Hasil untuk "Medicine"

Jia Li, Yu Hou, Rui Zhang

Automatically discovering personalized sequential events from large-scale time-series data is crucial for enabling precision medicine in clinical research, yet it remains a formidable challenge even for contemporary AI models. For example, while transformers capture rich associations, they are mostly agnostic to event timing and ordering, thereby bypassing potential causal reasoning. Intuitively, we need a method capable of evaluating the "degree of alignment" among patient-specific trajectories and identifying their shared patterns, i.e., the significant events in a consistent sequence. This necessitates treating timing as a true \emph{computable} dimension, allowing models to assign ``relative timestamps'' to candidate events beyond their observed physical times. In this work, we introduce LITT, a novel Timing-Transformer architecture that enables temporary alignment of sequential events on a virtual ``relative timeline'', thereby enabling \emph{event-timing-focused attention} and personalized interpretations of clinical trajectories. Its interpretability and effectiveness are validated on real-world longitudinal EHR data from 3,276 breast cancer patients to predict the onset timing of cardiotoxicity-induced heart disease. Furthermore, LITT outperforms both the benchmark and state-of-the-art survival analysis methods on public datasets, positioning it as a significant step forward for precision medicine in clinical AI.

Mark S. Keller, Nicholas Lucarelli, Yijiang Chen et al.

Application of machine learning techniques enables segmentation of functional tissue units in histology whole-slide images (WSIs). We built a pipeline to apply previously validated segmentation models of kidney structures and extract quantitative features from these structures. Such quantitative analysis also requires qualitative inspection of results for quality control, exploration, and communication. We extend the Vitessce web-based visualization tool to enable visualization of segmentations of multiple types of functional tissue units, such as, glomeruli, tubules, arteries/arterioles in the kidney. Moreover, we propose a standard representation for files containing multiple segmentation bitmasks, which we define polymorphically, such that existing formats including OME-TIFF, OME-NGFF, AnnData, MuData, and SpatialData can be used. We demonstrate that these methods enable researchers and the broader public to interactively explore datasets containing multiple segmented entities and associated features, including for exploration of renal morphometry of biopsies from the Kidney Precision Medicine Project (KPMP) and the Human Biomolecular Atlas Program (HuBMAP).

Yilun Zhang, Dexing Kong

Large Language Models (LLMs) show promise in medicine but are prone to factual and logical errors, which is unacceptable in this high-stakes field. To address this, we introduce the "Haibu Mathematical-Medical Intelligent Agent" (MMIA), an LLM-driven architecture that ensures reliability through a formally verifiable reasoning process. MMIA recursively breaks down complex medical tasks into atomic, evidence-based steps. This entire reasoning chain is then automatically audited for logical coherence and evidence traceability, similar to theorem proving. A key innovation is MMIA's "bootstrapping" mode, which stores validated reasoning chains as "theorems." Subsequent tasks can then be efficiently solved using Retrieval-Augmented Generation (RAG), shifting from costly first-principles reasoning to a low-cost verification model. We validated MMIA across four healthcare administration domains, including DRG/DIP audits and medical insurance adjudication, using expert-validated benchmarks. Results showed MMIA achieved an error detection rate exceeding 98% with a false positive rate below 1%, significantly outperforming baseline LLMs. Furthermore, the RAG matching mode is projected to reduce average processing costs by approximately 85% as the knowledge base matures. In conclusion, MMIA's verifiable reasoning framework is a significant step toward creating trustworthy, transparent, and cost-effective AI systems, making LLM technology viable for critical applications in medicine.

Jean-Paul Ainam, Rahul, Lora Cavuoto et al.

Airway management skills are critical in emergency medicine and are typically assessed through subjective evaluation, often failing to gauge competency in real-world scenarios. This paper proposes a machine learning-based approach for assessing airway skills, specifically endotracheal intubation (ETI), using human gaze data and video recordings. The proposed system leverages an attention mechanism guided by the human gaze to enhance the recognition of successful and unsuccessful ETI procedures. Visual masks were created from gaze points to guide the model in focusing on task-relevant areas, reducing irrelevant features. An autoencoder network extracts features from the videos, while an attention module generates attention from the visual masks, and a classifier outputs a classification score. This method, the first to use human gaze for ETI, demonstrates improved accuracy and efficiency over traditional methods. The integration of human gaze data not only enhances model performance but also offers a robust, objective assessment tool for clinical skills, particularly in high-stress environments such as military settings. The results show improvements in prediction accuracy, sensitivity, and trustworthiness, highlighting the potential for this approach to improve clinical training and patient outcomes in emergency medicine.

