Communications medicine最新文献

{"title":"Subgrouping patients with ischemic heart disease by means of the Markov cluster algorithm.","authors":"Amalie D Haue, Peter C Holm, Karina Banasik, Kenny Emil Aunstrup, Christian Holm Johansen, Agnete T Lundgaard, Victorine P Muse, Timo Röder, David Westergaard, Piotr J Chmura, Alex H Christensen, Peter E Weeke, Erik Sørensen, Ole B V Pedersen, Sisse R Ostrowski, Kasper K Iversen, Lars V Køber, Henrik Ullum, Henning Bundgaard, Søren Brunak","doi":"10.1038/s43856-025-01077-1","DOIUrl":"10.1038/s43856-025-01077-1","url":null,"abstract":"<p><strong>Background: </strong>Ischemic heart disease (IHD) is heterogeneous with respect to onset, burden of symptoms, and disease progression. We hypothesized that unsupervised clustering analysis could facilitate identification of distinct and clinically relevant multimorbidity clusters.</p><p><strong>Methods: </strong>We included IHD patients who underwent coronary angiography (CAG) or coronary computed tomography angiography (CCTA) between 2004 and 2016 and used the earliest procedure as the index date. Patient health records were obtained from the Danish National Patient Registry, the Danish National Prescription Registry, and two in-hospital laboratory database systems. Genetic data were obtained from the Copenhagen Hospital Biobank. Using registered pre-index diagnosis codes (n = 3046), patients were clustered by application of the Markov Cluster algorithm. Multimorbidity clusters were then characterized using Cox regressions (new ischemic events, non-IHD mortality, and all-cause mortality) and enrichment analysis to explore both risks and phenotypical characteristics.</p><p><strong>Results: </strong>In a cohort of 72,249 patients with IHD (mean age 63.9 years, 63.1% males), 31 distinct clusters (C1-31, 67,136 patients) are identified. Comparing each cluster to the 30 others, seven clusters (9,590 patients) have significantly higher or lower risk of new ischemic events (five and two clusters, respectively). A total of 18 clusters (35,982 patients) have higher or lower risk of death from non-IHD causes (12 and six clusters, respectively), and 23 clusters have a statistically significant higher or lower risk for all-cause mortality. Cardiovascular or inflammatory diseases are commonly enriched in clusters (13). Distributions for 24 laboratory test results differ significantly across clusters. Polygenic risk scores are increased in a total of 15 clusters (48.4%).</p><p><strong>Conclusions: </strong>Based on prior disease profiles, unsupervised clustering robustly stratify patients with IHD in subgroups with similar clinical features and outcomes.</p>","PeriodicalId":72646,"journal":{"name":"Communications medicine","volume":"5 1","pages":"372"},"PeriodicalIF":5.4,"publicationDate":"2025-08-26","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC12381225/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144980822","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Molecular profiles of the great obstetrical syndromes reveal common features and dynamic changes in early pregnancy.","authors":"Xiaojing Zeng, Yuexin Gan, Jian Zhao, Lin Zhang, Qianlong Zhang, Lisong Shen, Xipeng Wang, Kun Sun, Jun Zhang","doi":"10.1038/s43856-025-01103-2","DOIUrl":"10.1038/s43856-025-01103-2","url":null,"abstract":"<p><strong>Background: </strong>The great obstetrical syndromes (GOS) encompass a spectrum of pregnancy-related complications that are the major determinants of maternal and neonatal mortality. Although shared placental pathology underlies these complications, knowledge of the molecular commonalities before clinical presentation remains incomplete.</p><p><strong>Methods: </strong>In this case-control study nested within the prospective Shanghai Birth Cohort, we investigated the molecular characteristics underlying GOS conditions in the pre-symptomatic phase using serum samples from 203 nulliparas who subsequently developed GOS and 181 controls. A multi-omics approach combining protein, metabolite, and gene analysis was employed.