Headlines

A WHO forum meeting report argues that global data standards such as FHIR, SNOMED CT, LOINC, and ICD are essential for pandemic intelligence and cross-border health data exchange.

A mini review used Swiss Target Prediction to map the probable protein targets of seven major classes of antidiabetic drugs used in type 2 diabetes.

Researchers built a seven-gene tumor stemness signature from single-cell and bulk transcriptomic data that predicts survival, immunotherapy response, and oxaliplatin sensitivity in hepatocellular carcinoma.

An engineered compact Cas12m variant enables multiplexed CRISPR gene activation from a single AAV vector in mouse models.

The EXTRA Trial will test whether 24 weeks of supervised aerobic and resistance training improves fitness, immune markers, and graft-related outcomes in kidney transplant recipients.

A new graph-based framework groups similar learners and augments sparse data to improve cognitive diagnosis predictions across three public datasets.

Braided electrospun PU/ALMA scaffolds carrying decellularized tendon ECM and umbilical cord stem cells showed stronger mechanics, slower degradation and better tendon regeneration in rabbits.

A new MCDA-CRISPR/Cas12a assay detects the aminoglycoside resistance gene armA in about 70 minutes with 100 percent concordance on 129 sequenced clinical isolates.

The FDA has outlined steps drugmakers can take to seek regulatory approval for gene editing medicines tailored to fix a single person's disease-causing mutation. The announcement is part of an ongoing ...

CDMOs already play a critical role in the cell and gene therapy development and manufacturing space and, says Rose Biosolutions, with emerging bioprocessing technologies, production methods, and strategies, are creating even more opportunities for contractors. The post Decentralization, AI, and Gene Editing are Opportunities for CGT Contractors appeared first on GEN - Genetic Engineering and Biotechnology News.

Disorders of the small intestine are often difficult to diagnose and treat. More than 20 feet (6.1 meters) long, an inch (2.5 centimeters) in diameter and located deep within the body, this organ is harder to access than other parts of the digestive tract, such as the colon or stomach.

AI Clauses in Fashion Contracts A draft Calvin Klein contract asked a model to accept, for no extra

A Hiroshima University researcher outlines advantages of Japan's new law that prohibits the transfer of genome-edited human embryos or germ cells into the human reproductive tract and imposes criminal penalties for violations.

A team from the Research Institute of the McGill University Health Center (The Institute) have discovered a new type of extracellular vesicle released by migrating cancer cells, revealing a previously unknown way in which these cells shed their DNA as they move. Published today in the journal Nature Cell Biology, the findings show that these cells release tiny structures called motile vesicles (MoVes), which carry their genetic material, including cancer-associated mutations. The discovery raises important questions about the role these vesicles may play in cancer progression.

Samsung partners with Lark Health to offer eligible US seniors AI diabetes prevention coaching through Samsung Health, with Medicare eligibility restrictions. The post Samsung Health to Offer Free AI Diabetes Prevention Coaching for Seniors appeared first on TechRepublic.

A Hiroshima University researcher contends that Japan's newly enacted law regulating human genome-edited embryos and germ cells marks a significant advance in the governance of reproductive genome editing. Tsutomu Sawai, professor at Hiroshima University's Graduate School of Humanities and Social Sciences, praises the new law for defining prohibited activities and penalties for violation but also stresses the need for stronger international coordination to prevent the prohibitions from being circumvented by individuals seeking procedures abroad.

A Chinese biotech startup has raised more than $100 million in Series A funding for its AI-driven protein design. BioGeometry, a 4-year-old company headquartered in Beijing, said in a statement Wednesday ...

A JAMA study found epidural steroid injections fell 15% among Original Medicare patients in six states after an AI-assisted prior authorization pilot began, but patient benefits and harms remain unclear.

Artificial intelligence can be helpful in analyzing medical images and support radiotherapy planning workflows. However, huge amounts of sensitive patient data are needed to train robust AI models. Researchers at the Karlsruhe Institute of Technology (KIT) have investigated how hospitals can jointly develop AI models for radiation therapy applications without exchanging such data. In their project, they use federated learning, a process in which a single model is trained at a decentralized location while the data remains in place. Their results demonstrate not only the technical feasibility of the approach but also its safe, more sustainable, and accessible use.

