A Rapid, Reproducible AI-Assisted Workflow for Generating Course-Aligned Anki Decks in Biomedical Sciences Education: A Tutorial (Preprint)

<sec> <title>BACKGROUND</title> <p>The Marian University Biomedical Science (BMS) Master’s program is an intensive graduate-level program designed to mimic the rigor of the first year of medical school. This requires students to implement self-directed strategies to manage cognitive load and support retention. Anki is a digital flashcard program that enables spaced repetition and active recall, but manual flashcard generation is time consuming. Here we describe a reproducible workflow using ChatGPT 5.2 to generate Anki decks aligned to course objectives and lecture content.</p> </sec> <sec> <title>OBJECTIVE</title> <p>To describe and demonstrate a reproducible AI-assisted workflow for rapidly generating course-aligned Anki decks from instructor-provided lecture materials in medical sciences education curriculum.</p> </sec> <sec> <title>METHODS</title> <p>Lecture slides were downloaded as editable PowerPoint files and modified during class to remove identifying information, delete non-testable information, clarify key points, and condense extra content. After lecture, edited slides were exported as a PDF and provided to a ChatGPT 5.2 Thinking session with a lecture-specific prompt. ChatGPT generated two import-ready CSV files per lecture, which were imported into Anki.</p> </sec> <sec> <title>RESULTS</title> <p>Across six completed examination blocks, 17,222 total Anki cards were generated from BMS coursework. Mean deck size was 106 cards per lecture. Cards comprised 56% Basic (9,644) and 44% Cloze (7,578). Deck generation required ~3-5 minutes per lecture, and &gt;99% of cards were retained in the decks with only minor edits for clarity or focus</p> </sec> <sec> <title>CONCLUSIONS</title> <p>This AI-assisted workflow functions as a scalable learner-support tool to rapidly produce high-yield, objective-aligned Anki decks in an accelerated graduate curriculum, enabling immediate post-lecture studying and recall practice customizable to course content and individual student needs. To our knowledge, this represents one of the largest implementations of AI-assisted flashcard generation within medical curriculum.</p> </sec>