</p><p><strong>Results: </strong>Here we show a shared molecular background of the spectrum of all tested GOS conditions at the co-expression module, molecule, pathway, multi-omics network and genome levels. In early pregnancy, one protein module mainly involved in the immune system and platelet function shows significant associations across the spectrum of tested GOS conditions. Different conditions share several hub proteins and enriched pathways in the innate immune system and platelet activation, signaling and aggregation. Common molecular changes could be observed before 14 weeks of gestation. Different groups of hub proteins demonstrate the potential to differentiate between normal and complicated pregnancies before and after 14 weeks of gestation.</p><p><strong>Conclusions: </strong>We highlight the shared molecular signatures among different GOS conditions in the pre-symptomatic phase, suggesting the potential of a common screening and intervention strategy. Our results further support the notion that the prevention of GOS should start at 14 weeks of gestation or earlier, when the molecular signature changes have already emerged.</p>","PeriodicalId":72646,"journal":{"name":"Communications medicine","volume":"5 1","pages":"369"},"PeriodicalIF":5.4,"publicationDate":"2025-08-25","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC12378939/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144980801","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Predictive computational framework to provide a digital twin for personalized cardiovascular medicine.","authors":"Mengzhe Lyu, Ryo Torii, Ce Liang, Xuehuan Zhang, Xifu Wang, Qiaoqiao Li, Yiannis Ventikos, Duanduan Chen","doi":"10.1038/s43856-025-01055-7","DOIUrl":"10.1038/s43856-025-01055-7","url":null,"abstract":"<p><strong>Background: </strong>In percutaneous coronary intervention (PCI), the ability to predict post-PCI fractional flow reserve (FFR) and stented vessel informs procedural planning. However, highly precise and effective methods to quantitatively simulate coronary intervention are lacking. This study developed and validated a virtual coronary intervention (VCI) technique for non-invasive physiological and anatomical assessment of PCI.</p><p><strong>Methods: </strong>In this study, patients with substantial lesions (pre-PCI CT-FFR of less than 0.80) were enrolled. VCI framework was used to predict vessel reshape and post-PCI CT-FFR. The accuracy of predicted post-VCI CT-FFR, luminal cross-sectional area (CSA) and centreline curvature was validated with post-PCI computed tomography (CT) angiography datasets.</p><p><strong>Results: </strong>Overall, 30 patients are initially screened; 21 meet the inclusion criteria, and 9 patients (9 vessels) are included in the final analysis. The average PCI-simulation time is 24.92 ± 1.00 s on a single processor. The calculated post-PCI CT-FFR is 0.92 ± 0.09, whereas the predicted post-VCI CT-FFR is 0.90 ± 0.08 (mean difference: -0.02 ± 0.05 FFR units; limits of agreement: -0.08 to 0.05). Morphologically, the predicted CSA is 16.36 ± 4.41 mm² and the post-CSA is 17.91 ± 4.84 mm² (mean difference: -1.55 ± 1.89 mm²; limits of agreement: -5.22 to 2.12). The predicted centreline curvature across the stented segment (including ~2 mm proximal and distal margins) is 0.15 ± 0.04 mm⁻¹, while the post-PCI centreline curvature is 0.17 ± 0.03 mm⁻¹ (mean difference: -0.02 ± 0.06 mm⁻¹; limits of agreement: -0.12 to 0.09).</p><p><strong>Conclusions: </strong>The proposed VCI technique achieves non-invasive pre-procedural anatomical and physiological assessment of coronary intervention. The proposed model has the potential to optimize PCI pre-procedural planning and improve the safety and efficiency of PCI.