Four seriously ill Chinese boys who would almost certainly have died in their twenties from a rare inherited disease received the same highly anticipated gene-editing drug. More than a year later, a set of cruel and contradictory data emerged. Two boys gained the strength to walk farther. One boy showed no improvement in gait and subsequently withdrew from the trial. And the fourth boy’s heart stopped beating 81 hours after receiving the medication intravenously, owing to a severe systemic...

A qualitative study of 45 stakeholders in Shenzhen finds concerns about safety, regulatory clarity, responsibility allocation and cross-sector cooperation in medical AI governance.

A year ago, Endpoints Signal surveyed almost 1,500 biopharma insiders about the very near-term prospects of AI accelerating drug discovery. Scientists, interestingly, were the most skeptical. Now, as we revisit the question in our

For decades, stroke prevention has revolved around a list of separate—and often separately treated—risk factors: blood pressure, cholesterol, diabetes, smoking and inactivity. According to Dr. Gustavo Saposnik, it's time to start considering biological age as a risk factor for stroke.

Researchers cloned a 1,874-bp ABCD4 cDNA from Ujimqin sheep liver, expressed the recombinant protein in E. coli, and generated a polyclonal antibody for future studies.

“This is the first documented instance of large- scale automated/bot interference in a formal public health consultation,” the researchers write.

Interim Phase 3 results from 80 women suggest Humacyte's lab-grown blood vessel provided usable dialysis access sooner than fistulas, reducing catheter time, though independent review is still pending.

A new review paper proposes a Regenerative Resource Governance framework that integrates safety, circular economy principles, and community stewardship, illustrated through India's Chilika Lagoon.

A case-control study of 180 infertile women found no statistically significant associations between polycystic ovary syndrome, personality traits, and adherence to the Mediterranean diet.

Mice with transplanted human brain organoids improved their gait and maze abilities. The resulting cells suggest new ways to study cerebral palsy, epilepsy, and autism.

A two-stage AI framework uses a large language model and evolutionary feature selection to halve the symptom set for Traditional Chinese Medicine respiratory syndrome differentiation without a statistically detectable performance cost.

A gene-edited Monascus strain fermented for 21 days showed the strongest glucose uptake and lipid improvements in insulin-resistant liver cells, though the mechanisms remain candidate associations.

Researchers report a lightweight, imbalance-aware intrusion detection system achieving up to 0.9998 accuracy on IoT and medical IoT benchmark datasets.

Deleting the strigolactone biosynthesis gene MAX1 in lettuce produces a strong stay-green phenotype in darkness while redirecting part of the senescence transcriptional program toward a distinct stress-associated state.

A genome-wide CRISPR screen identifies UGCG-driven sphingolipid metabolism and protective autophagy as drivers of Niraparib resistance in ovarian cancer.

AI and Early Pancreatic Cancer Every scan in the study had already been read and cleared as normal.

Mayo Clinic’s Zombie Cell Studies Not all aging cells cause the same harm, and scientists now have better

arXiv:2609.33795v2 Announce Type: replace Abstract: Hydrogen-diesel dual-fuel (H2DF) combustion engines offer a promising pathway for decarbonising hard-to-electrify transport sectors, yet their highly nonlinear dynamics and coupled process variables demand constraint-aware control strategies. Model Predictive Control (MPC) meets these requirements but requires an online optimisation in every combustion cycle, which limits deployment on low-cost embedded hardware. This paper trains a feedforward deep neural network (DNN) by behaviour cloning (BC) to imitate an MPC expert, using 86,000 engine cycles of demonstration data collected at 1500 min-1 on a modified Cummins QSB 4.5-litre hydrogen dual-fuel engine. Two variants, one with process feedback and one without, are validated experimentally. Both track unseen fast-transient load steps (3-8 bar indicated mean effective pressure, IMEP) with normalised root mean square error (NRMSE) values of 7.80% and 9.03% against the expert's 8.01%,

