Front Endocrinol (Lausanne). 2025 Jan 14;16:1513929. doi: 10.3389/fendo.2025.1513929. eCollection 2025.

NO ABSTRACT

PMID:39882100 | PMC:PMC11772191 | DOI:10.3389/fendo.2025.1513929

Sovrem Tekhnologii Med. 2024;16(4):5-13. doi: 10.17691/stm2024.16.4.01. Epub 2024 Aug 30.

ABSTRACT

The aim of the study is to develop a method for diagnosing fungal keratitis based on the analysis of photographs of the anterior segment of the eye using deep learning algorithms with subsequent evaluation of sensitivity and specificity of the method on a test data set in comparison with the results of practicing ophthalmologists.

MATERIALS AND METHODS: The study has included the stages of data acquisition, image pre-training and markup, selection of training approach and neural network architecture, training with input data augmentation, validation with hyperparameter correction, evaluation of algorithm performance on a test sample, and determination of sensitivity and specificity of fungal keratitis detection by practicing doctors. A total of 274 anterior segment images were used, including 130 photographs of the eyes affected by fungal keratitis and 144 photographs illustrating normal eyes, keratitis of other etiologies, and various anterior segment pathologies. Photographs taken after the treatment onset, illustrations of keratitis of mixed etiology and corneal perforation were excluded from the study. Images of the training sample were marked up using the VGG Image Annotator web application and then used to train the YOLOv8 convolutional neural network. Images from the test data set were also offered to practicing ophthalmologists to determine the diagnostic accuracy of fungal keratitis.

RESULTS: The sensitivity of the model was 56.0%, the specificity level reached 96.1%, and the proportion of correct answers of the algorithm was 76.5%. The accuracy of image recognition by practicing ophthalmologists was 50.0%, specificity - 41.7%, sensitivity - 57.7%.

CONCLUSION: The study showed the high potential of deep learning algorithms in the diagnosis of fungal keratitis and its advantages in accuracy compared to expert judgment in the absence of metadata. The use of computer vision technologies may find application as a complementary diagnostic method in decision making in complex cases and in telemedicine care settings. Further research is required to compare the developed model with alternative approaches, to expand and standardize databases.

PMID:39881837 | PMC:PMC11773139 | DOI:10.17691/stm2024.16.4.01

Health Inf Sci Syst. 2025 Jan 27;13(1):19. doi: 10.1007/s13755-025-00339-5. eCollection 2025 Dec.

ABSTRACT

Coronary artery disease (CAD) remains the leading cause of death globally, highlighting the critical need for accurate diagnostic tools in medical imaging. Traditional segmentation methods for coronary angiograms often struggle with vessel discontinuity and inaccuracies, impeding effective diagnosis and treatment planning. To address these challenges, we developed the Local Adaptive Segmentation Framework (LASF), enhancing the YOLOv8 architecture with dilation and erosion algorithms to improve the continuity and precision of vascular image segmentation. We further enriched the ARCADE dataset by meticulously annotating both proximal and distal vascular segments, thus broadening the dataset's applicability for training robust segmentation models. Our comparative analyses reveal that LASF outperforms well-known models such as UNet and DeepLabV3Plus, demonstrating superior metrics in precision, recall, and F1-score across various testing scenarios. These enhancements ensure more reliable and accurate segmentation, critical for clinical applications. LASF represents a significant advancement in the segmentation of vascular images within coronary angiograms. By effectively addressing the common issues of vessel discontinuity and segmentation accuracy, LASF stands to improve the clinical management of CAD, offering a promising tool for enhancing diagnostic accuracy and patient outcomes in medical settings.

PMID:39881813 | PMC:PMC11772642 | DOI:10.1007/s13755-025-00339-5

Insights Imaging. 2025 Jan 29;16(1):25. doi: 10.1186/s13244-024-01896-1.

ABSTRACT

PURPOSES: The presence of clinically significant prostate cancer (csPCa) is equivocal for patients with prostate imaging reporting and data system (PI-RADS) category 3. We aim to develop deep learning models for re-stratify risks in PI-RADS category 3 patients.

