Although hypoxia plays a central role in the pathobiology of cardiovascular and cerebrovascular diseases, controlled hypoxic stimuli, such as intermittent hypoxia conditioning (IHC), can trigger beneficial adaptations that counteract those diseases. IHC has been studied as a readily administered, non-pharmacological intervention for almost a century, it has not yet been widely adopted in clinical practice. Growing evidence demonstrates the beneficial preventive and therapeutic effects of IHC on the cardiovascular system, and although less conclusively established, on the brain. Moreover, IHC is highly feasible and raises few safety concerns. A notable limitation, similar to that of remote ischemic conditioning, is the lack of refined IHC protocols for personalized, disease-specific applications in the individual patient. Significant progress has been achieved in addressing this limitation. For example, increasingly sophisticated modern hypoxia equipment permits customization of IHC parameters according to each individual’s responses to and tolerance of hypoxia. Still to be evaluated are the persistence of IHC’s effects and the capacity of IHC booster protocols to renew its benefits. Comparative analyses of the mechanisms and efficacy of IHC and remote ischemic conditioning could aid in defining their respective applications and possible synergistic effects. This perspective article summarizes the historical development of IHC, presents a selection of preclinical and clinical studies exemplifying recent research progress, and discusses strategies to address the lingering concerns limiting clinical IHC applications.
Background Cisplatin is a widely used chemotherapeutic agent, but its clinical application is limited by serious ototoxic side effects, including hearing loss, tinnitus, and vertigo. Inflammation is now recognized as a significant contributor to cisplatin-induced hearing loss, but the upstream regulators remain largely undefined. This study aimed to identify novel regulators of cisplatin-induced ototoxicity through CRISPR-Cas9 screening and to determine the role of growth arrest and DNA damage-inducible alpha (GADD45A) in cochlear inflammation and hearing loss. Methods To identify upstream regulators of cisplatin-induced ototoxicity, a CRISPR-Cas9-based loss-of-function screen was first performed in OC1 cells. Following the identification of Gadd45a as a candidate gene, its effects on cisplatin-induced cytotoxicity were evaluated using cell viability assays, flow cytometry, and TUNEL staining. Integrated transcriptomic and proteomic analyses were subsequently conducted to elucidate the downstream molecular mechanisms. Finally, the protective role of Gadd45a inhibition was validated in vivo through siRNA delivery into the posterior semicircular canal and in Gadd45a conditional knockout (cKO) mice. Results Gadd45a knockout significantly inhibited cisplatin-induced cell death in the OC1 cell line while simultaneously enhancing autophagy signaling. Transcriptomic profiling revealed a marked downregulation of the expression of C-X-C motif chemokine ligand (CXCL) family cytokines following Gadd45a loss, which was further supported by proteomic analysis of culture supernatants. Mechanistically, Gadd45a knockout specifically inhibited nuclear factor κB subunit 1 (NF-κB1) but not RELA proto-oncogene, NF-κB subunit (RELA). Immunofluorescence staining demonstrated that Gadd45a knockout inhibited the nuclear translocation of the NF-κB1 protein, and subsequent studies revealed that the NF-κB1 protein was degraded through the lysosomal pathway. In vivo, both si-Gadd45a injection into the semicircular canal and the knockout of Gadd45a effectively attenuated cisplatin-induced ototoxicity. Conclusions This study identifies GADD45A as a key regulator of cisplatin ototoxicity and links the balance between apoptosis and autophagy with NF-κB1-CXCL-mediated inflammation. Targeted inhibition of GADD45A signaling may represent a promising therapeutic strategy to prevent hearing loss in patients receiving cisplatin chemotherapy.