arXiv:2610.37533v1 Announce Type: new Abstract: Failure diagnosers for robot manipulation are increasingly vision--language models (VLMs), trained and evaluated on single-arm failures produced by perturbing a successful demonstration. Bimanual tool use adds failures of other kinds: a tool that slips out of its socket while still held, two correct gestures made at different times, and attempts that fail for two reasons at once. On a simulated bimanual tool-use task whose failure causes are read from the recorded trace, we measure what each sensing channel lets a diagnoser tell. The tool's slip is read on the tool hand; which hand was late is read on both arms, while a sensor on one hand reads its own hand's delay and at most part of the other's; and whether a late attempt was also pushed too weakly is not identifiable from passive sensing when the tool hand is late. Within two attempts, a repair policy that escalates when the cause is not identifiable repairs as many failures as one

arXiv:2610.36852v1 Announce Type: new Abstract: Robot foundation models may generalize across tasks and embodiments, but their actions still depend on sensors that can fail. When the policy is frozen or inaccessible, can an external monitor diagnose and repair corrupted inputs without access to its internals? We test a structured surrogate that separates periodic gait from sensor-dependent corrections to guide input repair around a frozen locomotion policy. On simulated Go2, offline diagnosis raises walking success from 259/640 to 429/640 episodes; online repair reaches 382/640, with no triggers in 64 held-out fault-free episodes. A matched boosted-tree surrogate predicts clean actions similarly well but yields no detected repair improvement over no repair. Separate fits improve walking on Go1 and ANYmal C, although phase-aware gradients perform better. Evidence remains limited to one policy and a partly screened simulated fault suite; robot foundation model and hardware validation

arXiv:2610.36247v1 Announce Type: new Abstract: Offline-to-online reinforcement learning uses offline pretraining to accelerate subsequent online learning. While many methods assume the offline dataset remains available throughout fine-tuning, there are many settings where this is not the case. Prior work attributes failures in this setting primarily to critic recalibration under distribution shift. We identify a complementary actor-side failure: once the offline data disappears, continued behavioral cloning (BC) no longer preserves the offline behavior distribution and instead refits the behavioral model to the agent's limited, and nonstationary, online experience. We demonstrate that stopping online behavioral fitting while preserving the pretrained expressive prior substantially improves no-retention fine-tuning across implicit flow Q-learning (IFQL), flow Q-learning (FQL), and action-chunked RL (QC) on seven OGBench domains. We trace this failure to a behavioral feedback loop:

A new study identifies DPP4-positive zone 2 hepatocytes as mechanosensory cells that trigger liver regeneration through PIEZO1-mediated mechanotransduction.

Two LAMP-CRISPR diagnostic pipelines for Mycobacterium tuberculosis reached a limit of detection of 10 copies per microlitre, with the Cas12a version outperforming the Cas13a pipeline on clinical samples.

A narrative review finds AI and digital pathology show potential to improve adult CNS tumor diagnostics, while generalizability, transparency, and regulation remain barriers.

A Rhode Island grand jury indicted seven people tied to a Providence health agency over an alleged $2 million Medicaid fraud scheme involving AI-generated patient records and backdated training documents.

A new review finds stem cell and extracellular vesicle therapies show encouraging preclinical results for Alzheimer's disease, but published clinical trials have so far tested only safety and feasibility.

A zebrafish study shows that the ATP receptor P2RY11 sustains macrophage mitochondrial function through the SIRT1–PGC-1α axis and is required for fin regeneration.

A new review finds AI excels within single bone marrow modalities but genuine multimodal fusion systems remain unrealized.

A single-surgeon prospective study found that acetabular-only robot-assisted total hip arthroplasty was associated with a lower surgeon stress index and lower perceived workload than conventional surgery despite longer operative time.

A six-year internal evaluation of the fully virtual EDIT AI program reports 284 high school participants, improved self-reported familiarity with biomedical AI concepts, and 39 peer-reviewed publications.