</p>","PeriodicalId":72646,"journal":{"name":"Communications medicine","volume":"5 1","pages":"370"},"PeriodicalIF":5.4,"publicationDate":"2025-08-25","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC12379278/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144980778","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Forecasting the diabetic retinopathy progression using generative adversarial networks.","authors":"Huiyu Qiao, Feilong Tang, Huanfen Zhou, Yun Cai, Kairou Guo, Jin Wang, Tong Ma, Lie Ju, Wei Feng, Zhiqiang Ma, Juan Chen, Yuan Luo, Bin Wang, Zongyuan Ge, Qiansu Yang","doi":"10.1038/s43856-025-01092-2","DOIUrl":"10.1038/s43856-025-01092-2","url":null,"abstract":"<p><strong>Background: </strong>Diabetic retinopathy (DR) is the leading cause of blindness worldwide, making early prediction of DR progression crucial for effectively preventing visual loss. This study introduces a prediction framework DRForecastGAN (Diabetic Retinopathy Forecast Generative Adversarial Network), and investigates its clinical value in predicting DR development.</p><p><strong>Methods: </strong>DRForecastGAN model, consisting of a generator, discriminator, and registration network, was trained, validated, and tested in training (12,852 images), internal validation (2734 images), and external test (8523 images) datasets. A pre-trained ResNet50 classification model identified the DR severity on synthetic images. The performance of the proposed DRForecastGAN model was compared with the CycleGAN and Pix2Pix models in image reality and DR severity of the synthesized fundus images by calculating Fréchet Inception Distance (FID), Peak Signal-to-Noise Ratio (PSNR), Structural Similarity Index (SSIM), and area under the curve (AUC).</p><p><strong>Results: </strong>DRForecastGAN model has the lowest FID, highest PSNR and highest SSIM on internal validation (FID: 27.3 vs. 32.8 vs. 34.4; PSNR: 25.3 vs. 17.0 vs. 16.9; SSIM: 0.93 vs. 0.79 vs. 0.65) and external test (FID: 37.6 vs.45.1 vs.48.4; PSNR: 20.7 vs.15.2 vs.14.7; SSIM: 0.86 vs.0.69 vs.0.63) datasets compared with Pix2Pix and CycleGAN models. In the prediction of DR severity, our DRForecastGAN model outperforms both Pix2Pix and CycleGAN models, achieving the highest AUC values on both internal validation (0.87 vs. 0.76 vs. 0.75) and external test (0.85 vs. 0.70 vs. 0.69) datasets.</p><p><strong>Conclusions: </strong>The proposed DRForecastGAN model can effectively visualize DR development by synthesizing future fundus images, offering potential utility for both treatment and ongoing monitoring of DR.</p>","PeriodicalId":72646,"journal":{"name":"Communications medicine","volume":"5 1","pages":"368"},"PeriodicalIF":5.4,"publicationDate":"2025-08-23","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC12375066/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144980914","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Using the PET/CT radiotracer [⁶⁸Ga]Ga-DOTA-mDesmo to target V1b receptors and localize corticotropinoma in Cushing's disease.","authors":"Somit Pandey, Rama Walia, Gurvinder Kaur, Kumud Pandav, Imran Rather, Nivedita Rana, Sushant Sahoo, Bhagwant Rai Mittal, Jaya Shukla","doi":"10.1038/s43856-025-01033-z","DOIUrl":"10.1038/s43856-025-01033-z","url":null,"abstract":"<p><strong>Background: </strong>Cushing's disease, the most common cause of Cushing's syndrome, is driven by pituitary tumors (corticotropinoma) and characterized by the overexpression of CRH R1 and V1b receptors. Accurate detection of these tumors remains challenging. This study aims to develop and evaluate a corticotropinoma specific radiopharmaceutical, [⁶⁸Ga]Ga-DOTA-mDesmo, that targets the V1b receptor for both anatomical and functional identification of corticotropinomas.</p><p><strong>Methods: </strong>Molecular docking was used to validate the binding affinity of [⁶⁸Ga]Ga-DOTA-mDesmo to the V1b receptor. Radiolabeling was optimized with gallium-68, followed by quality controls and physicochemical characterization. ACTH-release assay was performed using primary-cultured corticotropinoma cells. Receptor specificity was confirmed via radioimmunoassay using recombinant human V1b receptor. Ex vivo biodistribution studies were performed in healthy male Wistar rats at 30-, 60-, and 120-min post-injection (8 ± 1.1 MBq).