Related Experiment Video
Updated: May 12, 2026
09:44
Manual Muscle Testing: A Method of Measuring Extremity Muscle Strength Applied to Critically Ill Patients
Published on: April 12, 2011
83.4K
Evaluating the Diagnostic Accuracy of ChatGPT-4.0 for Classifying Multimodal Musculoskeletal Masses: A Comparative
Wolfram A Bosbach1,2, Luca Schoeni1,2, Claus Beisbart3,4
1Department of Nuclear Medicine, Inselspital, Bern University Hospital, University of Bern, Bern, Switzerland.
Rofo : Fortschritte Auf Dem Gebiete Der Rontgenstrahlen Und Der Nuklearmedizin
|June 3, 2025
Summary
Artificial intelligence, specifically ChatGPT-4.0, showed lower accuracy than human raters in diagnosing musculoskeletal tumors. Including secondary diagnoses improved AI performance, suggesting its potential as an assistive tool for radiologists.
Area of Science:
- Radiology and Medical Imaging
- Artificial Intelligence in Medicine
- Musculoskeletal Oncology
Background:
- Artificial intelligence (AI) tools offer potential for enhancing medical productivity and treatment quality.
- Evaluating the diagnostic capabilities of advanced AI models like ChatGPT-4.0 in complex medical scenarios is crucial.
Purpose of the Study:
- To assess the accuracy of ChatGPT-4.0 in interpreting multimodal musculoskeletal tumor cases.
- To compare the diagnostic performance of ChatGPT-4.0 against human raters.
Main Methods:
- Twenty-five musculoskeletal tumor cases with multimodal imaging (X-ray, CT, MRI, scintigraphy) were used.
- ChatGPT-4.0 and human raters classified cases, allowing for primary and secondary diagnoses.
- Performance was evaluated based on diagnostic accuracy, with power analysis confirming sample size adequacy.
Main Results:
- Human raters achieved significantly higher accuracy (87%) than ChatGPT-4.0 (44%) for primary diagnoses.
- When secondary diagnoses were considered, the accuracy gap narrowed, with human raters at 94% and ChatGPT-4.0 at 71%.
- AI demonstrated lower performance but showed improvement when secondary diagnoses were included.
Conclusions:
- ChatGPT-4.0 currently underperforms human experts in primary diagnosis of musculoskeletal tumors.
- The inclusion of secondary diagnoses substantially improves AI performance, narrowing the diagnostic gap.
- AI tools show promise as valuable assistive systems for clinicians in radiological workflows, despite current limitations.
Related Concept Videos
01:48Classification of Skeletal Muscle Fibers
Skeletal muscles continuously produce ATP to provide the energy that enables muscle contractions. Skeletal muscle fibers can be categorized into three types based on differences in their contraction speed and how they produce ATP, as well as physical differences related to these factors. Most human muscles contain all three muscle fiber types, albeit in varying proportions.
Slow-Twitch Muscle Fibers
Slow oxidative, muscle fibers appear red due to large numbers of capillaries and high levels of...
Slow-Twitch Muscle Fibers
Slow oxidative, muscle fibers appear red due to large numbers of capillaries and high levels of...
01:18Classification of Bones
The bones of the human skeletal system are of varied shapes, sizes, and functions. They can be classified based on their shape and function into four major classes: long bones, short bones, flat bones, and irregular bones. Some classifications include a fifth type, the sesamoid bones, as a separate class, whereas others categorize them under short bones.
Long and Short Bones
The appendicular skeleton, particularly the upper and lower limbs, is primarily made of long and short bones. The long...
Long and Short Bones
The appendicular skeleton, particularly the upper and lower limbs, is primarily made of long and short bones. The long...