A study in BMC Medical Genomics finds that miR-183-5p is upregulated in the serum of adolescent idiopathic scoliosis patients, shows diagnostic potential, and inhibits bone-forming differentiation of human bone marrow mesenchymal stem cells by negatively regulating GNG12.

A randomized trial found that injecting cefuroxime into bone before robotic-assisted knee replacement achieved far higher tissue concentrations than intravenous dosing, with 750 milligrams identified as a candidate dose for further evaluation.

University of Missouri researchers used the AI-designed, open-source gene editor OpenCRISPR-1 to create rice plants that resist bacterial blight, with commercial varieties still years away.

A new study combines LSTM-based deep learning with LIME and SHAP explainability to estimate EV battery state of health across an IoT–Fog–Cloud 6G architecture.

A retrospective single-center study of 163 Asian kidney transplant recipients found basiliximab induction offered no significant short-term benefit over no induction in one-year rejection, graft survival, function, or infection risk.

Researchers studying Bolotana, Sardinia, found that women had exceptionally high estimated late-life survival during 2021–2025, with their probability of surviving from age 70 to 100 exceeding regional and national averages by more than threefold. Earlier demographic estimates showed a much smaller survival advantage, prompting the authors to classify Bolotana as a candidate longevity blue zone that requires further investigation.

LANCASTER, PA — This spring, Lorraine and Mark Martin, a Mennonite family with a tax business, adopted a stubby-legged cow named Royal Lite. Lorraine, Mark and Royal Lite all have one thing in common: They ...

University of Missouri researchers have created rice plants that resist bacterial blight, a devastating disease that threatens one of the world's most important food crops. The advance could help farmers in parts of Asia and Africa, where the disease can destroy harvests and threaten livelihoods.

A four-tier wearable health monitoring framework combining multimodal deep learning, federated learning, differential privacy and blockchain auditability has reported strong benchmark results in preclinical simulation.

A qualitative interview study of hospital staff identified six ethical themes in AI-assisted clinical practice and led researchers to propose a lifecycle-based, multi-actor governance framework that has not yet been tested.

A systematic review and meta-analysis of 43 studies reports pooled diagnostic performance of artificial intelligence models for dengue, chikungunya, and Zika, while cautioning that most studies were at high risk of bias.

A retrospective international analysis found male metastatic urothelial carcinoma patients survived longer on enfortumab vedotin than females, with higher NECTIN-4 expression in male tumors.

Today, the National Comprehensive Cancer Network (NCCN)-an alliance of leading cancer centers in the United States responsible for guidelines that patients and providers rely on worldwide to improve outcomes-published its first-ever NCCN Guidelines for Patients focused specifically on how to support loved ones with cancer.

Hormone replacement therapy, or HRT, may lead to longer, more frequent migraines, according to research presented at The Menopause Society’s annual meeting.“Migraines can have a significant impact on quality of life, particularly during the menopause transition when many women experience changes in their symptoms,” Flavia Fairbanks Ruano, MD, PhD, lead author from the University of Miami Miller School of Medicine, said in a press release. “More research is needed to better understand the relationship between hormone therapy and migraine and to help guide individualized treatment decisions.”

A three-year review at a major US referral center found that 96% of completed uterine transplant applications were excluded on initial screening, with medical contraindications, age, and BMI the leading reasons.

A new teacher-student framework aggregates knowledge from multiple medical imaging datasets and consistently improves segmentation, classification, and detection performance across MRI, CT, and X-ray benchmarks.

TUESDAY, Oct. 27, 2026 — Is that brown spot a sign of aging, or is it skin cancer?A new AI tool promises to help doctors suss out skin cancers, researchers reported Oct. 22 in the journal Mayo Clinic Proceedings.The tool, called Dermoscopy...

A new Transplant International special issue shows that older donors and older recipients can achieve strong transplant outcomes when cold ischemia time, infection surveillance and machine perfusion are managed carefully.

