patient-centered decision-making in health care.Keywords: artificial intelligence, clinical reasoning, diagnostic accuracy, digital health, internal medicine, natural language processing |
