BMC Cancer. 2025 Apr 9;25(1):648. doi: 10.1186/s12885-025-13611-4.

ABSTRACT

OBJECTIVE: The assessment of immunotherapy plays a pivotal role in the clinical management of skin melanoma. Graph neural networks (GNNs), alongside other deep learning algorithms and bioinformatics approaches, have demonstrated substantial promise in advancing cancer diagnosis and treatment strategies.

METHODS: GNNs models were developed to predict the response to immunotherapy and to pinpoint key pathways. Utilizing the genes from these key pathways, multi-omics bioinformatics methods were employed to refine the construction of a gene signature, termed responseScore, aimed at enhancing the precision of immunotherapy response predictions. Subsequently, responseScore was explored from the perspectives of prognosis, genetic variation, pathway enrichment, and the tumor microenvironment. Concurrently, the association among 13 genes contributing to responseScore and factors such as immunotherapy response, prognosis, and the tumor microenvironment was investigated. Among these genes, PSMB6 was subjected to an in-depth analysis of its biological effect through experimental approaches like transfection and co-culture.

RESULTS: In the finalized model utilizing GNNs, it has revealed an AUC of 0.854 within the training dataset and 0.824 within the testing set, pinpointing key pathways such as R-HSA-70,268. The indicator named as responseScore excelled in its predictive accuracy regarding immunotherapy response and patient prognosis. Investigations into genetic variation, pathway enrichment, tumor microenvironment disclosed a profound association between responseScore and the enhancement of immune cell infiltration and anti-tumor immunity. A negative correlation was observed between the expression of PSMB6 and immune genes, with elevated PSMB6 expression correlating with poor prognosis. ELISA detection after co-cultivation experiments revealed significant reductions in the levels of cytokines IL-6 and IL-1β in specimens from the PCDH-PSMB6 group.

CONCLUSION: The GNNs prediction model and the responseScore developed in this research effectively indicate the immunotherapy response and prognosis for patients with skin melanoma. Additionally, responseScore provides insights into the tumor microenvironment and the characteristics of tumor immunity of melanoma. Thirteen genes identified in this study show promise as potential tumor markers or therapeutic targets. Notably, PSMB6 emerges as a potential therapeutic target for skin melanoma, where its elevated expression exhibits an inhibitory effect on the tumor immunity.

PMID:40205338 | DOI:10.1186/s12885-025-13611-4

J Imaging Inform Med. 2025 Apr 9. doi: 10.1007/s10278-025-01491-w. Online ahead of print.

ABSTRACT

This study aimed to develop and validate a predictive model for early recurrence of high-grade glioma (HGG) within 180 days, assess the prognostic value of preoperative and postoperative temporalis muscle metrics (area and thickness), and explore their significance in postoperative follow-up. Seventy-one molecularly confirmed HGG patients were included, with data sourced from local data and TCIA (The Cancer Imaging Archive) RHUH-GBM (Río Hortega University Hospital Glioblastoma) dataset. Tumor segmentation was performed using deep learning, and radiomic features were extracted following comparison with manual segmentation. Feature selection was conducted using mutual information and recursive feature elimination. A comprehensive model integrating 3D tumor radiomics and temporalis muscle metrics was developed and compared with a tumor-only model to identify the optimal predictive framework. SHAP analysis was used to evaluate model interpretability and feature importance. The TM_Tumor_HistGradientBoosting model, incorporating 16 features including temporalis muscle metrics, outperformed the tumor-only model in accuracy (0.89), recall (0.87), and F1 score (0.88). SHAP analysis highlighted that preoperative temporalis muscle cross-sectional area was strongly associated with early recurrence risk, while postoperative temporalis muscle thickness significantly contributed to recurrence prediction. Combining temporalis muscle metrics with preoperative tumor MRI substantially improved the accuracy of early recurrence prediction in HGG. Temporalis muscle metrics serve as objective and sustainable prognostic indicators with significant clinical value in postoperative follow-up.

PMID:40205255 | DOI:10.1007/s10278-025-01491-w

Nature. 2025 Apr;640(8058):470-477. doi: 10.1038/s41586-025-08829-y. Epub 2025 Apr 9.