</p><p><strong>Results: </strong>Here we show that [⁶⁸Ga]Ga-DOTA-mDesmo binds effectively to the V1b receptor, with a binding energy of -13.98 kcal/mol and the key interacting residues of V1b are GLN301, SER304, ASN309, and ASP314. Radiolabeling achieves high yield (~96%) and purity (>99%), with human serum stability for up to 4 h. In vitro studies confirm that DOTA-mDesmo acts as an agonist in corticotropinoma cells. Excess cold DOTA-mDesmo results in a 50% blocking in binding. Biodistribution in rats indicates renal clearance, with high %ID/g in the kidneys (7.37 ± 0.58) and urinary bladder (4.32 ± 0.48), and negligible uptake in the pituitary gland, our organ of interest.</p><p><strong>Conclusions: </strong>These finding support [⁶⁸Ga]Ga-DOTA-mDesmo as a promising radiotracer for non-invasive, receptor-targeted PET/CT imaging of corticotropinomas in patients with Cushing's disease.</p>","PeriodicalId":72646,"journal":{"name":"Communications medicine","volume":"5 1","pages":"367"},"PeriodicalIF":5.4,"publicationDate":"2025-08-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC12373759/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144980839","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"The International Health Cohorts Consortium (IHCC) advances population health research and genomic discovery.","authors":"John J Connolly, Scott Sundseth, Grant M Wood, Chisom Nwaneri, Geoffrey Ginsburg, Philip Awadalla, Michele Ramsay, Thomas Keane, Peter Goodhand, Adam S Butterworth, Hakon Hakonarson","doi":"10.1038/s43856-025-01026-y","DOIUrl":"10.1038/s43856-025-01026-y","url":null,"abstract":"","PeriodicalId":72646,"journal":{"name":"Communications medicine","volume":"5 1","pages":"366"},"PeriodicalIF":5.4,"publicationDate":"2025-08-21","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC12371067/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144980818","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Liquid biopsy diagnostics for non-small cell lung cancer via elucidation of tRNA signatures.","authors":"Zhuokun Feng, Masaki Nasu, Gehan Devendra, Ayman A Abdul-Ghani, Owen T M Chan, Jeffrey A Borgia, Zitong Gao, Hanqiu Zhang, Yu Chen, Ting Gong, Gang Luo, Hua Yang, Lang Wu, Yuanyuan Fu, Youping Deng","doi":"10.1038/s43856-025-01068-2","DOIUrl":"10.1038/s43856-025-01068-2","url":null,"abstract":"<p><strong>Background: </strong>Lung cancer, particularly non-small cell lung cancer (NSCLC), accounts for about 85% of all lung cancer cases and remains a major global health challenge. Traditional diagnostic methods, such as chest X-rays and low-dose CT scans, have limitations, including high false-positive rates, radiation risks, and the invasiveness of tissue biopsies. This study aims to develop a non-invasive liquid biopsy approach for early NSCLC diagnosis.</p><p><strong>Methods: </strong>We developed a machine-learning model to analyze small RNA sequencing data from 1446 tissue samples to identify a diagnostic tRNA signature. This signature was independently validated using the in-house data of 233 plasma exosome samples. Diagnostic performance was assessed using Area Under the Curve (AUC) metrics. Signature tRNAs were then evaluated across various clinical and demographic variables, with further survival analysis and functional studies to explore the molecular role of the signature tRNAs.</p><p><strong>Results: </strong>We identify a robust six-tRNA signature with strong diagnostic performance, achieving AUC values of 0.97 in discovery, 0.96 in hold-out validation, and 0.84 in independent validation. The signature effectively distinguishes cancerous from benign samples (AUC = 0.85) and consistently performs across clinical and demographic variables, with AUC values exceeding 0.80, particularly for early-stage lung cancer diagnosis. Additionally, three signature tRNAs demonstrate prognostic value for independent survival prediction. Functional studies suggest potential regulatory roles of specific tRNAs and their associated fragments in tumor metabolism pathways.