Venous leg ulcers (VLUs) result from a multifactorial interplay of etiological factors, posing significant challenges in clinical management. Effective treatment requires a comprehensive assessment to identify the underlying pathophysiological mechanisms, followed by the implementation of an individualized, multidisciplinary approach that integrates the expertise of vascular surgeons, wound care specialists, and other pertinent healthcare professionals. Although both domestic and international vascular surgery societies, as well as other relevant specialties, have issued consensus statements on the diagnosis and treatment of VLUs or clinical guidelines for managing chronic lower extremity venous diseases, there remains a notable gap in systematic guidance specifically focused on wound repair for VLUs. In particular, standardized protocols for surgical wound repair are either lacking or insufficiently detailed in these documents. With the establishment and advancement of wound repair as a specialized field in China, Chinese experts have accumulated substantial clinical experience in the diagnosis and management of VLU-related wound healing. To standardize the surgical repair of VLUs and improve both therapeutic outcomes and patients' quality of life, the Wound Repair Professional Committee of the Chinese Medical Doctor Association convened a panel of multidisciplinary experts for extensive deliberations. Based on a synthesis of current international and domestic evidence and clinical practice, the committee reached consensus on evidence-based recommendations for key issues in VLU wound management, including debridement techniques, indications, timing, and approaches to wound repair for VLUs. This consensus is intended to serve as a practical reference for clinicians involved in the care of patients with VLUs.
Background:Severe bacterial sepsis caused by Klebsiella pneumoniae (KP) is characterized by dysregulated inflammation, multiorgan injury, and high mortality, yet the key molecular drivers of this process remain incompletely understood. This study aimed to elucidate how the Z-DNA binding protein 1 (ZBP1) regulates inflammatory injury during KP-induced sepsis and to identify potential therapeutic targets with translational relevance. Methods:To investigate the role of macrophage ZBP1 in sepsis, we combined transcriptomic datasets from human sepsis cohorts and murine sepsis and KP-infection models with single-cell RNA sequencing of infected tissues. We further established both global and myeloid-specific Zbp1 knockout (Zbp1 fl/fl Lyz2-Cre +/- ) mice, as well as myeloid-specific Lgmn (encoded Legumain) knockout (Lgmn fl/fl Lyz2-Cre +/- , Lgmn CKO ) mice, to delineate the macrophage-dependent mechanisms of immune regulation. In addition, macrophage-targeted adeno-associated virus 9 (AAV9) vectors driven by the F4/80 promoter were used to overexpress Lgmn in vivo. Flow cytometry, immunofluorescence, and survival analyses were performed to evaluate systemic inflammation, organ injury, and sepsis severity. The effect of recombinant LGMN supplementation on macrophage activation and barrier restoration was also assessed. Results:ZBP1 activation was driven by both type I interferon (IFN-I) signaling and mitochondrial damage. Mechanistically, ZBP1 directly interacted with signal transducer and activator of transcription 3 (STAT3), suppressed its phosphorylation and nuclear translocation, and consequently inhibited transcription of Lgmn, a gene associated with anti-inflammatory macrophage polarization and tissue repair. Genetic and myeloid-specific deletion of Zbp1 resulted in reduced proinflammatory cytokine production and improved survival in both KP- and cecal ligation and puncture (CLP)-induced sepsis models. In contrast, Lgmn CKO aggravated systemic inflammation, organ injury, and barrier dysfunction, whereas macrophage-targeted AAV9-mediated Lgmn overexpression or recombinant LGMN supplementation alleviated macrophage inflammation, restored epithelial function, and improved survival. Conclusion:These findings reveal a previously unrecognized ZBP1-STAT3-LGMN signaling axis that contributes to immune and inflammatory dysregulation in KP-induced sepsis and suggest that targeting ZBP1 or restoring LGMN activity may represent a promising therapeutic strategy for severe bacterial infections.