ABSTRACT

The complexity of neural circuits makes it challenging to decipher the brain's algorithms of intelligence. Recent breakthroughs in deep learning have produced models that accurately simulate brain activity, enhancing our understanding of the brain's computational objectives and neural coding. However, it is difficult for such models to generalize beyond their training distribution, limiting their utility. The emergence of foundation models1 trained on vast datasets has introduced a new artificial intelligence paradigm with remarkable generalization capabilities. Here we collected large amounts of neural activity from visual cortices of multiple mice and trained a foundation model to accurately predict neuronal responses to arbitrary natural videos. This model generalized to new mice with minimal training and successfully predicted responses across various new stimulus domains, such as coherent motion and noise patterns. Beyond neural response prediction, the model also accurately predicted anatomical cell types, dendritic features and neuronal connectivity within the MICrONS functional connectomics dataset2. Our work is a crucial step towards building foundation models of the brain. As neuroscience accumulates larger, multimodal datasets, foundation models will reveal statistical regularities, enable rapid adaptation to new tasks and accelerate research.

PMID:40205215 | DOI:10.1038/s41586-025-08829-y

Nature. 2025 Apr;640(8058):368-374. doi: 10.1038/s41586-025-08854-x. Epub 2025 Apr 9.

ABSTRACT

Over the past decade, photonics research has explored accelerated tensor operations, foundational to artificial intelligence (AI) and deep learning1-4, as a path towards enhanced energy efficiency and performance5-14. The field is centrally motivated by finding alternative technologies to extend computational progress in a post-Moore's law and Dennard scaling era15-19. Despite these advances, no photonic chip has achieved the precision necessary for practical AI applications, and demonstrations have been limited to simplified benchmark tasks. Here we introduce a photonic AI processor that executes advanced AI models, including ResNet3 and BERT20,21, along with the Atari deep reinforcement learning algorithm originally demonstrated by DeepMind22. This processor achieves near-electronic precision for many workloads, marking a notable entry for photonic computing into competition with established electronic AI accelerators23 and an essential step towards developing post-transistor computing technologies.

PMID:40205212 | DOI:10.1038/s41586-025-08854-x

J Med Ultrason (2001). 2025 Apr 9. doi: 10.1007/s10396-025-01528-1. Online ahead of print.

ABSTRACT

PURPOSE: In recent years, computer-aided diagnosis (CAD) using deep learning methods for medical images has been studied. Although studies have been conducted to classify ultrasound images of tumors of the liver into four categories (liver cysts (Cyst), liver hemangiomas (Hemangioma), hepatocellular carcinoma (HCC), and metastatic liver cancer (Meta)), no studies with additional information for deep learning have been reported. Therefore, we attempted to improve the classification accuracy of ultrasound images of hepatic tumors by adding hepatitis virus infection information to deep learning.

METHODS: Four combinations of hepatitis virus infection information were assigned to each image, plus or minus HBs antigen and plus or minus HCV antibody, and the classification accuracy was compared before and after the information was input and weighted to fully connected layers.

RESULTS: With the addition of hepatitis virus infection information, accuracy changed from 0.574 to 0.643. The F1-Score for Cyst, Hemangioma, HCC, and Meta changed from 0.87 to 0.88, 0.55 to 0.57, 0.46 to 0.59, and 0.54 to 0.62, respectively, remaining the same for Hemangioma but increasing for the rest.

CONCLUSION: Learning hepatitis virus infection information showed the highest increase in the F1-Score for HCC, resulting in improved classification accuracy of ultrasound images of hepatic tumors.

PMID:40205118 | DOI:10.1007/s10396-025-01528-1

NPJ Digit Med. 2025 Apr 10;8(1):198. doi: 10.1038/s41746-025-01593-3.

ABSTRACT

Transthoracic echocardiography (TTE), commonly used for initial screening of pulmonary hypertension (PH), often lacks sufficient accuracy. To address this gap, we developed and validated a multimodal fusion model for improved PH screening (MMF-PH). The study was registered in the ClinicalTrials.gov (NCT05566002, 09/30/2022). The MMF-PH underwent extensive training, validation, and testing, including comparisons with TTE and evaluations across various patient subgroups to assess robustness and reliability. We analyzed 2451 patients who underwent right heart catheterization, supplemented by a prospective dataset of 477 patients and an external dataset. The MMF-PH demonstrated robust performance across different datasets. The model outperformed TTE in terms of specificity and negative predictive value across all test datasets. An ablation study using the external test dataset confirmed the essential role of each module in the MMF-PH. The MMF-PH significantly advances PH detection, offering robust and reliable diagnostic accuracy across diverse patient populations and clinical settings.