METHODS: This retrospective study included a bi-parametric MRI of 1567 consecutive male patients from six centers (Centers 1-6) between Jan 2015 and Dec 2020. Deep learning models with double channel attention modules based on MRI (AttenNet) for predicting PCa and csPCa were constructed separately. Each model was first pretrained using 1144 PI-RADS 1-2 and 4-5 images and then retrained using 238 PI-RADS 3 images from three training centers (centers 1-3), and tested using 185 PI-RADS 3 images from the other three testing centers (centers 4-6).

RESULTS: Our AttenNet models achieved excellent prediction performances in testing cohort of center 4-6 with the area under the receiver operating characteristic curves (AUC) of 0.795 (95% CI: [0.700, 0.891]), 0.963 (95% CI: [0.915, 1]) and 0.922 (95% CI: [0.810, 1]) in predicting PCa, and the corresponding AUCs were 0.827 (95% CI: [0.703, 0.952]) and 0.926 (95% CI: [0.846, 1]) in predicting csPCa in testing cohort of center 4 and center 5. In particular, 71.1% to 92.2% of non-csPCa patients were identified by our model in three testing cohorts, who can spare from invasive biopsy or RP procedure.

CONCLUSIONS: Our model offers a noninvasive screening clinical tool to re-stratify risks in PI-RADS 3 patients, thereby reducing unnecessary invasive biopsies and improving the effectiveness of biopsies.

CRITICAL RELEVANCE STATEMENT: The deep learning model with MRI can help to screen out csPCa in PI-RADS category 3.

KEY POINTS: AttenNet models included channel attention and soft attention modules. 71.1-92.2% of non-csPCa patients were identified by the AttenNet model. The AttenNet models can be a screen clinical tool to re-stratify risks in PI-RADS 3 patients.

PMID:39881076 | DOI:10.1186/s13244-024-01896-1

Nat Cancer. 2025 Jan 29. doi: 10.1038/s43018-024-00904-z. Online ahead of print.

ABSTRACT

The diagnostic landscape of brain tumors integrates comprehensive molecular markers alongside traditional histopathological evaluation. DNA methylation and next-generation sequencing (NGS) have become a cornerstone in central nervous system (CNS) tumor classification. A limiting requirement for NGS and methylation profiling is sufficient DNA quality and quantity, which restrict its feasibility. Here we demonstrate NePSTA (neuropathology spatial transcriptomic analysis) for comprehensive morphological and molecular neuropathological diagnostics from single 5-µm tissue sections. NePSTA uses spatial transcriptomics with graph neural networks for automated histological and molecular evaluations. Trained and evaluated across 130 participants with CNS malignancies and healthy donors across four medical centers, NePSTA predicts tissue histology and methylation-based subclasses with high accuracy. We demonstrate the ability to reconstruct immunohistochemistry and genotype profiling on tissue with minimal requirements, inadequate for conventional molecular diagnostics, demonstrating the potential to enhance tumor subtype identification with implications for fast and precise diagnostic workup.

PMID:39880907 | DOI:10.1038/s43018-024-00904-z

Sci Rep. 2025 Jan 29;15(1):3640. doi: 10.1038/s41598-025-86532-8.

ABSTRACT

Air pollution in cities, especially NO2, is linked to numerous health problems, ranging from mortality to mental health challenges and attention deficits in children. While cities globally have initiated policies to curtail emissions, real-time monitoring remains challenging due to limited environmental sensors and their inconsistent distribution. This gap hinders the creation of adaptive urban policies that respond to the sequence of events and daily activities affecting pollution in cities. Here, we demonstrate how city CCTV cameras can act as a pseudo-NO2 sensors. Using a predictive graph deep model, we utilised traffic flow from London's cameras in addition to environmental and spatial factors, generating NO2 predictions from over 133 million frames. Our analysis of London's mobility patterns unveiled critical spatiotemporal connections, showing how specific traffic patterns affect NO2 levels, sometimes with temporal lags of up to 6 h. For instance, if trucks only drive at night, their effects on NO2 levels are most likely to be seen in the morning when people commute. These findings cast doubt on the efficacy of some of the urban policies currently being implemented to reduce pollution. By leveraging existing camera infrastructure and our introduced methods, city planners and policymakers could cost-effectively monitor and mitigate the impact of NO2 and other pollutants.

PMID:39880905 | DOI:10.1038/s41598-025-86532-8

Commun Biol. 2025 Jan 29;8(1):141. doi: 10.1038/s42003-025-07568-0.