Background:Epilepsy is a major public health challenge affecting individuals of all ages, especially in low- and middle-income countries (LMICs). Reliable prevalence projections are critical for healthcare planning and resource allocation. This study aimed to forecast the prevalence of epilepsy and its trends in LMICs by age, sex, year, and income level by 2050. Methods:Using data from the Global Burden of Disease Study (GBD) 2023, we projected the prevalence and number of idiopathic and secondary epilepsy cases in LMICs from 2024 to 2050. We developed a hybrid deep neural network (DNN)-Transformer framework that integrates Poisson regression and Autoregressive Integrated Moving Average (ARIMA) models for prevalence projection. Decomposition analysis was applied to quantify the contributions of population growth, aging, and prevalence change to the increase in epilepsy cases. Dementia-attributable epilepsy was independently projected to address secondary causes not included in GBD 2023. Results:By 2050, the age-standardized prevalence rate (ASPR) of epilepsy in LMICs was projected to reach 907.22 per 100,000 [95% uncertainty interval (UI) 731.01-1083.56], a 33.32% increase from 2023, with cases rising to 72.04 million (95% UI 57.86-86.23), a 58.68% increase. The ASPRs of idiopathic and secondary epilepsy were estimated at 323.11 and 584.10 per 100,000 in 2050, respectively, with the increase in secondary epilepsy being more than 7-fold that of idiopathic epilepsy since 2023. The ASPR of secondary epilepsy due to neonatal disorders was projected to rise by 65.76%. Model validation demonstrated good predictive performance (root mean squared error <0.001). From 2023 to 2050, the increases in idiopathic and secondary epilepsy cases were forecast to be highest in low-income countries (LICs; 76.12% and 241.50%, respectively), with growth declining as income levels increased. Population growth (21.40%) primarily drove the increase in idiopathic epilepsy cases, whereas changes in prevalence (59.89%) predominantly drove the rise in secondary epilepsy cases. Dementia-attributable secondary epilepsy was projected to reach 3.40 million cases by 2050. Conclusions:We forecast a continuous increase in the prevalence and number of epilepsy cases in LMICs through 2050, with secondary epilepsy increasing more rapidly than idiopathic epilepsy. LICs may exhibit the greatest increases over the next three decades, necessitating targeted interventions and further investigation.
Background:Hypertension affects around one billion adults worldwide, with abnormal glucose metabolism and vascular smooth muscle cell (VSMC) phenotype switch playing crucial roles in its pathogenesis. Pyruvate kinase M2 (PKM2) is a key glycolytic enzyme, but its regulation and roles in VSMC phenotype switch and hypertension are unknown. Using the Gene Importance Calculator (GIC) to predict gene essentiality, we identified ribosomal modification protein rimK-like family member A (RIMKLA) as a highly relevant gene and explored its regulatory contributions to hypertension. Methods:Internal mammary arteries from patients with hypertension and normotension, as well as arteries from angiotensin II (Ang II)-induced hypertensive mice, salt-sensitive hypertensive Dahl/SS rats, and spontaneously hypertensive rats, were analyzed in this study. Adenoviruses and adeno-associated viruses were used for ex vivo and in vivo gene overexpression. VSMC-specific RIMKLA or PKM2 knockout mice were generated using the Cre-Loxp system. Arterial tension was measured by wire myography, and blood pressure was assessed by the tail-cuff method and remote radio-telemetry. Protein-protein interactions were determined by co-immunoprecipitation with mass spectrometry. Non-targeted metabolomics, in vitro phosphorylation, adenosine triphosphate (ATP), reactive oxygen species (ROS), and cytoplasmic calcium assays were performed to identify the signaling axis involved. Results:RIMKLA expression was increased in the medial layer of the internal mammary arteries of patients with hypertension, as well as in hypertensive rat and mouse arteries. RIMKLA overexpression in the mesenteric arteries of Sprague-Dawley rats significantly increased vessel contractility. VSMC-specific RIMKLA overexpression increased arterial contractility and blood pressure in mice. VSMC-specific RIMKLA deletion attenuated Ang II-induced hypertension in mice. Mechanistically, we identified RIMKLA as a scaffold protein that binds to protein-tyrosine phosphatase 1B (PTP1B) and PKM2. RIMKLA phosphorylated PTP1B at tyrosine (Tyr)66, leading to PKM2 dephosphorylation and activation at Tyr105. Once activated, PKM2 enhanced glucose metabolism, increased ROS production, and boosted ATP secretion, driving VSMC phenotype switch. RIMKLA-induced vasoconstriction and hypertension were reversed by VSMC-specific PKM2 deletion. Additionally, a PKM2 inhibitor reduced arterial contractility and blood pressure. Conclusions:RIMKLA functioned as a scaffold protein kinase, recruiting both PTP1B and PKM2, and orchestrating PTP1B phosphorylation and its subsequent recruitment to activate PKM2. These findings position RIMKLA as a key regulator of PKM2 activation, promoting VSMC phenotype switch and contributing to hypertension.