PMID:40205021 | DOI:10.1038/s41746-025-01593-3

PLoS One. 2025 Apr 9;20(4):e0319299. doi: 10.1371/journal.pone.0319299. eCollection 2025.

ABSTRACT

Traditional crack detection methods initially relied on manual observation, followed by instrument-assisted techniques. Today, road surface inspection leverages deep learning to achieve automated crack detection. However, in the domain of deep learning-based road surface damage classification, the heterogeneous and complex nature of road environments introduces significant background noise and unstructured features. These factors often undermine the robustness and generalization capability of models, thereby adversely affecting classification accuracy. To address this challenge, this research incorporates edge priors by integrating traditional edge detection techniques with deep convolutional neural networks (DCNNs). This paper proposes an innovative mechanism called Edge-Enhanced Multi-Scale Feature Fusion (EE-MSFF), which enhances edge information through multi-scale feature extraction, thereby mitigating the impact of complex backgrounds and improving the model's focus on crack regions. Specifically, the proposed mechanism leverages classical edge detection operators such as Sobel, Prewitt, and Laplacian to perform multi-scale edge information extraction during the feature extraction phase of the model. This process captures both local edge features and global structural information in crack regions, thereby enhancing the model's resistance to interference from complex backgrounds. By employing multi-scale receptive fields, the EE-MSFF mechanism facilitates hierarchical fusion of feature maps, guiding the model to learn edge information that is correlated with crack regions. This effectively strengthens the model's ability to perceive damaged pavement features in complex environments, improving classification accuracy and stability. In this study, the model underwent systematic training and validation on both the complex-background dataset RDD2020 and the simple-background dataset Concrete_Data_Week3. Experimental results demonstrate that the proposed model achieved a classification accuracy of 88.68% on the RDD2020 dataset and 99.5% on the Concrete_Data_Week3 dataset, where background interference is minimal. Furthermore, ablation studies were conducted to analyze the independent contributions of each module, highlighting the performance improvements associated with the integration of multi-scale edge features.

PMID:40203245 | DOI:10.1371/journal.pone.0319299

Nano Lett. 2025 Apr 9. doi: 10.1021/acs.nanolett.5c01435. Online ahead of print.

ABSTRACT

Point-of-care testing (POCT) with multiplexed capability, ultrahigh sensitivity, affordable smart devices, and user-friendly operation is critically needed for clinical diagnostics and food safety. This study presents a deep-learning-assisted microfluidic immunoassay platform that uses a smartphone-based imaging transcoding system, polystyrene microsphere-based encoding, and artificial-intelligence-assisted decoding. Microspheres of varying sizes act as multiprobes, with their quantities correlating to target concentrations after an immunoreaction and separation-filtration within the microfluidic chip. A smartphone with intelligent decoding software captures images of multiprobes from the chip and performs classification, counting, and concentration calculations. The "encoding-decoding" strategy and integrated microfluidic chip design allow these processes to be completed in simple steps, eliminating the need for additional immunomagnetic separation. As a proof of concept, this platform successfully detected multiple respiratory viruses and antibiotics in various real samples with high sensitivity within 30 min, demonstrating great potential as a smart, universal toolkit for next-generation POCT applications.

PMID:40203242 | DOI:10.1021/acs.nanolett.5c01435

PLoS One. 2025 Apr 9;20(4):e0320184. doi: 10.1371/journal.pone.0320184. eCollection 2025.

ABSTRACT

BACKGROUND: Head and neck cancer (HNC) becomes a vital global health burden. Accurate assessment of the disease burden plays an essential role in setting health priorities and guiding decision-making.

METHODS: This study explores data from the Global Burden of Disease (GBD) 2021 study, involving totally 204 countries during the period from 1980 to 2021. The analysis focuses on age-standardized incidence, mortality, and disability-adjusted life years (DALYs) for HNC. A Transformer-based model, HNCP-T, is used for the prediction of future trends from 2022 to 2030, quantified based on the estimated annual percentage change (EAPC).

RESULTS: The global age-standardized incidence rate (ASIR) for HNC has escalated between 1980 and 2021, with men bearing a higher burden than women. In addition, the burden rises with age and exhibits regional disparities, with the greatest impact on low-to-middle sociodemographic index (SDI) regions. Additionally, the model predicts a continued rise in ASIR (EAPC = 0.22), while the age-standardized death rate (ASDR) is shown to decrease more sharply for women (EAPC = -0.92) than men (EAPC = -0.54). The most rapid increase in ASIR is projected for low-to-middle SDI countries, while ASDR and DALY rates are found to decrease in different degrees across regions.