ABSTRACT

In mammalian oocytes, large-scale chromatin organization regulates transcription, nuclear architecture, and maintenance of chromosome stability in preparation for meiosis onset. Pre-ovulatory oocytes with distinct chromatin configurations exhibit profound differences in metabolic and transcriptional profiles that ultimately determine meiotic competence and developmental potential. Here, we developed a deep learning pipeline for the non-invasive prediction of chromatin structure and developmental potential in live mouse oocytes. Our Fluorescence prediction and Classification on Bright-field (Fluo-Cast-Bright) pipeline achieved 91.3% accuracy in the classification of chromatin state in fixed oocytes and 85.7% accuracy in live oocytes. Importantly, transcriptome analysis following non-invasive selection revealed that meiotically competent oocytes exhibit a higher expression of transcripts associated with RNA and protein nuclear export, maternal mRNA deadenylation, histone modifications, chromatin remodeling and signaling pathways regulating microtubule dynamics during the metaphase-I to metaphase-II transition. Fluo-Cast-Bright provides fast and non-invasive selection of meiotically competent oocytes for downstream research and clinical applications.

PMID:39880880 | DOI:10.1038/s42003-025-07568-0

J Cheminform. 2025 Jan 29;17(1):12. doi: 10.1186/s13321-024-00945-7.

ABSTRACT

G protein-coupled receptors (GPCRs) play vital roles in various physiological processes, making them attractive drug discovery targets. Meanwhile, deep learning techniques have revolutionized drug discovery by facilitating efficient tools for expediting the identification and optimization of ligands. However, existing models for the GPCRs often focus on single-target or a small subset of GPCRs or employ binary classification, constraining their applicability for high throughput virtual screening. To address these issues, we introduce AiGPro, a novel multitask model designed to predict small molecule agonists (EC50) and antagonists (IC50) across the 231 human GPCRs, making it a first-in-class solution for large-scale GPCR profiling. Leveraging multi-scale context aggregation and bidirectional multi-head cross-attention mechanisms, our approach demonstrates that ensemble models may not be necessary for predicting complex GPCR states and small molecule interactions. Through extensive validation using stratified tenfold cross-validation, AiGPro achieves robust performance with Pearson's correlation coefficient of 0.91, indicating broad generalizability. This breakthrough sets a new standard in the GPCR studies, outperforming previous studies. Moreover, our first-in-class multi-tasking model can predict agonist and antagonist activities across a wide range of GPCRs, offering a comprehensive perspective on ligand bioactivity within this diverse superfamily. To facilitate easy accessibility, we have deployed a web-based platform for model access at https://aicadd.ssu.ac.kr/AiGPro . Scientific Contribution We introduce a deep learning-based multi-task model to generalize the agonist and antagonist bioactivity prediction for GPCRs accurately. The model is implemented on a user-friendly web server to facilitate rapid screening of small-molecule libraries, expediting GPCR-targeted drug discovery. Covering a diverse set of 231 GPCR targets, the platform delivers a robust, scalable solution for advancing GPCR-focused therapeutic development. The proposed framework incorporates an innovative dual-label prediction strategy, enabling the simultaneous classification of molecules as agonists, antagonists, or both. Each prediction is further accompanied by a confidence score, offering a quantitative measure of activity likelihood. This advancement moves beyond conventional models focusing solely on binding affinity, providing a more comprehensive understanding of ligand-receptor interactions. At the core of our model lies the Bi-Directional Multi-Head Cross-Attention (BMCA) module, a novel architecture that captures forward and backward contextual embeddings of protein and ligand features. By leveraging BMCA, the model effectively integrates structural and sequence-level information, ensuring a precise representation of molecular interactions. Results show that this approach is highly accurate in binding affinity predictions and consistent across diverse GPCR families. By unifying agonist and antagonist bioactivity prediction into a single model architecture, we bridge a critical gap in GPCR modeling. This enhances prediction accuracy and accelerates virtual screening workflows, offering a valuable and innovative solution for advancing GPCR-targeted drug discovery.

PMID:39881398 | DOI:10.1186/s13321-024-00945-7

BMC Oral Health. 2025 Jan 29;25(1):153. doi: 10.1186/s12903-025-05431-6.

