Dear AuntMinnie.com Member,
Widespread adoption of radiology AI technology has yet to be achieved, owing in large part to reimbursement issues; payors don’t reimburse providers for use of most types of medical imaging AI software due to a perceived lack of impact on patient outcomes.
In a column that was our most highly read article this week, Sanjay Parekh, PhD, of AI software developer Perspectum shared his thoughts on business cases for imaging AI.
Nonhospital sites are facing a more competitive environment for oncology-related imaging, according to an article from AuntMinnie.com’s sister company IMV Medical Information Division. You can access this week’s second-most viewed story by clicking here.
January is always a popular month for resolutions and predictions. Signify Research has offered its top 10 predictions for medical imaging in 2024, including five for imaging IT and AI and five for imaging modalities.
In other stories attracting the attention of AuntMinnie.com members, a multimodal AI model that combined data from mammography and ultrasound images as well as other characteristics yielded promising performance for predicting disease-free survival of patients with breast cancer.
Also, ChatGPT-4 delivered better results than ChatGPT-3.5 for assessing text-based medical imaging cases, according to a new study.
See the full list below of our top stories of the week:
- Reimbursement issues give impetus to AI adoption
- Nonhospital sites see increased competition for oncology-based imaging
- Signify Research’s medical imaging predictions for 2024
- Deep learning combines breast imaging data to predict cancer prognosis
- ChatGPT-4 shows improvement over GPT-3.5 in text-based imaging cases
- Can ictal SPECT make a comeback in epilepsy patients?
- Cancer rates decline, health disparities exist
- IDR process financially unfeasible for radiologists
- Nanox mourns passing of company founder
- Imvaria gets FDA marketing authorization for Fibresolve
- Medical students gain confidence in radiology boot camp
- Researchers offer age-specific dose estimator for fluoroscopy exams
- Automated protocoling algorithm reduces radiologist CT workload
- ACS: Cancer mortality declines, but incidence rates rise



















![A normal mammogram confirmed by three-year radiologic follow-up illustrates reader-marked regions of interest (ROIs) during (A) unaided (round 1) and (B) artificial intelligence (AI)–assisted (round 2) reading. Each colored dot represents an ROI for recall by a human reader. Readers could mark more than one ROI per case, represented by multiple dots of the same color. During AI-assisted reading, the AI system displayed three visible prompts: two with suspicion of malignancy scores of 35% (left mediolateral oblique [L MLO] and craniocaudal [L CC]) and one with a suspicion of malignancy score of 10% (right craniocaudal [R CC]), shown as polygonal overlays. Without AI, six of 10 readers (60%) marked a false-positive ROI. With AI assistance, this fell to two of 10 (20%). R MLO = right mediolateral oblique.](https://img.auntminnie.com/mindful/smg/workspaces/default/uploads/2026/07/2026-07-14-radiology-mammogram-ai-auto-bias.H0bYO8QlWs.jpg?auto=format%2Ccompress&fit=crop&h=112&q=70&w=112)