</p><p><strong>Conclusions: </strong>This research underscores the diagnostic power of tRNA signature for NSCLC liquid biopsy and provides epigenetic insights that enhance our understanding of oncogenic molecular pathophysiology.</p>","PeriodicalId":72646,"journal":{"name":"Communications medicine","volume":"5 1","pages":"364"},"PeriodicalIF":5.4,"publicationDate":"2025-08-21","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC12370967/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144980663","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Associations between monthly temperature anomalies and mortality among children younger than five years in low- and middle-income countries.","authors":"Pengfei Li, Jingyi Wu, Tao Xue, Tong Zhu","doi":"10.1038/s43856-025-01101-4","DOIUrl":"10.1038/s43856-025-01101-4","url":null,"abstract":"<p><strong>Background: </strong>This study examined associations between anomalous temperatures and under-five mortality (U5M) in low- and middle-income countries (LMICs).</p><p><strong>Methods: </strong>Between 1998 and 2019, data were collected on 1,745,132 live births across 56 LMICs. The median age was 27.0 months (interquartile range: 12.0, 43.0), and 51.0% were male. Monthly temperature anomalies were calculated as deviations from the long-term average temperature for the same calendar months. We developed a two-dimensional exposure-response function (ERF) based on a varying-coefficient model to examine the relationship between U5M and both long-term average temperatures and anomalous temperatures.</p><p><strong>Results: </strong>Here we show, in the multicenter model, each 1 °C increase in positive temperature anomalies or decrease in negative temperature anomalies is associated with an 8% (95% confidence interval [CI]: 6%, 11%) or 2% (95% CI: -1%, 4%) excess risk of U5M, respectively. The two-dimensional ERF reveals considerable heterogeneity in the impact of anomalous temperatures on U5M across different temperature zones. Positive temperature anomalies pose a significant hazard to children in warm and hot zones, whereas negative anomalies are primarily hazardous in cold zones but appear protective in hot zones. The sibling-matched model demonstrates nonlinear and heterogeneous effect estimates consistent with the multicenter model findings.</p><p><strong>Conclusions: </strong>Anomalous temperatures of both heat and cold are associated with an increased risk of child mortality in LMICs, with the effect varying by temperature zone, reflecting the adaptive capacity of populations to local climates. There is an urgent need to develop regionally adaptive strategies to protect child health amid global climate change.</p>","PeriodicalId":72646,"journal":{"name":"Communications medicine","volume":"5 1","pages":"365"},"PeriodicalIF":5.4,"publicationDate":"2025-08-21","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC12370998/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144980887","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Leveraging sequences missing from the human genome to diagnose cancer.","authors":"Ilias Georgakopoulos-Soares, Ofer Yizhar-Barnea, Ioannis Mouratidis, Candace S Y Chan, Michail Patsakis, Akshatha Nayak, Rachael Bradley, Mayank Mahajan, Jasmine Sims, Dianne Laboy Cintron, Ryder Easterlin, Julia S Kim, Emmalyn Chen, Geovanni Pineda, Guillermo E Parada, John S Witte, Christopher A Maher, Felix Feng, Ioannis Vathiotis, Nikolaos Syrigos, Emmanouil Panagiotou, Andriani Charpidou, Konstantinos Syrigos, Jocelyn Chapman, Mark Kvale, Martin Hemberg, Nadav Ahituv","doi":"10.1038/s43856-025-01067-3","DOIUrl":"10.1038/s43856-025-01067-3","url":null,"abstract":"<p><strong>Background: </strong>Cancer diagnosis using cell-free DNA (cfDNA) has the potential to improve treatment and survival but has several technical limitations.</p><p><strong>Methods: </strong>In this study, we developed a prediction model based on neomers, DNA sequences 13-17 nucleotides in length that are predominantly absent from the genomes of healthy individuals and are created by tumor-associated mutations.