Week in Review: AI and radiologist liability | PET and alcohol abuse | SalaryScan

If a radiologist missed a key finding but AI software didn’t, how would a jury determine legal culpability? Researchers from Brown University recently explored that question and came up with some interesting findings. 

After surveying hundreds of jurors, the study team concluded that although that scenario would negatively affect the defense, the effect could be substantially mitigated by providing the panel with data on the algorithm’s error rates. Discover what else they concluded in our most highly viewed article this week.

In a new study, PET has found sex-related differences in the brains of those with alcohol use disorder. Women had significantly lower levels of protein involved in the brain’s immune response, according to the authors.

Other popular stories this week included reports on how AI-directed breast MRI could yield shorter scan times, how an explainable deep-learning model could aid in diagnosis of liver cancer, and how radiology AI saves time in a clinical setting. In addition, learn how technology-enabled interventions were able to help reduce unnecessary recommendations for additional imaging. 

Finally, we’re also excited to launch SalaryScan, our annual initiative that provides our members with the latest data on compensation and benefits in radiology. Please take a few minutes now to fill out this short survey. As always, your responses will be kept totally anonymous and will help to produce the most accurate results. 

See below for the full list of our top stories of the week: 

1. If AI finds an abnormality that a radiologist misses, who’s at fault? 

2. PET reveals differences in people with alcohol abuse disorder 

3. Experts weigh in on prior authorization reform bill

4. AI-directed breast MRI scanning may lead to shorter scan times 

5. Explainable DL MRI model shows promise diagnosing liver cancer 

6. Interventions help reduce additional wasteful imaging recommendations 

7. MRI-based fusion model determines HER2 status in breast cancer 

8. Men appear more vulnerable to dynamic carotid plaque 

9. The PACSman Pontificates: Should AI be free?

10. Chemotherapy negatively affects women’s cognitive function 

11. Researchers urge better LCS for smokers with mental health conditions 

12. Podcast: Florence Doo on how AI is affecting radiology

13. AI model saves time in live radiology clinical practice setting 

14. IMV: patient satisfaction tops mammography department priorities