CONCLUSIONS: The current work offers a detailed analysis of the global burden of HNC based on the GBD 2021 dataset and demonstrates the accuracy of the HNCP-T model in predicting future trends. Significant regional and gender-based differences are found, with incidence rates rising, especially among women and in low-middle SDI regions. Furthermore, the results underscore the value of deep learning models in disease burden prediction, which can outperform traditional methods.

PMID:40203229 | DOI:10.1371/journal.pone.0320184

PLoS One. 2025 Apr 9;20(4):e0320816. doi: 10.1371/journal.pone.0320816. eCollection 2025.

ABSTRACT

Advancements in robotics and other technological innovations have accelerated the development of surgical procedures, increasing the demand for training environments that accurately replicate human anatomy. This study developed a system that utilizes the AutoSegmentator extension of 3D Slicer, based on nnU-Net, a state-of-the-art deep learning framework for automatic organ extraction, to import automatically extracted organ surface data into CAD software along with original DICOM-derived images. This system allows medical experts to manually refine the automatically extracted data, making it more accurate and closer to the ideal dataset. First, Python programming is used to automatically generate and save JPEG-format image data from DICOM data for display in Blender. Next, DICOM data imported into 3D Slicer is processed by AutoSegmentator to extract surface data of 104 organs in bulk, which is then exported in STL format. In Blender, a custom-developed Python script aligns the image data and organ surface data within the same 3D space, ensuring accurate spatial coordinates. By using Blender's CAD functionality within this space, the automatically extracted organ boundaries can be manually adjusted based on the image data, resulting in more precise organ surface data. Additionally, organs and blood vessels that cannot be automatically extracted can be newly created and added by referencing the image data. Through this process, a comprehensive anatomical dataset encompassing all required organs and blood vessels can be constructed. The dataset created with this system is easily customizable and can be applied to various surgical simulations, including 3D-printed simulators, hybrid simulators that incorporate animal organs, and surgical simulators utilizing augmented reality (AR). Furthermore, this system is built entirely using open-source, free software, providing high reproducibility, flexibility, and accessibility. By using this system, medical professionals can actively participate in the design and data processing of surgical simulation systems, leading to shorter development times and reduced costs.

PMID:40203219 | DOI:10.1371/journal.pone.0320816

Langmuir. 2025 Apr 9. doi: 10.1021/acs.langmuir.4c04969. Online ahead of print.

ABSTRACT

Geopolymer is regarded as a novel type of eco-friendly material that may replace cement. To improve the prediction accuracy of mechanical properties of fly ash-slag-based geopolymer (FASGG), as well as optimize material composition and mix design, this study utilizes seven key parameters as variables, and compressive and flexural strengths were as outputs. Deep learning techniques were applied to train and predict 600 sets of experimental data, developing a novel predictive model of MK-CNN-GRU, which integrated Maximal Information Coefficient-K-median algorithm, Convolutional Neural Network, and Gated Recurrent Unit algorithms. Results indicated that the ranking of input parameters which were related with compressive strength was curing age, Ca/Si ratio, fly ash-to-slag ratio, Si/Al ratio, water-to-binder ratio, alkali activator modulus, and alkali equivalent. Three classical models were selected as benchmarks for predicting compressive strength at different curing ages. The MK-CNN-GRU model could fully exploit the internal features of experimental data and learn its variation patterns, resulting in more stable predictive performance. An ablation study of the submodels confirms that MK-CNN-GRU model considers temporal dependencies, long- and short-term features, as well as local dependencies and hierarchical feature representations within the data. Experimental data suggested an exponential relationship between flexural and compressive strengths of FASGG. The predictions for flexural strength indicated that the MK-CNN-GRU model effectively captured variations, demonstrating good generalization ability and applicability. This model enhances the estimation accuracy regarding the mechanical behavior of FASGG, offering a theoretical framework for refining its composition and mix design.

PMID:40203137 | DOI:10.1021/acs.langmuir.4c04969

PLoS Comput Biol. 2025 Apr 9;21(4):e1012997. doi: 10.1371/journal.pcbi.1012997. Online ahead of print.