Large language models (LLMs), such as ChatGPT and Claude, are increasingly being used to provide health information to the public. Although this is beneficial for improving access to treatment information, their limitations were revealed by a new study that found LLMs were biased in their treatment choices and that using these outputs in practice could drive up treatment costs.

AI training offers doctors a lucrative new side gig. And like the tech itself, it puts them at an intersection of money and responsibility for the future of healthcare.

Microsoft's MAI-DxO achieved higher diagnostic accuracy than generalist physicians in a peer-reviewed study, while selecting tests with lower estimated costs under controlled simulation conditions.

A pH- and ROS-responsive hydrogel carrying CRISPR-Cas9 silenced the TRPV4 calcium channel in rat discs, restoring autophagic flux and slowing degeneration over eight weeks.

NEW ORLEANS — In this video, Jennifer I. Lim, MD, discusses phase 2 data on OCU410 presented at the American Academy of Ophthalmology meeting.The trial evaluated OCU410 (Ocugen) subretinal gene therapy in patients with geographic atrophy. The one-time injection reduced the rate of ellipsoid zone loss by 27% and the rate of retinal pigment epithelium loss by 45%.Lim called OCU410 a “promising potential new treatment” and noted that the phase 3 ArMaDa study of the therapy is ongoing.Reference:

A study combining retrospective human DXA data and mouse experiments reports that statins promote bone formation in estrogen deficiency by acting on LEPR+ skeletal stem and progenitor cells through a RAC1- and NF-κB-dependent pathway that converges on osteolectin.

A three-dimensional co-culture of human blastoids and endometrial assembloids reveals reciprocal transcriptional adaptations, including SDC1-positive MMP7-positive trophoblast subsets linked to stromal CSF3 signalling, at the onset of implantation-like interaction.

A new review in the Journal of Translational Medicine examines how cancer stem cells, vascularization and immune evasion contribute to therapeutic resistance in central nervous system tumors.

A platinum catalyst on nitrogen-doped carbon with isolated iron sites resists oxidation and nitrogen poisoning, sustaining electrochemical ammonia cracking for hydrogen production for 500 hours at 1.5 A cm−2.

Senescent cells accumulate with age to drive the onset and progression of age-related disease and loss of function. Over the past decade, a broad range of research and development efforts have emerged to attempt to address this cause of aging. Senostatic therapies reduce the pace at which senescent cells are created. Senolytic therapies selectively destroy senescent cells. Senomorphic therapies suppress aspects of the harmful behavior of senescent cells, principally the pro-inflammatory signaling that contributes to the chronic inflammation of aging and is disruptive to tissue structure and function. Researchers here describe the discovery of a senomorphic microRNA in the context of an attempt to reduce the burden of senescent cells generated by cancer therapy. Many forms of cancer treatment leave the patient with increased numbers […]

A new review proposes measuring robotic bronchoscopy by clinically actionable yield — whether one procedure delivers diagnosis, molecular profiling and staging sufficient to inform treatment decisions within 30 days.

Egypt Healthcare Authority (EHA) is probing  co-operation with South Korea in medical technology, digital transformation, artificial intelligence and smart hospitals, with a view to transferring Korean expertise and technology to Egypt. EHA Chairman Ahmed el-Sobky held talks on Tuesday with President of the Korea Health Industry Development Institute (KHIDI), Ko Sang-baek, on the sidelines of […] The post Egypt, S. Korea probe co-operation in medical technology, AI appeared first on Egyptian Gazette.

A network meta-analysis of sixteen randomized trials finds brain-computer interface-triggered functional electrical stimulation shows a clearer motor recovery signal after stroke than robotic control, though the comparison between the two relies on indirect evidence.