Background:Osteoarthritis (OA) is currently the most common age-related degenerative joint disease, but there remains a lack of disease-modifying therapy for OA treatment. This study aimed to elucidate the critical roles and underlying mechanism of the fat mass and obesity associated gene (FTO) in OA pathogenesis induced by high fat diet (HFD) and HFD coupled with destabilized medial meniscus (DMM) surgery, and to ascertain the synergistic action of the two in OA pathogenesis. Methods:Mouse models were established using solely HFD feeding and combined HFD feeding and DMM surgery. Primary mouse chondrocytes were treated with palmitic acid (PA) and interleukin-1β (IL-1β) to simulate lipotoxicity and inflammation. FTO expression was modulated via genetic, adenoviral, or pharmacological methods. Downstream pyruvate dehydrogenase phosphatases 2 (PDP2) and YTH N6-methyladenosine RNA binding protein 2 (YTHDF2) were similarly manipulated by siRNA or overexpression plasmids; adenovirus-mediated knockdown/overexpression of Pdp2 was applied in vivo. Mechanistic studies included RNA sequencing (RNA-Seq) and methylated RNA immunoprecipitation sequencing (MeRIP-Seq). Results:The results revealed reduced FTO expression in the cartilage of both obese OA patients and mouse models. Genetically, adenovirus-induced or pharmacologically-induced FTO inhibition exacerbated OA progression in HFD-fed and HFD+DMM mice. Mechanistically, Fto knockdown downregulated the PDP2 level in an m6A-dependent manner via YTHDF2. Pdp2 knockdown exacerbated OA progression in HFD-fed and HFD+DMM mice, whereas PDP2 overexpression markedly alleviated cartilage degeneration. Moreover, YTHDF2 overexpression reversed the role of Fto knockdown in lipid deposition and cartilage degeneration. Conclusions:Our study identifies that downregulation of FTO exerts a pivotal effect on OA with obesity. FTO drives disease progression by regulating PDP2 activity, and YTHDF2 mediates the m6A modification of FTO to PDP2. Targeting the FTO/YTHDF2/PDP2 axis offers promising therapeutic potential for OA treatment.
Medical large language models (Med-LLMs) have shown considerable promise across a broad range of clinical tasks, including decision support, medical documentation, patient communication, multimodal analysis, and telemedicine. Their rapid development has generated growing interest in how large language models (LLMs) may support healthcare practice, while also raising important questions about reliability, clinical validity, and safe deployment. This review provides a structured overview of recent progress in Med-LLMs by examining their major application areas, key challenges, and emerging future directions. Current evidence shows that the clinical usefulness of Med-LLMs cannot be judged by model performance alone. Their value in practice depends on whether they are supported by reliable evidence, remain consistent with current medical knowledge, and can be integrated into clinical workflows. Important challenges remain in evaluation, safety, knowledge updating, and real-world deployment. These issues reflect a gap between performance in controlled settings and clinical practice. Future progress will require stronger clinical validation, better alignment with medical practice, and more careful deployment across different settings. The clinical impact of Med-LLMs will depend on whether they can be used as reliable tools in clinical care.