ABSTRACT

In the mouse distal terminal phalanx (P3), it remains mystery why amputation at less than 33% of the digit results in regeneration, while amputation exceeding 67% leads to non-regeneration. Unraveling the molecular mechanisms underlying this disparity could provide crucial insights for regenerative medicine. In this study, we aim to investigate the tissues within the wound bed to understand the tissue microenvironment associated with regenerative versus non-regenerative outcomes. We employed a P3-specific amputation model in mice, integrated with time-series RNA-seq and a macrophage assay challenged with pro- and anti-inflammatory cytokines, to explore these mechanisms. Our findings revealed that non-regenerative digits exhibit a greater intense early transcriptional response in the wound bed compared to regenerative ones. Furthermore, early macrophage phenotypes differ distinctly between regenerative and non-regenerative outcomes. Regenerative digits also display unique co-expression modules related to Bone Morphogenetic Protein 2 (Bmp2). The differentially expressed genes (DEGs) between regenerative and non-regenerative digits are enriched in targets of several transcription factors, such as HOXA11 and HOXD11 from the HOX gene family, showing a time-dependent pattern of enrichment. These transcription factors, known for their roles in bone regeneration, skeletal patterning, osteoblast activity, fracture healing, angiogenesis, and key signaling pathways, may act as master regulators of the regenerative gene signatures. Additionally, we developed a deep learning AI model capable of predicting post-amputation time and level from RNA-seq data, indicating that the regenerative probability may be "encoded" in the transcriptomic response to amputation.

PMID:40203060 | DOI:10.1371/journal.pcbi.1012997

PLoS One. 2025 Apr 9;20(4):e0315340. doi: 10.1371/journal.pone.0315340. eCollection 2025.

ABSTRACT

Deep learning, as a high-performance data analysis method, has demonstrated superior efficiency and accuracy in side-channel attacks compared to traditional methods. However, many existing models enhance accuracy by stacking network layers, leading to increased algorithmic and computational complexity, overfitting, low training efficiency, and limited feature extraction capabilities. Moreover, deep learning methods rely on data correlation, and the presence of noise tends to reduce this correlation, increasing the difficulty of attacks. To address these challenges, this paper proposes the application of an InceptionNet-based network structure for side-channel attacks. This network utilizes fewer training parameters. achieves faster convergence and demonstrates improved attack efficiency through parallel processing of input data. Additionally, a LU-Net-based network structure is proposed for denoising side-channel datasets. This network captures the characteristics of input signals through an encoder, reconstructs denoised signals using a decoder, and utilizes LSTM layers and skip connections to preserve the temporal coherence and spatial details of the signals, thereby achi-eving the purpose of denoising. Experimental evaluations were conducted on the ASCAD dataset and the DPA Contest v4 dataset for comparative studies. The results indicate that the deep learning attack model proposed in this paper effectively enhances side-channel attack performance. On the ASCAD dataset, the recovery of keys requires only 30 traces, and on the DPA Contest v4 dataset, only 1 trace is needed for key recovery. Furthermore, the proposed deep learning denoising model significantly reduces the impact of noise on side-channel attack performance, thereby improving efficiency.

PMID:40203055 | DOI:10.1371/journal.pone.0315340

PLoS One. 2025 Apr 9;20(4):e0320085. doi: 10.1371/journal.pone.0320085. eCollection 2025.

ABSTRACT

An enhancer is a specific DNA sequence typically located within a gene at upstream or downstream position and serves as a pivotal element in the regulation of eukaryotic gene transcription. Therefore, the recognition of enhancers is highly significant for comprehending gene expression regulatory systems. While some useful predictive models have been proposed, there are still deficiencies in these models. To address current limitations, we propose a model, DNABERT2-Enhancer, based on transformer architecture and deep learning, designed for the recognition of enhancers (classified as either enhancer or non-enhancer) and the identification of their activity (strong or weak enhancers). More specifically, DNABERT2-Enhancer is composed of a BERT model for extracting features and a CNN model for enhancers classification. Parameters of the BERT model are initialized by a pre-training DNABERT-2 language model. The enhancer recognition task is then fine-tuned through transfer learning to convert the original sequence into feature vectors. Subsequently, the CNN network is employed to learn the feature vector generated by BERT and produce the prediction results. In comparison with existing predictors utilizing the identical dataset, our approach demonstrates superior performance. This suggests that the model will be a useful instrument for academic research on the enhancer recognition.

