Small study finds no clear link between endometriosis and ART outcomes
The retrospective analysis found no significant differences in embryo development or early pregnancy outcomes, but its authors said larger prospective studies are needed.
Search
2,230 results
581–600
HNF4α-enhanced microtissues show more liver-like behavior
A laboratory model using engineered Huh7 cells produced stronger liver markers and altered drug-response signals, but the shift remained partial.
Review finds varied neurological disease after immune checkpoint therapy
A review of 60 published cases describes symptoms, treatment and outcomes after immune checkpoint inhibitor therapy, but cannot measure individual risk.
European analysis reports seizure reductions in selected cenobamate group
In 75 adults with highly drug-resistant focal epilepsy, 33.3% were seizure-free at the last visit, but the selected study cannot show that cenobamate caused the result.
AI Tools for Autism Show Promise, but Data Gaps Remain
A peer-reviewed review of brain imaging and machine learning says broader datasets, shared standards and clinical validation are still needed.
Aphasia review finds promise in therapy, but no clear winner
A review of six randomized trials suggests intensive language therapy and neuromodulatory approaches may help, while uneven methods make the best approach uncertain.
Low energy availability is common among university athletes
A review links the pattern with disordered-eating risk, but mostly snapshot studies cannot show which comes first.
Small RNA linked to colorectal cancer shows promise as a blood marker
Researchers found higher levels in colorectal cancer tissue and plasma and linked the RNA to a proposed HARS2 pathway in cell and mouse experiments.
Epidural Injections Before Lumbar Fusion Linked to More Infections
A retrospective cohort study found the highest infection odds when the injection was given closest to surgery, but it did not establish a safe waiting period.
Review finds dosing software better at hitting pediatric drug targets
A peer-reviewed analysis of 26 studies and 11 platforms found better success at reaching drug targets, but direct patient outcomes remain untested.
AI Model Flags Seven HPK1 Compounds With Strong Lab Activity
A peer-reviewed study paired a multiscale affinity model with virtual screening, then found seven of 20 selected compounds exceeded 80% inhibition in a kinase assay.
Review Maps How Glioma Complement Proteins May Shield Tumors
A peer-reviewed systematic review of 21 experimental studies links complement regulators to immune evasion, tumor signaling and treatment resistance, but does not establish patient benefit.
Culturally tailored crystal meth toolkit gets positive early feedback
Preprint: In a six-year, community-led project, 73 beta testers largely found the toolkit clear and worth recommending; the study did not test whether it changes health outcomes.
Machine learning yields sharper drought signals in chickpea gene data
A comparison of three RNA-seq approaches found HDBSCAN more stress-specific in drought data, while limited salt and salinity samples weakened its resolution.
Review Finds Vascular Graft Infections Uncommon but Relevant
A 19-study analysis found a lower pooled incidence after EVAR than open surgery, while major differences between studies cloud the comparison.
ZipMVS Reports Competitive 3D Results With Less Memory
A preprint reports that ZipMVS uses fewer depth samples in its cost volume while delivering competitive reconstruction results on benchmark tests.
Model-based ad incentives posted higher profit in online tests
Preprint: In one online test, MB IQL recorded 7.9640% higher per-user net profit than TD3+BC, with a reported 1.4978% cost change.
A preprint reports higher 3D registration scores in noisy tests
The proposed Ex-Sim(3)-Reg method reported higher match quality and registration recall across disturbed, unseen and low-inlier indoor benchmark settings.
New Method Tracks Shifts in Long, Dependent Data
Preprint: A version-1 arXiv paper reports stronger results on longer simulated series and tests the method on sensor and Bitcoin data.
Review Finds AI Systems Are Judged More by Outputs Than Their Process
Preprint: A review of 259 works finds final artifacts are assessed more consistently than the revisions and decisions that produce them.