Background:Spinal cord injury (SCI) triggers persistent neuroinflammation, gliosis, neuronal loss, and demyelination, leading to motor deficits and neuropathic pain (NeP). Botulinum neurotoxin type A (BoNT/A) has shown anti-inflammatory and neuroprotective effects in acute SCI, but its potential in the chronic phase remains unclear. This study investigates whether combining BoNT/A with electrical muscle stimulation (EMS) enhances recovery in chronic SCI. Methods:Adult mice with severe thoracic SCI (paraplegic) underwent EMS (30 min/d for 10 non-consecutive days starting 3 d post-injury) or no stimulation. Fifteen days after SCI, animals received a single intrathecal injection of BoNT/A (15 pg/5 μl) or saline. Functional recovery was assessed up to 60 d as well as in moderate and mild SCI mice. NeP onset and maintenance were evaluated. Spinal cord tissue was analysed for astrocytic and microglial morphology, neuronal and oligodendroglial survival, myelin protein expression, and in vitro effects on oligodendrocyte precursor cells (OPCs). The phenotype of hindlimb muscles was evaluated through morphological and gene expression analyses. Results:EMS was able to counteract muscle atrophy and fibrosis, and when combined with BoNT/A, also denervation. Moreover, the combination restored hindlimb motor function in chronic SCI, whereas BoNT/A or EMS alone were ineffective. NeP, a common comorbidity associated with SCI, was mitigated by BoNT/A treatment even when administered in the chronic phase. BoNT/A reduced astrocytic hypertrophy and excitatory synapse association and was associated with a morphology-based redistribution of microglial profiles toward a resting-like classification, decreased apoptosis, and increased neuronal and oligodendroglial survival. Myelin basic protein (MBP) expression was significantly elevated in vivo. In vitro, BoNT/A promoted OPC differentiation into myelinating oligodendrocytes, increased process complexity, and upregulated MBP, galactocerebroside C, proteolipid protein, and myelin oligodendrocyte glycoprotein under both proliferative and differentiating conditions. Cleaved synaptosomal-associated protein 25 colocalization with OPC confirmed direct BoNT/A internalization and activity. Conclusions:BoNT/A exerts neuroprotective effects in chronic SCI by reducing neuroinflammation and supporting neuronal and oligodendroglial preservation. When combined with EMS, it also promotes remyelination and improves muscle homeostasis, suggesting that early stimulation creates a permissive environment for recovery. These findings support the clinical evaluation of BoNT/A as a therapeutic strategy for chronic SCI.
Background:Simulated microgravity, modeled by head-down tilt (HDT), induces cephalad fluid shifts that perturb intracranial hemodynamics and may affect cognitive function. However, the temporal adaptation of cerebral arterial blood flow (CaBF), both during simulated microgravity and throughout the recovery phase, remains incompletely understood. Methods:In this study, 38 healthy male participants underwent a 7-day -6° HDT protocol followed by a 5-day recovery phase. Four-dimensional flow magnetic resonance imaging (4D flow MRI) was performed at 8 time points [baseline, HDT 12 h, HDT 1 d, HDT 3 d, HDT 7 d, recovery (R) 1 d, R 3 d, and R 5 d] to quantify CaBF and total cerebral blood inflow (TCBI) in the basilar artery (BA), left and right internal carotid arteries (ICAL and ICAR), and left and right middle cerebral arteries (MCAL and MCAR). Systemic vitals and fasting cortisol/renin were collected, and a computerized reaching task assessed reaction time (RT), movement time (MT), and peak velocity (PV). Time effects were tested with repeated-measures analysis of variance (RM ANOVA) or the Friedman test. Predictors of ≥10% TCBI decrease during HDT and ≥10% TCBI increase during the recovery phase were assessed using logistic regression, and flow-behavior associations were examined using Spearman correlation. Results:No significant vessel lumen area changes were found after post-hoc analysis, despite an overall difference observed in the MCAR (χ²=17.40, P=0.015). However, average blood flow significantly changed in the ICAL (χ²=34.16, P<0.001), MCAL (χ²=73.11, P<0.001), and MCAR (χ²=49.02, P<0.001), while BA was stable (RM ANOVA F=0.787, P=0.599) and ICAR showed no significant pairwise effects despite an overall difference (χ²=16.35, P=0.022). TCBI progressively declined during HDT and rebounded rapidly at the onset of recovery (P<0.001). Logistic regression identified systolic blood pressure (SBP) as an independent predictor of a ≥10% TCBI reduction during HDT [P=0.044, odds ratio (OR)=3.004, 95% confidence interval (CI) 1.028-8.777], and baseline cortisol levels predicted significant TCBI decreases from baseline to HDT 7 d (P=0.047, OR=1.306, 95% CI 1.004-1.699). Cognitive-motor testing further revealed phase-dependent changes, with RT and MT generally shortening, most consistently in the no-beep condition, while PV remained stable with beep but increased without beep. Apart from an exploratory negative correlation between TCBI rebound and cued PV (r=-0.360, P=0.031), TCBI changes were largely decoupled from behavioral outcomes. Conclusions:This study demonstrates vessel-specific, lateralized adaptation of cerebral arterial inflow during 7 days of -6° HDT and 5 days of recovery, with anterior circulation more responsive to posture-induced fluid shifts and TCBI gradually decreasing then rapidly rebounding after re-ambulation. Interindividual TCBI susceptibility reflects blood pressure and endocrine status, while cognitive-motor changes remain weakly coupled, underscoring the importance of incorporating early-recovery assessments into HDT studies to better characterize cerebrovascular readaptation after re-ambulation. Clinical Trials Registry:ChiCTR2500096128.
