| Abstract | Background:
NotebookLM is Google’s AI-powered research and note-taking assistant that uses Gemini AI to help users synthesise, understand, and generate insights from their own documents.
One of its standout features is the Audio Overview, which generates accessible two-person AI-hosted podcasts from uploaded sources. However, concerns remain around factual accuracy and privacy, especially if real patient data is used. In transfusion science—a field critical to healthcare—tools that enhance comprehension are potentially valuable to educators, students, and medical laboratory scientists alike. However, it is important to assess whether NotebookLM can use textbook material to generate engaging, easy-to-understand podcasts with case studies, and solve a case study relevant in the field of transfusion science, such as in Haemolytic Disease of the Foetus and Newborn (HDFN). Investigations:
Test 1: A textbook excerpt about Quality Management in Blood bank was uploaded to evaluate the podcast’s accuracy, engagement, and content creation.
Test 2: An HDFN case study was inputted to assess its problem-solving ability.
A prompt was used for both: “Make easy to understand, focus on medical laboratory science students. End with a case study.” Results:
Test 1: The podcast accurately reflected the source content. Using its interactive beta feature, NotebookLM explained complex concepts through effective analogies. It also created a relevant case study on a mislabelled blood sample to illustrate quality systems.
Test 2: Although NotebookLM could not fully solve the HDFN case, it correctly identified the clinical context and generated a podcast explaining investigation protocols. Summary:
NotebookLM is a promising AI tool for making transfusion science education more engaging and accessible. However, it is currently limited to interpreting source material and requires expert oversight. Future development may expand its capability to perform more complex analyses independently. |
|---|