PMID:40203028 | DOI:10.1371/journal.pone.0320085

PLoS One. 2025 Apr 9;20(4):e0321028. doi: 10.1371/journal.pone.0321028. eCollection 2025.

ABSTRACT

Campus walking environments significantly influence college students' daily lives and shape their subjective perceptions. However, previous studies have been constrained by limited sample sizes and inefficient, time-consuming methodologies. To address these limitations, we developed a deep learning framework to evaluate campus walking perceptions across four universities in China's Yangtze River Delta region. Utilizing 15,596 Baidu Street View Images (BSVIs), and perceptual ratings from 100 volunteers across four dimensions-aesthetics, security, depression, and vitality-we employed four machine learning models to predict perceptual scores. Our results demonstrate that the Random Forest (RF) model outperformed others in predicting aesthetics, security, and vitality, while linear regression was most effective for depression. Spatial analysis revealed that perceptions of aesthetics, security, and vitality were concentrated in landmark areas and regions with high pedestrian flow. Multiple linear regression analysis indicated that buildings exhibited stronger correlations with depression (β = 0.112) compared to other perceptual aspects. Moreover, vegetation (β = 0.032) and meadows (β = 0.176) elements significantly enhanced aesthetics. This study offers actionable insights for optimizing campus walking environments from a student-centered perspective, emphasizing the importance of spatial design and visual elements in enhancing students' perceptual experiences.

PMID:40203019 | DOI:10.1371/journal.pone.0321028

PLoS One. 2025 Apr 9;20(4):e0321233. doi: 10.1371/journal.pone.0321233. eCollection 2025.

NO ABSTRACT

PMID:40203015 | DOI:10.1371/journal.pone.0321233

Invest Ophthalmol Vis Sci. 2025 Apr 1;66(4):22. doi: 10.1167/iovs.66.4.22.

ABSTRACT

PURPOSE: This study aims to develop a new method to quantify nonperfused retinal capillaries (NPCs) and evaluate NPCs in eyes with AMD and diabetic retinopathy (DR).

METHODS: We averaged multiple registered optical coherence tomography (OCT)/OCT angiography (OCTA) scans to create high-definition volumes. The deep capillary plexus slab was defined and segmented. A developed deep learning denoising algorithm removed tissue background noise from capillaries in en face OCT/OCTA. The algorithm segmented NPCs by identifying capillaries from OCT without corresponding flow signals in OCTA. We then investigated the relationships between NPCs and known features in AMD and DR.

RESULTS: The segmented NPC achieved an accuracy of 88.2% compared to manual grading of NPCs in DR. Compared to healthy controls, both the mean number and total length (mm) of NPCs was significantly increased in AMD and DR eyes (P < 0.001, P < 0.001). Compared to early and intermediate AMD, the number and total length of NPCs were significantly higher in advanced AMD (number: P < 0.001, P < 0.001; total length: P = 0.002, P = 0.003). Geography atrophy, macular neovascularization, drusen volume, and extrafoveal avascular area (EAA) significantly correlated with increased NPCs (P < 0.05). In DR eyes, NPCs correlated with the number of microaneurysms and EAA (P < 0.05). The presence of fluid did not significantly correlate with NPCs in AMD and DR.

CONCLUSIONS: A deep learning-based algorithm can segment and quantify retinal capillaries that lack flow using colocalized OCT/OCTA. This new biomarker may be useful in AMD and DR in predicting progression of these diseases.

PMID:40202734 | DOI:10.1167/iovs.66.4.22

Radiol Artif Intell. 2025 Apr 9:e240459. doi: 10.1148/ryai.240459. Online ahead of print.