Fine particulate matter (PM2.5) is a prevalent environmental pollutant that has been well established as a contributor to morbidity on a global scale. An increasing body of scientific evidence suggests that PM2.5 promotes various modes of regulated cell death (RCD), including apoptosis, necroptosis, pyroptosis, PANoptosis, ferroptosis, and autophagy-dependent cell death, as well as potentially cuproptosis, across a broad spectrum of human diseases. These interconnected pathways demonstrate how exposure to pollutants induces oxidative stress, mitochondrial dysfunction, endoplasmic reticulum stress (ERS), and inflammatory responses. Alterations in lysosomal permeability serve as a critical link, connecting environmental pollutants to health conditions. Importantly, dysregulation of RCD mechanisms is associated with exposure to PM2.5 and numerous health disorders, such as cardiovascular diseases (CVDs), neurological conditions, respiratory illnesses, renal and hepatic dysfunctions, reproductive health issues, and ocular diseases. Collectively, RCD functions as a central molecular framework through which PM2.5 exposure accelerates the progression of pathological conditions. This review synthesizes recent mechanistic insights, identifies promising therapeutic candidates, and highlights critical knowledge gaps. It offers a strategic framework to guide future research endeavors aimed at mitigating the impact of PM2.5-induced RCD in human diseases. Additionally, we will identify and explore the research gaps that need to be addressed to effectively translate preclinical PM2.5 studies into clinical trials. We will also provide a comprehensive overview that both highlights these gaps and offers pathways to bridge them.
Bone is a multifunctional organ essential for structural support, protection, hematopoiesis, and endocrine regulation. Emerging research increasingly highlights the pivotal role of the nervous system in bone metabolism, repair, and musculoskeletal disease progression, particularly the autonomic nervous system (ANS), which precisely regulates bone tissue cells via sympathetic and parasympathetic pathways. The ANS regulates the activity of mesenchymal stem cells (MSCs), osteoblasts, osteoclasts, chondrocytes, and nucleus pulposus cells (NPCs) via neurotransmitters such as norepinephrine (NE), acetylcholine (ACh), neuropeptide Y (NPY), and vasoactive intestinal peptide (VIP), thereby modulating core pathological processes in bone-related disorders including osteoporosis (OP), osteoarthritis (OA), intervertebral disc degeneration (IVDD), and traumatic fractures. Targeted neuromodulation is emerging as a promising therapeutic strategy for musculoskeletal disorders. This review aims to summarize the regulatory roles of the ANS in bone cell metabolism and bone-related pathologies and to evaluate current ANS-bone axis-based interventions to promote skeletal health and alleviate the burden of musculoskeletal diseases.
Brain tumors, including primary intracranial tumors and brain metastases (BrMs), represent a major threat to human health and are associated with extremely poor prognoses. The brain tumor microenvironment (BTME) is a complex ecosystem composed of various elements, including tumor cells, immune cells, neurons, the vascular system, the extracellular matrix, and cytokines. These elements not only coexist spatially but are also functionally connected, interacting through intricate networks that collectively influence tumor initiation, progression, and therapeutic efficacy. In recent years, the application of novel technologies such as single-cell and spatial transcriptomics has uncovered complex cellular heterogeneity and spatial organization within the BTME. This review comprehensively summarizes the key components and functions of the BTME in tumor development and therapy, the mechanisms underlying intercellular interactions, with an emphasis on their clinical potential and challenges.