Leonie Wagner, Jan Klocke, Philipp Enghard

Grace Y. Lam, Chuan Wen, Paul E. Ronksley et al.

Background Sex disparities in healthcare access were exacerbated during the coronavirus disease 2019 (COVID-19) pandemic. For respiratory conditions such as asthma, where females have poorer disease outcomes, it is unclear whether the pandemic further worsened this sex disparity. Thus, the objective of this study was to compare the impact of the pandemic on healthcare utilisation, exacerbations and mortality rates in males and females with asthma. Methods A retrospective population-based provincial-level analysis was conducted using linked administrative datasets from Alberta, Canada. We measured hospitalisation, emergency department and outpatient visits, and asthma outpatient exacerbations in female and males with asthma 18 months before and after 12 March 2020. Mortality data were compared pre- versus post-pandemic, taking into account confirmed COVID-19 infection within 30 days. Subgroup analysis was undertaken to determine whether healthcare utilisation differed in those with severe asthma. Results Acute care and outpatient encounters for patients with asthma declined for both females and males. Those with severe asthma of either sex experienced a reduction in hospitalisations during the pandemic. Total number of outpatient asthma visits, including both virtual and in-person, increased during the pandemic for both sexes but significantly more in females. Mortality rate was unchanged after adjusting for COVID-19-associated deaths pre- versus post-pandemic. Conclusion All patients with asthma accessed acute care resources less but outpatient visits increased during the pandemic. There was no increase in non-COVID-related mortality, regardless of sex, suggesting that the previously established sex disparity in asthma outcomes was not seen during the pandemic.

S. Manivannan, M.K. Mustafa, S. Ahmed et al.

Thushara Madathil, Praveen K. Neema

Yuting Zhang, Boyang Liu, Karina V. Bunting et al.

The echocardiographic measurement of left ventricular ejection fraction (LVEF) is fundamental to the diagnosis and classification of patients with heart failure (HF). In order to quantify LVEF automatically and accurately, this paper proposes a new pipeline method based on deep neural networks and ensemble learning. Within the pipeline, an Atrous Convolutional Neural Network (ACNN) was first trained to segment the left ventricle (LV), before employing the area-length formulation based on the ellipsoid single-plane model to calculate LVEF values. This formulation required inputs of LV area, derived from segmentation using an improved Jeffrey's method, as well as LV length, derived from a novel ensemble learning model. To further improve the pipeline's accuracy, an automated peak detection algorithm was used to identify end-diastolic and end-systolic frames, avoiding issues with human error. Subsequently, single-beat LVEF values were averaged across all cardiac cycles to obtain the final LVEF. This method was developed and internally validated in an open-source dataset containing 10,030 echocardiograms. The Pearson's correlation coefficient was 0.83 for LVEF prediction compared to expert human analysis (p<0.001), with a subsequent area under the receiver operator curve (AUROC) of 0.98 (95% confidence interval 0.97 to 0.99) for categorisation of HF with reduced ejection (HFrEF; LVEF<40%). In an external dataset with 200 echocardiograms, this method achieved an AUC of 0.90 (95% confidence interval 0.88 to 0.91) for HFrEF assessment. This study demonstrates that an automated neural network-based calculation of LVEF is comparable to expert clinicians performing time-consuming, frame-by-frame manual evaluation of cardiac systolic function.

Yifan Yang, Qiao Jin, Robert Leaman et al.

The remarkable capabilities of Large Language Models (LLMs) make them increasingly compelling for adoption in real-world healthcare applications. However, the risks associated with using LLMs in medical applications have not been systematically characterized. We propose using five key principles for safe and trustworthy medical AI: Truthfulness, Resilience, Fairness, Robustness, and Privacy, along with ten specific aspects. Under this comprehensive framework, we introduce a novel MedGuard benchmark with 1,000 expert-verified questions. Our evaluation of 11 commonly used LLMs shows that the current language models, regardless of their safety alignment mechanisms, generally perform poorly on most of our benchmarks, particularly when compared to the high performance of human physicians. Despite recent reports indicate that advanced LLMs like ChatGPT can match or even exceed human performance in various medical tasks, this study underscores a significant safety gap, highlighting the crucial need for human oversight and the implementation of AI safety guardrails.

