Jurors are more likely to find radiologists at fault if AI detects an abnormality they miss, but transparency about AI error rates can mitigate this effect.
Key Details
- 1Study evaluated over 1,300 mock jurors using vignettes of missed brain bleeds or cancer diagnoses.
- 2Jurors sided with plaintiffs 72.9% (brain bleed) and 78.7% (cancer) when AI flagged missed findings, versus 56.3% and 65.2% with no AI.
- 3Disclosure of AI's false omission (1%) or false discovery (50%) rates reduced perceived radiologist liability.
- 4If both radiologist and AI missed abnormality, jurors were less likely to fault the radiologist (50% for brain bleed, 63.5% for cancer).
- 5Providing AI error rates had stronger mitigating effects for brain bleed cases than for cancer.
Why It Matters
As AI tools become more integrated into radiology workflows, awareness of their error profiles and transparent communication may have significant legal implications. The findings can inform how radiologists document and present AI results to reduce liability risk in diagnostic misses.
Source
AuntMinnie
Related News
•Radiology Business
Study Highlights Limitations of AI in Prostate MRI Screening
New research points to several shortcomings in implementing AI for MRI-based prostate cancer screening.
•AuntMinnie
Deep Learning Model Predicts Brain Tumor MRI Enhancement Without Gadolinium
German researchers developed a deep learning approach to predict MRI contrast enhancement in brain tumors without the need for gadolinium-based agents.
•AuntMinnie
Multimodal LLMs Achieve High Accuracy Detecting Scoliosis on X-rays
Multimodal LLMs achieved up to 94% accuracy for scoliosis detection on spine x-rays, but struggled with lumbar stenosis on MRI.