ABSTRACT

BACKGROUND: The severity of furcation involvement (FI) directly affected tooth prognosis and influenced treatment approaches. However, assessing, diagnosing, and treating molars with FI was complicated by anatomical and morphological variations. Cone-beam computed tomography (CBCT) enhanced diagnostic accuracy for detecting FI and measuring furcation defects. Despite its advantages, the high cost and radiation dose associated with CBCT equipment limited its widespread use. The aim of this study was to evaluate the performance of the Vision Transformer (ViT) in comparison with several commonly used traditional deep learning (DL) models for classifying molars with or without FI on panoramic radiographs.

METHODS: A total of 1,568 tooth images obtained from 506 panoramic radiographs were used to construct the database and evaluate the models. This study developed and assessed a ViT model for classifying FI from panoramic radiographs, and compared its performance with traditional models, including Multi-Layer Perceptron (MLP), Visual Geometry Group (VGG)Net, and GoogLeNet.

RESULTS: Among the evaluated models, the ViT model outperformed all others, achieving the highest precision (0.98), recall (0.92), and F1 score (0.95), along with the lowest cross-entropy loss (0.27) and the highest accuracy (92%). ViT also recorded the highest area under the curve (AUC) (98%), outperforming the other models with statistically significant differences (p < 0.05), confirming its enhanced classification capability. The gradient-weighted class activation mapping (Grad-CAM) analysis on the ViT model revealed the key areas of the images that the model focused on during predictions.

CONCLUSION: DL algorithms can automatically classify FI using readily accessible panoramic images. These findings demonstrate that ViT outperforms the tested traditional models, highlighting the potential of transformer-based approaches to significantly advance image classification. This approach is also expected to reduce both the radiation dose and the financial burden on patients while simultaneously improving diagnostic precision.

PMID:39881302 | DOI:10.1186/s12903-025-05431-6

BMC Med Educ. 2025 Jan 29;25(1):143. doi: 10.1186/s12909-025-06716-8.

ABSTRACT

STUDY OBJECTIVE: Student-centered learning and unconventional teaching modalities are gaining popularity in medical education. One notable approach involves engaging students in producing creative projects to complement the learning of preclinical topics. A systematic review was conducted to characterize the impact of creative project-based learning on metacognition and knowledge gains in medical students.

METHODS: A systematic search was conducted using MEDLINE and Embase via Ovid, PubMed, CINAHL, Web of Science, Cochrane CENTRAL, and Scopus from January 1st, 1995, to July 6th, 2023. Studies using quantitative, qualitative, or mixed-methods approaches that explored the impact of creative project-based lessons on medical students' educational outcomes were included. Two investigators independently screened the titles and abstracts and extracted data from included articles. A narrative synthesis was conducted to summarize study designs and outcome measures. Content analysis was conducted to generate codes and themes. Study quality was assessed using the Mixed Methods Appraisal Tool in view of the range of study types employed.

RESULTS: The review included 17 studies published between 2010 to 2022. These studies implemented various creative project interventions such as handicraft models, drawings, and concept maps covering multiple topics, including anatomy, histopathology, and fundamental sciences. The identified themes of Enhanced Learning, Collaborative Learning, and Deep Learning led to further themes of Student Engagement, Student Disengagement, and Faculty Engagement. Collaborative learning involves students working in teams and benefitting from effective mentorship. Creative projects facilitated deep learning objectives via interdisciplinary learning and promoted new ways of perceiving concepts. Learning was enhanced through increased interactivity, high conceptual fidelity and improved knowledge retention.

CONCLUSION: Creative projects undertaken by medical students exhibit attributes that facilitate the acquisition of collaborative and deep learning objectives through self-directed learning, cognitive load modulation, and metacognitive behaviours. Faculty mentorship and group learning amongst peers facilitate these processes, although challenges such as high task demands, cognitive and emotional intensiveness, and mismatch with students; professional identities remain. Overall, students and faculty received these interventions well, thus, warranting further exploration for uses in medical curricula.

TRIAL REGISTRATION: Not applicable as this study is a systematic review.

PMID:39881268 | DOI:10.1186/s12909-025-06716-8

BMC Bioinformatics. 2025 Jan 29;26(1):33. doi: 10.1186/s12859-025-06047-x.

