Related Experiment Video
Updated: Jul 9, 2026
03:14
Augmenting Large Language Models via Vector Embeddings to Improve Domain-Specific Responsiveness
Published on: December 6, 2024
Augmenting medical data interpretation with Large Language Models (LLMs): a comparative analysis of patient
1Department of Information Systems and Business Analytics, College of Business, Florida International University (FIU), Modesto A. Maidique Campus, 11200 S.W. 8th St, RB 261B, Miami, FL, 33199, USA. pesmaeil@fiu.edu.
BMC Medical Informatics and Decision Making
|July 7, 2026
Summary
Large Language Models (LLMs) enhance patient understanding and efficiency in interpreting medical data, complementing healthcare professionals who excel in building trust and providing emotional support. Both approaches are valuable for patient care.
Area of Science:
- Medical informatics
- Human-computer interaction
- Digital health
Background:
- Traditional medical data interpretation by professionals can limit patient autonomy.
- Large Language Models (LLMs) offer potential for direct patient access to AI-generated interpretations.
- Limited comparative research exists on LLM effectiveness across medical data types and communication methods.
Purpose of the Study:
- To compare LLM-augmented interpretation with healthcare professional-led interpretation for medical data.
- To evaluate patient comprehension, empowerment, and technology acceptance across different modalities.
- To investigate the influence of data type and patient characteristics on interpretation effectiveness.
Main Methods:
- A mixed-methods approach with a within-subjects experimental design.
- 45 diverse participants experienced interpretations of blood work and medical imaging via phone, in-person, and a custom LLM interface (Medical Explainer AI).
- The LLM provided plain-language explanations, highlighted abnormalities, contextualized significance, and adapted complexity.
Main Results:
- LLM interaction significantly improved diagnostic comprehension, reduced cognitive load, and increased perceived control and time efficiency compared to phone consultations.
- In-person professional interpretation fostered greater trust, reduced anxiety, and enhanced decision-making confidence.
- LLM benefits were more pronounced for blood work than medical imaging; age, education, and health literacy moderated effectiveness.
Conclusions:
- LLMs serve as a valuable complement to, rather than a replacement for, healthcare professionals in medical data interpretation.
- LLMs excel at enhancing comprehension, control, and efficiency, while professionals provide crucial relational value.
- Implementation should strategically combine LLM strengths with professional expertise, considering data complexity and patient factors for equitable access.