arXiv:2610.35678v1 Announce Type: new Abstract: In observational survival studies, delayed treatment initiation can induce immortal time bias in effect estimation when treatment groups are defined by future uptake. Despite originating from distinct predictive or causal inference paradigms, time-dependent approaches (e.g. the time-dependent Cox model, landmark supermodel, multi-state model, and g-formula) are often treated as interchangeable in applied research when addressing immortal time bias. These models effectively reclassify person-time according to time-varying exposure, yet the interpretation of their target estimands depends on assumptions about the underlying disease process and clinically relevant time scales. We study effect estimation for allogeneic hematopoietic stem cell transplantation under baseline intention-to-transplant strategies in acute myeloid leukemia (i.e. transplantation is planned for a prespecified time, subject to the patient's eligibility), within a

arXiv:2610.35508v1 Announce Type: new Abstract: Single-cell CRISPR screens run in many cell types are used to compare how cell types respond to perturbation, to train models that transfer effects between them and to score such models. These uses assume that cells carrying a guide RNA had their target knocked down. Yet cells grow for days before sequencing: when an essential gene is knocked down, knocked-down cells are depleted or stop dividing while cells that escaped knockdown keep growing, and the measured response, an average over all cells carrying the guide, is diluted. We define the escape fraction of a perturbation as the share of its measured cells whose target was not effectively knocked down, and estimate it from each cell's counts of the targeted gene with a two-group mixture that recovers known fractions in simulated mixtures. In most cell types the escape fraction rises with how essential the gene is, it is higher in a later screen of the same cells, and it reaches about

arXiv:2610.33003v1 Announce Type: new Abstract: Given the challenges of pharmaceutical development, there is increasing interest in computational approaches for identifying novel drug targets. Transcriptomics technologies enable the analysis of disease mechanisms at unprecedented resolution, while spatially resolved platforms further preserve the spatial organization of cells within tissues, allowing analysis to focus on specific regions of disease localization. To derive actionable therapeutic targets from these data, it is necessary to identify mechanisms that recur across a population rather than those observed in individual patients, which requires integrated analysis across disease conditions and donor cohorts. Here, we present the groundwork for an AI co-scientist framework for the sequential analysis of multi-batch spatial transcriptomics data that integrates interpretable mathematical models, including optimal transport, tensor factorizations, and topological data analysis,

arXiv:2610.35294v1 Announce Type: new Abstract: While climate change disproportionately impacts African public health, AI predictive modeling remains heavily constrained. Using a Synthetic Knowledge Synthesis approach and VOSviewer, we conducted a bibliometric analysis of Scopus-indexed literature, comparing Global and African corpora. Network analysis reveals stark divergences: global research emphasizes complex, predictive models for planetary health, whereas African research focuses on localized, reactive solutions for immediate infectious disease triage. Severe structural barriers in Africa, including compute deficits, data silos, and energy instability, hinder advanced AI scaling. Ultimately, international funding must shift from siloed pilot projects toward building foundational digital infrastructure and localized AI frameworks.

arXiv:2610.34408v1 Announce Type: new Abstract: While recent advances in text-to-speech (TTS) synthesis have substantially improved speech quality and modeling flexibility, the development of lightweight and computationally efficient systems remains comparatively underexplored. We present LeanTTS, a bilingual complete synthesis stack that supports two complementary regimes: reference-prompt zero-shot cloning for unseen speakers and token-conditioned customization for recurring voices. LeanTTS combines a compact Mandarin polyphone front end, a structurally compressed flow-matching acoustic model, and a lightweight Vocos decoder with four-step MeanFlow generation. The resulting system contains 18.9M learned parameters, 86.12% fewer than the 136.2M-parameter ZipVoice-Vocos pipeline. On Seed-TTS-Eval, its zero-shot mode obtains 2.02 CER/62.3 SIM on Mandarin and 2.32 WER/48.3 SIM on English, exposing the quality-size trade-off under aggressive compression. With token-conditioned customized

arXiv:2610.33931v1 Announce Type: new Abstract: "Why is my period late?" Sometimes, the most useful response begins with another question. Yet conversational AI systems may respond to incomplete health queries without seeking essential context or grounding their guidance in verifiable evidence. This paper introduces HerHealthGPT, an evidence-grounded conversational AI system designed to address these challenges in women's reproductive health. HerHealthGPT integrates retrieval-augmented generation (RAG), context-aware clarification, and source validation within an interactive web interface. Rather than responding immediately to incomplete queries, the system maintains conversational context, asks targeted follow-up questions, checks for premature conclusions, and grounds its responses in retrieved evidence. We evaluate the AI components using 540 English messages derived from 90 reproductive-health cases and conduct a prototype questionnaire study with 30 adults. Evidence retrieval