</p><p><strong>Results: </strong>We show that neomer-based classifiers can accurately detect cancer, including early stages, and distinguish subtypes and features. Analysis of 2577 cancer genomes from 21 cancer types shows that neomers can distinguish tumor types with higher accuracy than state-of-the-art methods. Generation and analysis of 465 cfDNA whole-genome sequences demonstrates that neomers can precisely detect lung and ovarian cancer, including early stages, with an area under the curve ranging from 0.89 to 0.94. By testing various promoters or over 9000 candidate enhancer sequences with massively parallel reporter assays, we show that neomers can identify cancer-associated mutations that alter regulatory activity.</p><p><strong>Conclusions: </strong>Combined, our results identify a sensitive, specific, and simple cancer diagnostic tool that can also identify cancer-associated mutations in gene regulatory elements.</p>","PeriodicalId":72646,"journal":{"name":"Communications medicine","volume":"5 1","pages":"363"},"PeriodicalIF":5.4,"publicationDate":"2025-08-21","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC12371106/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144980703","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Computer vision detects covert voluntary facial movements in unresponsive brain injury patients.","authors":"Xi Cheng, Sujith Swarna, Jermaine Robertson, Nathaniel A Cleri, Jordan R Saadon, Chiemeka Uwakwe, Yindong Hua, Seyed Morsal Mosallami Aghili, Cassie Wang, Robert S Kleyner, Xuwen Zheng, Ariana Forohar, John Servider, Kurt Butler, Chao Chen, Jordane Dimidschstein, Petar M Djurić, Charles B Mikell, Sima Mofakham","doi":"10.1038/s43856-025-01042-y","DOIUrl":"10.1038/s43856-025-01042-y","url":null,"abstract":"<p><strong>Background: </strong>Many brain injury patients who appear unresponsive retain subtle, purposeful motor behaviors, signaling capacity for recovery. We hypothesized that low-amplitude movements precede larger-amplitude voluntary movements detectable by clinicians after acute brain injury. To test this hypothesis, we developed a novel, as far as we are aware, computer vision-based tool (SeeMe) that detects and quantifies low-amplitude facial movements in response to auditory commands.</p><p><strong>Methods: </strong>We enrolled 16 healthy volunteers and 37 comatose acute brain injury patients (Glasgow Coma Scale ≤8) aged 18-85 with no prior neurological diagnoses. We measured facial movements to command assessed using SeeMe and compared them to clinicians' exams. The primary outcome was the detection of facial movement in response to auditory commands. To assess comprehension, we tested whether movements were specific to command type (i.e., eye-opening to open your eyes and not stick out your tongue) with a machine learning-based classifier.</p><p><strong>Results: </strong>Here we show that SeeMe detects eye-opening in comatose patients 4.1 days earlier than clinicians. SeeMe also detects eye-opening in more comatose patients (30/36, 85.7%) than clinical examination (25/36, 71.4%). In patients without an obscuring endotracheal tube, SeeMe detects mouth movements in 16/17 (94.1%) patients. The amplitude and number of SeeMe-detected responses correlate with clinical outcome at discharge. Using our classifier, eye-opening is specific (81%) to the command open your eyes.</p><p><strong>Conclusion: </strong>Acute brain injury patients have low-amplitude movements before overt movements. Thus, many covertly conscious patients may have motor behavior currently undetected by clinicians.</p>","PeriodicalId":72646,"journal":{"name":"Communications medicine","volume":"5 1","pages":"361"},"PeriodicalIF":5.4,"publicationDate":"2025-08-20","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC12368019/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144980855","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}