LLM-based QA system shows feasibility for screening AI auto-contouring in male pelvic radiotherapy
Can an AI assistant help doctors check AI-drawn cancer treatment maps?
A feasibility study of 20 male pelvic CT scans found an LLM-based quality assurance (LAQUA) system demonstrated moderate to strong agreement…
An AI assistant can quickly check the quality of organ maps for radiation therapy, matching expert doctors' judgments on specific body parts…
Deep learning system with Grad-CAM shows potential for tuberculosis screening on chest X-rays
Can a simple phone app help spot tuberculosis on chest X-rays?
A technical evaluation of a deep learning system with Gradient-weighted Class Activation Mapping (Grad-CAM) for tuberculosis screening from …
An offline AI tool can spot tuberculosis on chest X-rays by highlighting the right lung areas, even on mobile phones without internet.
Deep learning QTMnet shows potential for glioma grading versus traditional 2CXM model
AI method shows promise for classifying brain tumor severity in small study
A study of 30 glioma patients compared a deep neural network-based perfusion estimation method (QTMnet) to a traditional two-compartment exc…
New AI analysis of brain scans outperformed traditional methods at telling the difference between low and high severity tumors in a small te…
Vision transformer-radiomics model predicts 6-month PFS in ovarian cancer patients with high AUC
Can a scan predict if ovarian cancer treatment will work?
A retrospective cohort study of 182 ovarian cancer patients found that a combined Vision Transformer and MedSAM embedding model achieved an …
A new computer model analyzing CT scans can predict which ovarian cancer patients will stay progression-free for six months after treatment.
Breast asymmetry analysis from mammograms shows association with short-term breast cancer risk
Can a mammogram's hidden patterns predict short-term breast cancer risk?
A matched case-control study in women with mammograms near diagnosis found that bilateral breast asymmetry measures were associated with bre…
Hidden patterns in raw mammogram images linked to breast cancer risk fade when pictures are processed for doctors to review.
Qualitative study finds cautiously positive attitudes toward AI in breast cancer screening in Serbia
Radiologists and women in Serbia show cautious openness to AI in breast cancer screening
A qualitative cohort study of radiologists and screening-aged women in Serbia found cautiously positive attitudes toward AI use in breast ca…
Serbian radiologists and women support using AI in breast cancer screening, but they remain cautious until safety and effectiveness are prov…
Partial splenic embolization in pediatric portal hypertension improves hematologic parameters
Can a partial spleen procedure help kids with liver complications?
A retrospective case series of 12 pediatric patients with portal hypertension-related hypersplenism found partial splenic embolization (PSE)…
Blocking part of a swollen spleen helped 12 children with liver pressure regain healthy blood cell counts without major complications.
Mini-review examines prophylactic embolization after endoscopic hemostasis in high-risk GI bleeding
Review examines preventive artery-blocking procedure after stomach bleeding treatment
A mini-review of 10 studies (two RCTs, three prospective, five retrospective) examined prophylactic transarterial embolization (pTAE) after …
Blocking arteries after stomach bleeding may lower rebleeding and surgery needs for high-risk patients, though guidelines do not suggest rou…
Review examines cerebral microbleeds as imaging markers in traumatic brain injury
Review examines brain microbleeds as potential markers for traumatic brain injury
A narrative review synthesizes literature on cerebral microbleeds (CMBs) in traumatic brain injury.
Tiny brain bleeds seen on MRI scans may signal how severe the damage is to the fibers connecting brain cells after a traumatic injury.
Meta-analysis finds 58% prevalence of dementia-relevant MRI abnormalities in LMIC patients
What does a brain scan show when dementia is suspected in poorer countries?
A systematic review and meta-analysis of 23 studies (2,513 participants) in low- and middle-income countries found a pooled prevalence of…
Brain scans in poorer countries often reveal physical changes linked to dementia, giving doctors a visual clue beyond memory tests for over …
Deep learning reconstruction improves 5.0 Tesla knee MRI quality and diagnostic agreement in 69-patient cohort
Deep learning algorithm improves knee MRI image quality in small study
A prospective cohort study of 69 patients undergoing knee arthroscopy and 5.0 Tesla MRI found that deep learning-based reconstruction (DLR) …
A computer program using deep learning improved knee MRI image clarity by up to 350% compared to standard scans in a small study.
AI software modestly improves AUROC and reduces interpretation time in breast ultrasound reading
AI software modestly improves breast ultrasound analysis speed and performance in small study
A retrospective multi-reader cohort study of six radiologists interpreting 258 breast ultrasound examinations found that Vis-BUS AI software…
AI software helped radiologists read breast ultrasound images twice as fast, but it did not improve their accuracy in this small study.