Purpose
A key challenge in medical education is helping students transfer basic science knowledge—typically learned through deductive reasoning (disease, e.g., pneumonia → symptoms) to clinical settings that require inductive reasoning (symptoms, e.g., cough, fever → differential diagnoses). Few tools train students to shift between these reasoning modes. To address this, we developed a dual-reasoning artificial intelligence driven (AI) app that fosters active recall, illness script building, and integration of theory with practice.

Methods
Users chose deductive or inductive reasoning modes at three difficulty levels. Deductive mode starts with a diagnosis, and inductive mode starts with symptoms, and users generate differentials. Cases unfold stepwise, with the case and then new findings to test users' reasoning, illness scripts, and anchoring bias. This step also involves knowledge recall through written justifications, followed by delayed feedback. AI feedback includes users vs “expert” comparison, key knowledge gaps, and next steps. Premed students (n= 4, Ignite program) guided usability enhancements. Clinicians reviewed the app for medical accuracy and clinical relevance.

Results
Clinicians supported the reasoning design but recommended safeguards to ensure clinical accuracy, prompting the use of authoritative medical sources and case validation. Student feedback led to added features such as medical term definitions (in basic mode), normal value references, and a persistent case-viewing panel. The System Usability Scale (SUS) score averaged 85.0 ± 7.9, reflecting high usability. Future work includes tracking expert vs. novice reasoning performance and curating validated cases.

Conclusion
By training both deductive and inductive reasoning, the app enhances cognitive flexibility and improves diagnostic thinking. The app helps bridge the gap between textbook learning and clinical decision-making. We plan to avail this instructional innovation tool soon.