Background:Laparoscopic gastrectomy is technically demanding and errors often arise from limited situational awareness. Artificial intelligence (AI)-based computer vision may enhance intraoperative recognition of instruments and anatomy, but evidence for generalizable, real-time deployment remains limited. Methods:We conducted a multicenter study using laparoscopic gastrectomy videos from 4 tertiary hospitals. Data were split by institution into training, internal test, and external validation sets. A You Only Look Once version 10 (YOLOv10)-based model was developed to recognize 18 categories (11 instruments and 7 organs) and integrated into a lightweight user interface for bedside use. Mean average precision at Intersection over Union (IoU) 0.50 (mAP@50) and AP@[0.50-0.95] for boxes and masks, recall, and F1 score were calculated. The feasibility of live operating-room deployment was assessed. Surgeon cognitive workload was assessed using the National Aeronautics and Space Administration Task Load Index (NASA-TLX), and system usability was evaluated with the System Usability Scale (SUS). Results:The training set comprised 4308 frames (41,066 labels); the internal testing set, 1803 frames (17,211 labels); and the external validation set, 721 frames (6685 labels). Training performance reached box precision 0.896, recall 0.891, mAP@50 0.937, and AP@[0.50-0.95] 0.846; mask AP@[0.50-0.95] was 0.809. On the internal testing set, box AP@[0.50-0.95] was 0.670 and mask AP@[0.50-0.95] 0.642. External validation achieved box precision 0.798, recall 0.773, F1 score 0.785, mAP@50 0.832, and AP@[0.50-0.95] 0.684; mask mAP@50 was 0.836 and AP@[0.50-0.95] 0.658. High-performing classes included stapler, ultrasonic scalpel, and forceps, whereas omentum and pancreas were relatively lower. The system operated in real time during laparoscopic gastrectomy on a laptop-grade GPU, providing on-demand overlays without workflow disruption. The NASA-TLX score was 26.72±6.51, indicating low cognitive burden, and the SUS score was 78.33±5.20, demonstrating good usability. Conclusions:We present development, external validation, and live deployment of a real-time AI navigation system for laparoscopic gastrectomy across multiple centers. The model demonstrated robust detection and segmentation of instruments and organs and was feasible for bedside use. Prospective trials are warranted to evaluate effects on intraoperative safety, efficiency, and training.
Background:A cancer diagnosis is usually associated with a substantial loss of life years. However, few studies have quantified life expectancy (LE) and years of life lost (YLL) among cancer patients, particularly in China. This study aims to estimate LE and YLL by cancer type, sex, age at diagnosis, and attained age (i.e., the age a cancer survivor has reached at a given time) for cancer patients in China. Methods:This is a comparative assessment based on population-based cancer registration and death surveillance in China. Data on all-cause deaths, cancer cases, and relative survival in 2021 were obtained from publicly available reports released by the National Cancer Center and the Chinese Center for Disease Control and Prevention. Life tables for the general population were constructed using age-specific all-cause mortality rates, whereas relative survival rates and parameterized long-term excess hazard functions were used to construct life tables for cancer patients. We used the standard period life table method to estimate LE of both the general population and cancer patients, assuming that mortality probabilities remained constant over time. The YLL was the LE difference between cancer patients and the sex- and age-matched general population. Results:For cancer patients diagnosed at the median age of 64 years, the LE is 8.8 years [95% confidence interval (CI) 8.7-9.0], corresponding to a YLL of 9.2 years (95% CI 9.0-9.3). Male patients have an LE and YLL of 6.7 years (95% CI 6.6-6.8) and 9.8 years (95% CI 9.7-9.9), respectively, while females have 11.2 years (95% CI 11.1-11.4) and 8.4 years (95% CI 8.3-8.6). Thyroid cancer had the highest LE [31.9 years (95% CI 31.5-32.0)], while pancreatic cancer had the lowest [2.6 years (95% CI 2.4-2.8)]. Female patients had better LEs than males across 20 cancer types that affect both sexes. Younger patients experienced greater YLL, except in cases of thyroid cancer, oropharyngeal cancer, ovarian cancer, and male bladder cancer. LE gradually increased in patients who survived the first three years; thereafter, the decline depended on attained age, approaching that of the general population. Conclusions:LE of cancer patients in China varied by sex, cancer type, age at diagnosis, and attained age. Cancer patients, healthcare providers, and policymakers may incorporate these estimates into their decision-making.
