- Liver Cancer Blood Test: Cedars-Sinai investigators developed a hepatocellular carcinoma (HCC) blood test that could allow earlier HCC detection and improve patient access compared to ultrasound-led screening. The blood test detected ~90% of HCC patients in a recent Phase II study (n = 106 & 72), setting the stage for larger validation trials and potentially clinical adoption. Cedars-Sinai’s liver cancer test is another sign that blood-based cancer screening is gaining momentum, joining lung cancer tests from MGH and MD Anderson and the ongoing Multi-Cancer Early Detection study.
- Lunit INSIGHT CXR’s Accuracy & Efficiency: A JAMA study highlighted Lunit INSIGHT CXR’s ability to improve radiologists’ chest X-ray interpretation accuracy and efficiency. Six radiologists of varying experience interpreted 407 exams with and without INSIGHT CXR, finding that AI assistance improved the radiologists’ sensitivity for detecting lung nodules (62.9% vs. 56.7%), pneumonia (71.9% vs. 67.3%), pleural effusion (89.5% vs. 88.9%), and pneumothorax (96.5% vs. 79.2%). The solution also cut the rads’ interpretation times by 10% (40.8 vs 36.9 seconds)
- Sectra Expands to Genomics IT: Sectra announced the launch of its new Genomics IT business unit, as the company works towards its goal to create an integrated multi-diagnostics platform. Sectra will collaborate with the University of Pennsylvania Health System to develop its genomics IT solution, which will become part of its enterprise imaging portfolio (joining radiology, cardiology, pathology, orthopedics, and ophthalmology).
- Detecting “Hot” Arteries: Researchers successfully used a novel PET-CT technique to predict which heart attack patients were most likely to experience a recurrent event. The researchers leveraged an inflammation-binding PET tracer (18F-NaF) to highlight “hot” coronary arterial plaque in 704 patients with recent MI and multivessel disease. Compared with patients with “cold” arteries (inactive disease), those with hot arteries were almost 2.5 times more likely to die of any cause, and 82% more likely to experience coronary heart disease death or another heart attack.
- Canadian Mammogram Review: Canadian health officials will review more than 16k mammograms performed in Newfoundland and Labrador since 2018, after a recent audit found that some health systems were reading screening mammograms with substandard 3-megapixel monitors (vs. 5-megapixel standard). Of 837 exams already reviewed, an initial auditing radiologist found four “potential discrepancies or differing interpretations,” prompting officials to consider enlisting more radiologists from surrounding areas.
- AI4CMR’s US & UK Clearances: AI4CMR is on a regulatory hot streak, after its cardiac MRI AI solution gained FDA 510k clearance and UKCA Marking (UK’s post-Brexit regulatory mark). It was already CE Marked. The AI4CMR solution automatically segments and interprets CMRI exams, quantifies various CMRI parameters, and identifies cases with suspicious left and right ventricle wall motion.
- Surveillance AB-MRI Beats DBT: A study out of South Korea found that abbreviated breast MRI (AB-MRI) detected cancer at a far higher rate than digital breast tomosynthesis (DBT) among women with a personal history of breast cancer. Researchers analyzed AB-MRI and DBT surveillance exams from 471 postoperative women (w/ 11 malignancies), finding that AB-MRI detected all 11 malignancies while DBT detected 6 (CDRs: 23.4 vs. 12.7 per 1k women).
- Hospital Margins in for a Rough Q4: July’s hospital margin update from Kaufman Hall is in, and it’s tough to describe the results as anything but bleak. Median operating margins plummeted to -0.98%, revenue dropped 4.8%, and labor expenses climbed 3.5% since June. Persistent labor shortages and a growing number of patients choosing ambulatory centers for surgical procedures helped reverse any margin gains made in recent months, causing Kaufman Hall to warn health systems not to lose sight of long-term capital planning despite the urgency of day-to-day pressures.
- Echo AI Superiority: A Cedars-Sinai team developed and tested an ultrasound AI system that was able to measure cardiac function more effectively than veteran sonographers (14yr avg. tenure). Cedars-Sinai cardiologists evaluated 3,495 transthoracic echo ejection fraction reports that were either generated by AI or sonographers, making corrections to a smaller share of the AI-based reports than the sonographer reports (16.8% vs. 27.2%) and finding that AI ejection fraction readings were closer to actual EF measurements (2.8 vs. 3.8 avg. pct point variation).
- Bridging the BCa Screening Digital Divide: A new study in Radiology encouraged mammography practices to continue supporting pre-digital communications, warning that digital scheduling and patient portals can exacerbate breast cancer screening disparities. The survey of 7.2k women revealed that only 18% used a computer to schedule screening appointments, while 70% of the women who scheduled via computer were White, insured, and college educated.
- BrainKey Adds fMRI: BrainKey added fMRI to its unique AI-based brain health analysis platform, joining its existing brain MRI, genetics, and demographic analysis capabilities. The BrainKey Platform produces reports that help individuals understand their brain health (w/ longitudinal brain MRI and fMRI) and the factors influencing their future brain longevity (w/ genetics and demographics), potentially allowing early detection and personalized treatments.