ABSTRACT

BACKGROUND: Single-cell RNA sequencing (scRNA-seq) has transformed biological research by offering new insights into cellular heterogeneity, developmental processes, and disease mechanisms. As scRNA-seq technology advances, its role in modern biology has become increasingly vital. This study explores the application of deep learning to single-cell data clustering, with a particular focus on managing sparse, high-dimensional data.

RESULTS: We propose the SMD deep learning model, which integrates nonlinear dimensionality reduction techniques with a porous dilated attention gate component. Built upon a convolutional autoencoder and informed by the negative binomial distribution, the SMD model efficiently captures essential cell clustering features and dynamically adjusts feature weights. Comprehensive evaluation on both public datasets and proprietary osteosarcoma data highlights the SMD model's efficacy in achieving precise classifications for single-cell data clustering, showcasing its potential for advanced transcriptomic analysis.

CONCLUSION: This study underscores the potential of deep learning-specifically the SMD model-in advancing single-cell RNA sequencing data analysis. By integrating innovative computational techniques, the SMD model provides a powerful framework for unraveling cellular complexities, enhancing our understanding of biological processes, and elucidating disease mechanisms. The code is available from https://github.com/xiaoxuc/scSMD .

PMID:39881248 | DOI:10.1186/s12859-025-06047-x

BMC Cancer. 2025 Jan 29;25(1):173. doi: 10.1186/s12885-025-13582-6.

ABSTRACT

BACKGROUND: To develop and test the performance of a fully automated system for classifying renal tumor subtypes via deep machine learning for automated segmentation and classification.

MATERIALS AND METHODS: The model was developed using computed tomography (CT) images of pathologically proven renal tumors collected from a prospective cohort at a medical center between March 2016 and December 2020. A total of 561 renal tumors were included: 233 clear cell renal cell carcinomas (RCCs), 82 papillary RCCs, 74 chromophobe RCCs, and 172 angiomyolipomas. Renal tumor masks manually drawn on contrast-enhanced CT images were used to develop a 3D U-Net-based deep learning model for fully automated tumor segmentation. After segmentation, the entire classification pipeline, including feature extraction and subtype classification, was conducted without any manual intervention. Both conventional radiological features (Hounsfield units, HUs) and radiomic features extracted from areas predicted by the deep learning models were used to develop an algorithm for classifying renal tumor subtypes via a random forest classifier. The performance of the segmentation model was evaluated using the Dice similarity coefficient, while the classification model was assessed based on accuracy, sensitivity, and specificity.

RESULTS: For tumors larger than 4 cm, the Dice similarity coefficient (DSC) for automated segmentation was 0.83, while for tumors smaller than 4 cm, the DSC was 0.65. The classification accuracy (ACC) for distinguishing RCC subtypes was 0.77 for tumors larger than 4 cm and 0.68 for tumors smaller than 4 cm. Additionally, the accuracy for benign versus malignant classification was 0.85.

CONCLUSIONS: Our automatic segmentation and classifier model showed promising results for renal tumor segmentation and classification.

PMID:39881216 | DOI:10.1186/s12885-025-13582-6

Sci Rep. 2025 Jan 30;15(1):3734. doi: 10.1038/s41598-025-88210-1.

ABSTRACT

Infertility has emerged as a significant global health concern. Assisted reproductive technology (ART) assists numerous infertile couples in conceiving, yet some experience repeated, unsuccessful cycles. This study aims to identify the pivotal clinical factors influencing the success of fresh embryo transfer of in vitro fertilization (IVF). We introduce a novel Non-negative Matrix Factorization (NMF)-based Ensemble algorithm (NMFE). By combining feature matrices from NMF, accelerated multiplicative updates for non-negative matrix factorization (AMU-NMF), and the generalized deep learning clustering (GDLC) algorithm. NMFE exhibits superior accuracy and reliability in analyzing the in vitro fertilization and embryo transfer (IVF-ET) dataset. The dataset comprises 2238 cycles and 85 independent clinical features, categorized into 13 categories based on feature correlation. Subsequently, the NMFE model was trained and reached convergence. Then the features of 13 categories were sequentially masked to analyze their individual effects on IVF-ET live births. The NMFE analysis highlights the significant influence of therapeutic interventions, Embryo transfer outcomes, and ovarian response assessment on live births of IVF-ET. Therapeutic interventions, including ovarian stimulation protocols, ovulation stimulation drugs, and pre-and intra-stimulation cycle acupuncture play prominent roles. However, their impacts on the IVF-ET model are reduced, suggesting a potential synergistic effect when combined. Conversely, factors like basic information, diagnosis, and obstetric history have a lesser influence. The NMFE algorithm demonstrates promising potential in assessing the influence of clinical features on live births in IVF fresh embryo transfer.