arXiv:2610.33355v1 Announce Type: new Abstract: Molecular cloning is one of the most routine wet-lab workflows, yet existing evaluations of language models in biology mostly test knowledge recall or isolated steps rather than end-to-end experimental planning. We introduce BenchBench-Cloning, a benchmark for large language models comprising 65 molecular cloning tasks grounded in sequence records, laboratory notebooks, and published experimental studies. Planning a molecular cloning experiment requires defining a set of constructs needed for a scientific objective and a construction route to produce them. BenchBench-Cloning contains 50 protocol-development tasks, in which agents produce an executable protocol for a specified construct, and 15 construct-design tasks, in which agents determine the constructs a scientific objective calls for. Protocol-development tasks are evaluated by execution in a cloning simulator; construct-design tasks are evaluated with task-specific deterministic

arXiv:2610.33198v1 Announce Type: new Abstract: A neural classifier can fail completely inside a layered defense without any end-to-end number moving. We demonstrate this on BARBICAN, a fail-closed gateway for language-model tool mediation whose arbitration stage spent an entire recorded evaluation receiving text mangled by a misconfigured tokenizer. Correcting the misconfiguration lifts that stage's corpus-wide area under the ROC curve (AUC) from 0.392 to 0.855 and pushes system recall down, from 80.67% to 80.34%; a stub that answers "malicious" scores 80.77%, higher than either. The reason emerges once the stage's input distribution is examined: upstream routing hands it a band that is 96.82% malicious, and the enforcement design lets an incriminating verdict block while an exculpating one releases an input only through one narrow branch. Yet deleting it costs 673 detections. Substituting its verdict with a constant instead moves between 12 and 140 of 25,000 decisions. Transplanted

A student-led study at an osteopathic medical college found that three paid AI platforms varied widely in reference quality and osteopathic content when queried on pharmacology topics.

A gut microbial bile acid, isoallolithocholic acid, prolonged survival and reduced rejection injury in a mouse liver transplant model by activating the NR4A1/FOXP3 signaling axis to shift intrahepatic Th1/Treg balance toward tolerance.

A dual-stream AI network that jointly predicts lithium-ion battery temperature and state of health achieved a 92.6 percent timely warning rate with an average lead time of 48.6 cycles across evaluated test cases.

A single-center cohort study of 752 kidney transplant recipients found that invasive fungal infections, most often Pneumocystis pneumonia, were associated with sharply elevated risks of death or graft failure, particularly within six months of transplantation and the first 90 days after infection.

A new review maps four molecular routes by which CRISPR systems can detect cancer-linked DNA methylation in cell-free DNA, and finds that none has yet reached independent clinical validation.

A randomised phase 1 clinical trial conducted by scientists and doctors from Shanghai and Guangzhou has achieved encouraging results in the treatment of Type 1 diabetes using stem cells extracted from the mouth’s gum tissue. More than half of those who received the treatment showed varying degrees of improvement, while one subject was able to discontinue exogenous insulin therapy completely, the researchers said. Results of the study, which tested a therapy based on gingival mesenchymal stem...

A large integrated transcriptomic study shows that rising cancer stemness tracks prostate cancer progression, lineage plasticity, therapy resistance, and poor survival.

A case report describes a 52-year-old man with dilated cardiomyopathy who developed a rare inferolateral left ventricular aneurysm within one month and required heart transplantation.

Laboratory experiments with human tonsil-derived fibroblasts show that senescent lymphoid stromal cells suppress B cell proliferation through contact-dependent mechanisms despite normal B cell activation.

A new narrative review finds that medical vision–language AI models show strong benchmark results but suffer from inconsistent evaluation reporting, hallucination risks, and narrow training data.