Login
Treatment with quercetin inhibits SARS-CoV-2 N protein-induced acute kidney injury by blocking Smad3-dependent G1 cell-cycle arrest
(Molecular Therapy 31, 344β361; February 2023)
Reading view
In vivo-directed evolution identifies AAV-WM04 as a next-generation vector for potent and sustained hearing restoration in DFNB9
AAV-WM04, an AAV vector identified through in-vivo-directed screening in the adult cochlea, enables highly efficient and selective inner hair cell transduction. Dual-AAV delivery of OTOF using AAV-WM04 restores hearing in a DFNA9 deafness mouse model at low doses, highlighting its translational potential for gene therapy.
Show-Harness: Just a VLM Agent Can Play Robots
arXiv:2609.10522v1 Announce Type: cross
Abstract: Foundation vision-language models (VLMs) exhibit broad intelligence about the world, yet translating this intelligence into robot control remains challenging. We present Show-Harness, an Embodied Harness that enables VLMs to "play" robots through a compact semantic interface linking intent to action. Show-Harness exposes discrete semantic action units that VLMs can naturally reason over, while embodiment-specific interpreters deterministically ground them into local robot actions, keeping the VLM directly responsible for fine-grained physical decisions. Through the same interface, Show-Harness demonstrates the feasibility of (1) directly unlocking closed-source frontier VLMs for zero-shot robot control, and (2) adapting small-scale open-source VLMs for low-cost deployment with just a few GPU-hours of fine-tuning. We further develop GUMI (GUI Manipulation Interface), which extends the same semantic action space to GUI-based demonstration collection, allowing humans and agents to "play" robots across embodiments without specialized teleoperation hardware. Extensive experiments show that Show-Harness-equipped VLM agents generalize robustly across tasks, embodiments, and environments, outperforming representative agentic and VLA paradigms. These results suggest that the right interface can unlock substantial embodied capability from foundation VLMs, without requiring additional model capacity or costly embodiment-specific pretraining.
Artificial intelligence in respiratory medicine: From diagnosis to treatment and future directions
Chin Med J Pulm Crit Care Med. 2026 Jun 6;4(2):99-116. doi: 10.1016/j.pccm.2026.05.005. eCollection 2026 Jun.
ABSTRACT
Lung diseases-including lung cancer, chronic obstructive pulmonary disease (COPD), asthma, interstitial lung diseases (ILDs), and rare conditions like cystic fibrosis-remain major drivers of global morbidity and mortality. Timely diagnosis and individualized treatment are frequently challenged by heterogeneous clinical phenotypes and the complexity of multimodal data. This review provides a critical synthesis of the transformative role of artificial intelligence (AI) in respiratory care, tracing the paradigm shift from classical machine learning to emerging large language models (LLMs) and multimodal foundation models. We evaluate the performance of AI across the patient care continuum: beginning with radiologist-level nodule detection and automated diagnostics, advancing into AI-powered clinical decision support systems (CDSS) and surgical/radiotherapeutic interventions, and culminating in prognostic modeling and "digital twin" simulations for longitudinal patient management. Furthermore, we explore the translational frontier of precision medicine, examining how AI leverages multi-omics and liquid biopsies to drive novel biomarker discovery and accelerate drug repurposing. Finally, we address persistent sociotechnical barriers-including data sovereignty, legal liability, and the critical need for prospective clinical validation-proposing a translational roadmap for the safe integration of generalist medical AI into clinical workflows.
PMID:42396189 | PMC:PMC13323542 | DOI:10.1016/j.pccm.2026.05.005