Christina W. Zhou, Nikki L. B. Freeman, Katharine L. McGinigle et al.

Precision medicine leverages patient heterogeneity to estimate individualized treatment regimens, formalized, data-driven approaches designed to match patients with optimal treatments. In the presence of competing events, where multiple causes of failure can occur and one cause precludes others, it is crucial to assess the risk of the specific outcome of interest, such as one type of failure over another. This helps clinicians tailor interventions based on the factors driving that particular cause, leading to more precise treatment strategies. Currently, no precision medicine methods simultaneously account for both survival and competing risk endpoints. To address this gap, we develop a nonparametric individualized treatment regime estimator. Our two-phase method accounts for both overall survival from all events as well as the cumulative incidence of a main event of interest. Additionally, we introduce a multi-utility value function that incorporates both outcomes. We develop random survival and random cumulative incidence forests to construct individual survival and cumulative incidence curves. Simulation studies demonstrated that our proposed method performs well, which we applied to a cohort of peripheral artery disease patients at high risk for limb loss and mortality.

Boah Kim, Yujin Oh, Bradford J. Wood et al.

Blood vessel segmentation in medical imaging is one of the essential steps for vascular disease diagnosis and interventional planning in a broad spectrum of clinical scenarios in image-based medicine and interventional medicine. Unfortunately, manual annotation of the vessel masks is challenging and resource-intensive due to subtle branches and complex structures. To overcome this issue, this paper presents a self-supervised vessel segmentation method, dubbed the contrastive diffusion adversarial representation learning (C-DARL) model. Our model is composed of a diffusion module and a generation module that learns the distribution of multi-domain blood vessel data by generating synthetic vessel images from diffusion latent. Moreover, we employ contrastive learning through a mask-based contrastive loss so that the model can learn more realistic vessel representations. To validate the efficacy, C-DARL is trained using various vessel datasets, including coronary angiograms, abdominal digital subtraction angiograms, and retinal imaging. Experimental results confirm that our model achieves performance improvement over baseline methods with noise robustness, suggesting the effectiveness of C-DARL for vessel segmentation.

Shih-Cheng Huang, Zepeng Huo, Ethan Steinberg et al.

Synthesizing information from multiple data sources plays a crucial role in the practice of modern medicine. Current applications of artificial intelligence in medicine often focus on single-modality data due to a lack of publicly available, multimodal medical datasets. To address this limitation, we introduce INSPECT, which contains de-identified longitudinal records from a large cohort of patients at risk for pulmonary embolism (PE), along with ground truth labels for multiple outcomes. INSPECT contains data from 19,402 patients, including CT images, radiology report impression sections, and structured electronic health record (EHR) data (i.e. demographics, diagnoses, procedures, vitals, and medications). Using INSPECT, we develop and release a benchmark for evaluating several baseline modeling approaches on a variety of important PE related tasks. We evaluate image-only, EHR-only, and multimodal fusion models. Trained models and the de-identified dataset are made available for non-commercial use under a data use agreement. To the best of our knowledge, INSPECT is the largest multimodal dataset integrating 3D medical imaging and EHR for reproducible methods evaluation and research.

