Machine learning models achieve a pooled AUC of 0.834 for predicting acute kidney injury occurrence
Machine learning models show promise in predicting acute kidney injury
This systematic review and meta-analysis analyzed 188 estimates from 7,343,170 patients to evaluate machine learning (ML) models for predict…
New analysis shows machine learning can accurately predict sudden kidney damage and death risks, helping doctors identify high-risk patients…
AI models demonstrate high diagnostic accuracy for identifying osteonecrosis of the femoral head in imaging data
Artificial intelligence shows high accuracy in detecting hip bone death
This meta-analysis of 12 studies involving 16,189 hip joints found that AI models have a pooled sensitivity of 0.91 and specificity of 0.95 …
New data shows that AI models can accurately identify a serious hip condition called osteonecrosis using medical imaging like X-rays and MRI…
Medical Students Demonstrate Positive Attitudes Toward Artificial Intelligence Integration in Clinical Education
Medical students show positive attitudes toward using AI in healthcare
A meta-analysis of 96 studies reveals high student interest in AI, though significant ethical concerns and varying trust levels persist.
Most medical students express a positive outlook on artificial intelligence, showing willingness to learn and integrate it into their future…
EUS-BD shows higher technical success rates than ERCP for biliary obstruction with gastrointestinal reconstruction
Endoscopic Ultrasound Drainage Offers Better Success for Patients with Complex Stomach Surgery
This meta-analysis of observational studies evaluated EUS-BD versus ERCP in patients with biliary obstruction and prior gastrointestinal rec…
Doctors are finding that endoscopic ultrasound drainage works better than traditional methods for patients who have had complex stomach surg…
Privacy-enhancing technologies present distinct technical trade-offs and risks in biomedical AI applications
Privacy Technologies Offer Trade-offs for Biomedical AI Systems
This systematic scoping review maps privacy-enhancing technologies (PETs) across the biomedical AI lifecycle, including federated learning a…
New research maps how privacy-enhancing technologies protect patient data in medical AI while highlighting specific technical risks and perf…
NeVa stent retriever achieves near-complete recanalization of superior sagittal sinus in cerebral venous sinus thrombosis
Stent retriever helps woman with blood clot in brain vein
This case report describes the use of a NeVa stent retriever for mechanical thrombectomy in a patient with cerebral venous sinus thrombosis.…
A new way to remove a dangerous blood clot from a brain vein helped a 32-year-old woman recover her health and mobility.
DTI-ALPS index reduced in traumatic brain injury, meta-analysis finds
Brain imaging shows how head injuries affect fluid drainage
This meta-analysis of 694 TBI patients and 503 healthy controls found significantly reduced DTI-ALPS values in TBI, suggesting glymphatic im…
A new look at brain scans reveals a specific marker for impaired fluid drainage in people with traumatic brain injuries.
Elevated intracranial pressure prevalence reaches 38% to 43% in children with single-suture craniosynostosis
New Data Shows Elevated Pressure Risk in Children with Craniosynostosis
This meta-analysis of 720 children with isolated single-suture craniosynostosis found a pooled prevalence of elevated intracranial pressure …
A large review of children with single-suture craniosynostosis shows that nearly 40% may have elevated intracranial pressure.
CT-based CPAK measurements show more varus MPTA and valgus LDFA than LLR
CT Scans Show Different Knee Alignment Than X-rays
This retrospective cohort study of 241 patients undergoing robotic-assisted total knee arthroplasty compared CT and long-leg radiograph (LLR…
New analysis shows that CT scans and traditional X-rays can produce different alignment measurements for patients undergoing knee replacemen…
Safety user interface bundles improve verification intentions in older adults using generative AI
Safety prompts boost older adults' AI verification
Researchers evaluated a safety user interface bundle designed to influence how older adults interact with generative AI. The intervention wa…
A simple interface change made older adults more likely to double-check AI answers, without hurting trust or ease of use.
LLM lay summaries improve patient satisfaction and subjective comprehension without increasing objective comprehension
AI Summaries Improve Patient Satisfaction After Brain MRI Reports
This randomized controlled trial of 2,727 adult participants found that LLM-generated lay summaries for brain MRI reports significantly impr…
Adding AI-generated summaries to brain MRI reports can significantly improve how patients feel about their results and their personal sense …
U-Net architecture shows higher sensitivity and precision for subdural hematoma detection in meta-analysis
Machine learning improves detection of brain bleeding in CT scans
This meta-analysis of 67,266 non-contrast CT scans evaluated machine learning models for subdural hematoma detection. The analysis found U-N…
A specific type of machine learning model shows higher accuracy and precision in finding subdural hematomas on head CT scans.