Related Experiment Video
Updated: Jun 23, 2026
10:17
Guidelines and Experience Using Imaging Biomarker Explorer IBEX for Radiomics
Published on: January 8, 2018
13.2K
Quantitative Evaluation of Large Language Models to Streamline Radiology Report Impressions: A Multimodal
Rushabh Doshi1, Kanhai S Amin1, Pavan Khosla1
1From the Yale School of Medicine (R.D., P.K.) and Department of Radiology and Biomedical Imaging (K.S.A., S.S.B., S.C., H.P.F.), Yale School of Medicine, 333 Cedar St, New Haven, CT 06510; Yale School of Management, New Haven, Conn (H.P.F.); and Department of Health Policy and Management, Yale School of Public Health, New Haven, Conn (H.P.F.).
Radiology
|March 26, 2024
Summary
Large language models (LLMs) can simplify complex radiology reports for patients. Four tested LLMs successfully reduced report readability across various prompts, making medical information more accessible.
Area of Science:
- Medical Informatics
- Artificial Intelligence in Healthcare
- Radiology Communication
Background:
- Complex medical terminology in radiology reports can cause patient confusion and anxiety.
- Increased patient access to electronic health records necessitates improved report readability.
- Large language models (LLMs) offer a potential solution for simplifying clinical documentation.
Purpose of the Study:
- To evaluate the effectiveness of four publicly available LLMs in simplifying radiology report impressions.
- To compare the performance of ChatGPT-3.5, ChatGPT-4, Gemini (formerly Bard), and Bing in generating patient-friendly radiology summaries.
- To assess how different prompting strategies influence LLM simplification capabilities.
Main Methods:
- A retrospective analysis of 750 anonymized radiology report impressions from the MIMIC-IV database.
- Utilized three distinct prompts to test LLM simplification: general simplification, patient-specific simplification, and seventh-grade reading level simplification.
- Assessed simplification using readability scores derived from four established indexes and compared results with the Wilcoxon signed-rank test.
Main Results:
- All four LLMs demonstrated a statistically significant ability to simplify radiology report impressions across all tested prompts (P < .001).
- Specifying the audience as a patient or requesting a seventh-grade reading level further reduced the complexity of the simplified reports for all models.
- Variations in simplification performance were observed among the LLMs depending on the prompt used.
Conclusions:
- All evaluated LLMs successfully simplified radiology report impressions, indicating their potential utility in improving patient understanding.
- Prompt wording significantly impacts the degree of simplification achieved by LLMs.
- LLMs show promise in enhancing patient comprehension of radiology findings across diverse imaging modalities and anatomical regions.
Related Concept Videos
02:00Molecular Models
Physical models representing molecular architectures of chemical compounds play essential roles in understanding chemistry. The use of molecular models makes it easier to visualize the structures and shapes of atoms and molecules.
01:07What is a Mode?
The mode is one of the commonly used measures of a central tendency. It is defined as the most frequent value in a data set.
There can be more than one mode in a data set if multiple values have the same highest frequency. For instance, suppose that the Statistics exam scores of 20 students are: 50; 53; 59; 59; 63; 63; 72; 72; 72; 72; 72; 76; 78; 81; 83; 84; 84; 84; 90; 93. Here, the mode is 72, as it occurs most frequently, five times.
A data set with two modes is called bimodal. For example,...
There can be more than one mode in a data set if multiple values have the same highest frequency. For instance, suppose that the Statistics exam scores of 20 students are: 50; 53; 59; 59; 63; 63; 72; 72; 72; 72; 72; 76; 78; 81; 83; 84; 84; 84; 90; 93. Here, the mode is 72, as it occurs most frequently, five times.
A data set with two modes is called bimodal. For example,...
01:23Mechanistic Models: Compartment Models in Individual and Population Analysis
Mechanistic models are utilized in individual analysis using single-source data, but imperfections arise due to data collection errors, preventing perfect prediction of observed data. The mathematical equation involves known values (Xi), observed concentrations (Ci), measurement errors (εi), model parameters (ϕj), and the related function (ƒi) for i number of values. Different least-squares metrics quantify differences between predicted and observed values. The ordinary least squares (OLS)...
01:29Mechanistic Models: Compartment Models in Algorithms for Numerical Problem Solving
Mechanistic models play a crucial role in algorithms for numerical problem-solving, particularly in nonlinear mixed effects modeling (NMEM). These models aim to minimize specific objective functions by evaluating various parameter estimates, leading to the development of systematic algorithms. In some cases, linearization techniques approximate the model using linear equations.
In individual population analyses, different algorithms are employed, such as Cauchy's method, which uses a...
In individual population analyses, different algorithms are employed, such as Cauchy's method, which uses a...
01:12Modeling and Similitude
Scaled modeling is a fundamental technique in engineering, enabling the study of large and complex systems by creating smaller, manageable replicas that recreate critical characteristics of the original. In hydrology and civil infrastructure, for example, scaled models of dams help analyze water flow, turbulence, and pressure. This method allows for accurate predictions of real-world behavior within a controlled environment, significantly reducing the cost and time involved in full-scale...
01:30Typical Model Studies
Fluid mechanics model studies often utilize scaled-down systems to predict fluid behavior in full-scale environments, such as river flows, dam spillways, and structures interacting with open surfaces. Maintaining Froude number similarity in river models is crucial, as it replicates surface flow features like wave patterns and velocities.