Deng-Ho Yang, Hsiang-Cheng Chen, James Cheng-Chung Wei

Abstract Background Gouty arthritis (GA) is a chronic systemic disease with recurrent acute monoarthritis. In a previous study, a higher incidence of acute flares was observed during the initial marked decrease in serum urate level. Our study evaluated the effect of early urate-lowering therapy in patients with acute GA flares. Methods This study included 40 patients with acute GA; of them, 20 received colchicine 0.5 mg colchicine twice daily, while 20 received probenecid 500 mg and colchicine 0.5 mg twice daily. We evaluated GA severity and laboratory data for 2 weeks after the initial therapy. Medians and interquartile ranges (IQRs) were calculated to evaluate clinical presentations between these two groups. Results Rapidly decreasing median serum uric acid levels was found in the patients treated with probenecid and colchicine compared with the patients treated with colchicine alone on day 8 (− 1.9 [IQR, − 3.7 to 0] vs 0.8 [IQR, − 0.1–2.2]; P < 0.001). However, the median decrease in visual analog scale score did not differ significantly between the two groups (− 5.5 [IQR, − 8.0 to − 3.0] vs − 3.5 [IQR, − 5.9 to − 2.0]; P = 0.080). Conclusion No significant increase was noted in acute gout flare severity or duration among GA patients treated with early aggressive control of hyperuricemia using probenecid plus colchicine.

Jun Xiong, Mengrui Wang, Jie Gao et al.

<h4>Objective</h4>General anesthesia can disturb the hormone levels in surgical patients. Hormone deficiency is one of the major symptoms of craniopharyngioma (CP) in pediatric patients. The aim of this prospective randomized controlled clinical study is to evaluate whether propofol and sevoflurane influence the perioperative hormone levels in these patients and to determine which anesthesia technique causes less impact on hormone levels.<h4>Materials</h4>Sixty-four ASA I and II pediatric patients with CP undergoing elective neurosurgery were randomly divided into the sevoflurane group (S group, n = 32) and the propofol group (P group, n = 32). Anesthesia was maintained with sevoflurane and propofol until the end of the operation. Demographic information, operation information and hemodynamic variables were recorded. The levels of hormones were evaluated preoperatively as the baseline (T0), 1h after the beginning of the operation (T1), immediately at the end of the operation (T2) and 72 h postoperatively (T3).<h4>Results</h4>There were no significant differences in the two groups in terms of patients' demographics and intraoperative information, such as operation duration, blood loss and transfusion volumes, and fluid infusion volume (P>0.05). In both groups, compared to those at T0, the levels of TSH, FT3, TT3 and ACTH at T1, T2 and T3 were significantly lower. The levels of FSH, PRL and GH at T3 were also significantly lower (P<0.05). The FT3 and TT3 levels of both groups at T2 and T3 were significantly lower than those at T1, but the ACTH level was significantly increased (P<0.05). Compared to the levels at T2, the TSH, FT3, FT4 and ACTH levels of the two groups at T3 were significantly reduced (P<0.05). The baseline hormone levels of both groups were similar (P>0.05). At T1, the FT3, TT3, FT4, TT4 and ACTH levels in the P group were significantly lower than those in the S group (P<0.05). At T2, the TT3 and ACTH levels of the P group were significantly lower than those of the S group (P<0.05) At T3, the TT4 level in the P group was significantly lower than that of the S group (P<0.05).<h4>Conclusion</h4>Propofol and sevoflurane could reduce the levels of hormones intraoperatively and postoperatively in pediatric patients with craniopharyngioma. However, propofol reduced hormone levels more intensively, mainly intraoperatively. Postoperatively, propofol and sevoflurane had similar inhibition effects on the shift in hormone levels. Therefore, in pediatric patients with craniopharyngioma undergoing neurosurgery, sevoflurane might be the preferred anesthetic because it causes less interruption of hormone levels. However, because of their similar postoperative effects, which long-term effects of sevoflurane or propofol could produce optimal clinical situations? Thus more extensive clinical studies are needed.<h4>Trial registration</h4>Clinical trial registration. This trail was registered at Chinese Clinical Trial Registry (http://www.chictr.org.cn, Jun Xiong) on 28/12/2021, registration number was ChiCTR2100054885.

Omayma O Abdelaleem, Olfat G. Shaker, Mohamed M. Mohamed et al.