ABSTRACT

"Just Accepted" papers have undergone full peer review and have been accepted for publication in Radiology: Artificial Intelligence. This article will undergo copyediting, layout, and proof review before it is published in its final version. Please note that during production of the final copyedited article, errors may be discovered which could affect the content. Purpose To evaluate the predictive value of deep learning (DL)-based coronary artery disease (CAD) extent analysis for major adverse cardiac events (MACEs) in patients with acute chest pain presenting to the emergency department (ED). Materials and Methods This retrospective multicenter observational study included consecutive patients with acute chest pain who underwent coronary CT angiography (CCTA) at three institutional EDs from January 2018 to December 2022. Patients were classified as having no CAD, nonobstructive CAD, or obstructive CAD using a DL model. The primary outcome was MACEs during follow-up, defined as a composite of cardiac death, nonfatal myocardial infarction, and hospitalization for unstable angina. Cox proportional hazards regression models were used to evaluate the predictors of MACEs. Results The study included 408 patients (224 male; mean age, 59.4 ± 14.6 years). The DL model classified 162 (39.7%) patients as having no CAD, 94 (23%) as having nonobstructive CAD, and 152 (37.3%) as having obstructive CAD. Sixty-three (15.4%) patients experienced MACEs during follow-up. Patients with MACEs had a higher prevalence of obstructive CAD than those without (P < .001). In multivariate analysis model 1 (clinical risk factors), dyslipidemia (Hazard ratio [HR], 2.15 and elevated Troponin-T (HR 2.13) predicted MACEs (all P < .05). In model 2 (clinical risk factors + DL-based CAD extent), obstructive CAD detected by the DL model was the most significant independent predictor of MACEs (HR, 88.07, P < .001). Harrell's C-statistic showed that DL-based CAD extent enhanced the risk stratification beyond clinical risk factors (Harrell's C-statistics: 0.94 versus 0.80, P < .001). Conclusion DL-based detection of obstructive CAD demonstrated stronger predictive value than clinical risk factors for MACEs in patients with acute chest pain presenting to the ED. ©RSNA, 2025.

PMID:40202417 | DOI:10.1148/ryai.240459

Radiol Artif Intell. 2025 Apr 9:e240271. doi: 10.1148/ryai.240271. Online ahead of print.

ABSTRACT

"Just Accepted" papers have undergone full peer review and have been accepted for publication in Radiology: Artificial Intelligence. This article will undergo copyediting, layout, and proof review before it is published in its final version. Please note that during production of the final copyedited article, errors may be discovered which could affect the content. Purpose To develop an adaptive dual-task deep learning model (ThyNet-S) for detection and classification of thyroid lesions at US screening. Materials and Methods The retrospective study used a multicenter dataset comprising 35008 thyroid US images of 23294 individual examinations (mean age, 40.4 years ± 13.1[SD], 17587 female) from 7 medical centers during January 2009 and December 2021. Of these, 29004 images were used for model development and 6004 images for validation. The model determined cancer risk for each image and automatically triaged images with normal thyroid and benign nodules by dynamically integrating lesion detection through pixel-level feature analysis and lesion classification through deep semantic features analysis. Diagnostic performance of screening assisted by the model (ThyNet-S triaged screening) and traditional screening (radiologists alone) was assessed by comparing sensitivity, specificity, accuracy and AUC using McNemar's test and Delong test. The influence of ThyNet-S on radiologist workload and clinical decision-making was also assessed. Results ThyNet-S-assisted triaged screening achieved higher AUC than original screening in six senior and six junior radiologists (0.93 versus 0.91, and 0.92 versus 0.88, respectively, all P < .001). The model improved sensitivity for junior radiologists (88.2% versus 86.8%, P <.001). Notably, the model reduced radiologists' workload by triaging 60.4% of cases as not potentially malignant, which did not require further interpretation. The model simultaneously decreased unnecessary fine needle aspiration rate from 38.7% to 14.9% and 11.5% when used independently or in combination with Thyroid Imaging Reporting and Data System, respectively. Conclusion ThyNet-S improved efficiency of thyroid cancer screening and optimized clinical decision-making. ©RSNA, 2025.

PMID:40202416 | DOI:10.1148/ryai.240271

ChemSusChem. 2025 Apr 9:e202500282. doi: 10.1002/cssc.202500282. Online ahead of print.

ABSTRACT

The parasitic hydrogen evolution reaction (HER) hinders electrolyte transport. It reduces the effective electrochemical surface area in the negative half-cell of vanadium redox flow batteries (VRFBs), resulting in substantial efficiency losses. We investigated the formation and evolution of hydrogen bubbles within VRFB electrodes through comprehensive experimental characterization and a detailed analysis of the resolved bubbles. The electrode was imaged using synchrotron X-ray tomography, and gas bubbles in the images were identified and characterized using a deep learning model combined with a morphological analysis tool. The HER intensity increases at more negative working electrode potentials, causing residual bubbles to grow and fuse in the electrode central region. In contrast, independent bubbles predominantly form at the electrode edges. Furthermore, the bubble growth leads to the gradual development of irregular shapes. These observations provide insights into bubble formation and evolution rules, contributing to a better understanding of the system.

PMID:40202080 | DOI:10.1002/cssc.202500282