Background: Ischemic stroke is one of the serious complications of diabetes. Non-coding RNAs are established as promising biomarkers for diabetes and its complications. The present research investigated the expression profiles of serum TUG1, LINC00657, miR-9, and miR-106a in diabetic patients with and without stroke.Methods: A total of 75 diabetic patients without stroke, 77 patients with stroke, and 71 healthy controls were recruited in the current study. The serum expression levels of TUG1, LINC00657, miR-9, and miR-106a were assessed using quantitative real-time polymerase chain reaction assays.Results: We observed significant high expression levels of LINC00657 and miR-9 in the serum of diabetic patients without stroke compared to control participants. At the same time, we found marked increases of serum TUG1, LINC00657, and miR-9 and a marked decrease of serum miR-106a in diabetic patients who had stroke relative to those without stroke. Also, we revealed positive correlations between each of TUG1, LINC00657, and miR-9 and the National Institutes of Health Stroke Scale (NIHSS). However, there was a negative correlation between miR-106a and NIHSS. Finally, we demonstrated a negative correlation between LINC00657 and miR-106a in diabetic patients with stroke.Conclusion: Serum non-coding RNAs, TUG1, LINC00657, miR-9, and miR-106a displayed potential as novel molecular biomarkers for diabetes complicated with stroke, suggesting that they might be new therapeutic targets for the treatment of diabetic patients with stroke.

Juan Huang, Luyuan Wang, Bin Tang et al.

Golden buckwheat (Fagopyrum cymosum) is used in Traditional Chinese Medicine. It has received attention because of the high value of its various medicinal and nutritional metabolites, especially flavonoids (catechin and epicatechin). However, the metabolites and their encoding genes in golden buckwheat have not yet been identified in the global landscape. This study performed transcriptomics and widely targeted metabolomics analyses for the first time on rhizomes of golden buckwheat. As a result, 10,191 differentially expressed genes (DEGs) and 297 differentially regulated metabolites (DRMs) were identified, among which the flavonoid biosynthesis pathway was enriched in both transcriptome and metabolome. The integration analyses of the transcriptome and the metabolome revealed a network related to catechin, in which four metabolites and 14 genes interacted with each other. Subsequently, an SG5 R2R3-MYB transcription factor, named FcMYB1, was identified as a transcriptional activator in catechin biosynthesis, as it was positively correlated to eight flavonoid biosynthesis genes in their expression patterns and was directly bound to the promoters of FcLAR2 and FcF3'H1 by yeast one hybrid analysis. Finally, a flavonoid biosynthesis pathway was proposed in the rhizomes of golden buckwheat, including 13 metabolites, 11 genes encoding 9 enzymes, and 1 MYB transcription factor. The expression of 12 DEGs were validated by qRT-PCR, resulting in a good agreement with the Pearson R ranging from 0.83 to 1. The study provided a comprehensive flavonoid biosynthesis and regulatory network of golden buckwheat.

Carl J. Lavie, Cemal Ozemek, Leonard A. Kaminsky

Aditi Arya, Loick P. Kojom Foko, Shewta Chaudhry et al.

Artemisinin-based combination therapies (ACT) are currently used as a first-line malaria therapy in endemic countries worldwide. This systematic review aims at presenting the current scenario of drug resistance molecular markers, either selected or involved in treatment failures (TF) during in vivo ACT efficacy studies from sub-Saharan Africa (sSA) and India. Eight electronic databases were comprehensively used to search relevant articles and finally a total of 28 studies were included in the review, 21 from sSA and seven from India. On analysis, Artemether + lumefantrine (AL) and artesunate + sulfadoxine-pyrimethamine (AS + SP) are the main ACT in African and Indian regions with a 28-day efficacy range of 54.3–100% for AL and 63–100% for AS + SP respectively. It was observed that mutations in the Pfcrt (76T), Pfdhfr (51I, 59R, 108N), Pfdhps (437G) and Pfmdr1 (86Y, 184F, 1246Y) genes were involved in TF, which varied with respect to ACTs. Based on studies that have genotyped the Pfk13 gene, the reported TF cases, were mainly linked with mutations in genes associated with resistance to ACT partner drugs; indicating that the protection of the partner drug efficacy is crucial for maintaining the efficacy of ACT. This review reveals that ACT are largely efficacious in India and sSA despite the fact that some clinical efficacy and epidemiological studies have reported some validated mutations (i.e., 476I, 539T and 561H) in circulation in these two regions. Also, the role of PfATPase6 in ART resistance is controversial still, while P. falciparum plasmepsin 2 (Pfpm2) in piperaquine (PPQ) resistance and dihydroartemisinin (DHA) + PPQ failures is well documented in Southeast Asian countries but studied less in sSA. Hence, there is a need for continuous molecular surveillance of Pfk13 mutations for emergence of artemisinin (ART) resistance in these countries.