PMID:39881210 | DOI:10.1038/s41598-025-88210-1

Sci Rep. 2025 Jan 29;15(1):3648. doi: 10.1038/s41598-024-77602-4.

ABSTRACT

The opioid crisis has disproportionately affected U.S. veterans, leading the Veterans Health Administration to implement opioid prescribing guidelines. Veterans who receive care from both VA and non-VA providers-known as dual-system users-have an increased risk of Opioid Use Disorder (OUD). The interaction between dual-system use and demographic and clinical factors, however, has not been previously explored. We conducted a retrospective study of 856,299 patient instances from the Washington DC and Baltimore VA Medical Centers (2012-2019), using a deep neural network (DNN) and explainable Artificial Intelligence to examine the impact of dual-system use on OUD and how demographic and clinical factors interact with it. Of the cohort, 146,688(17%) had OUD, determined through Natural Language Processing of clinical notes and ICD-9/10 diagnoses. The DNN model, with a 78% area under the curve, confirmed that dual-system use is a risk factor for OUD, along with prior opioid use or other substance use. Interestingly, a history of other drug use interacted negatively with dual-system use regarding OUD risk. In contrast, older age was associated with a lower risk of OUD but interacted positively with dual-system use. These findings suggest that within the dual-system users, patients with certain risk profiles warrant special attention.

PMID:39881142 | DOI:10.1038/s41598-024-77602-4

Insights Imaging. 2025 Jan 29;16(1):27. doi: 10.1186/s13244-025-01904-y.

ABSTRACT

OBJECTIVES: To develop and validate radiomics and deep learning models based on contrast-enhanced MRI (CE-MRI) for differentiating dual-phenotype hepatocellular carcinoma (DPHCC) from HCC and intrahepatic cholangiocarcinoma (ICC).

METHODS: Our study consisted of 381 patients from four centers with 138 HCCs, 122 DPHCCs, and 121 ICCs (244 for training and 62 for internal tests, centers 1 and 2; 75 for external tests, centers 3 and 4). Radiomics, deep transfer learning (DTL), and fusion models based on CE-MRI were established for differential diagnosis, respectively, and their diagnostic performances were compared using the confusion matrix and area under the receiver operating characteristic (ROC) curve (AUC).

RESULTS: The radiomics model demonstrated competent diagnostic performance, with a macro-AUC exceeding 0.9, and both accuracy and F1-score above 0.75 in the internal and external validation sets. Notably, the vgg19-combined model outperformed the radiomics and other DTL models. The fusion model based on vgg19 further improved diagnostic performance, achieving a macro-AUC of 0.990 (95% CI: 0.965-1.000), an accuracy of 0.935, and an F1-score of 0.937 in the internal test set. In the external test set, it similarly performed well, with a macro-AUC of 0.988 (95% CI: 0.964-1.000), accuracy of 0.875, and an F1-score of 0.885.

CONCLUSIONS: Both the radiomics and the DTL models were able to differentiate DPHCC from HCC and ICC before surgery. The fusion models showed better diagnostic accuracy, which has important value in clinical application.

CRITICAL RELEVANCE STATEMENT: MRI-based deep learning radiomics were able to differentiate DPHCC from HCC and ICC preoperatively, aiding clinicians in the identification and targeted treatment of these malignant hepatic tumors.

KEY POINTS: Fusion models may yield an incremental value over radiomics models in differential diagnosis. Radiomics and deep learning effectively differentiate the three types of malignant hepatic tumors. The fusion models may enhance clinical decision-making for malignant hepatic tumors.

PMID:39881111 | DOI:10.1186/s13244-025-01904-y

Insights Imaging. 2025 Jan 29;16(1):29. doi: 10.1186/s13244-025-01902-0.

ABSTRACT

OBJECTIVES: To determine whether deep learning-based reconstructions of zero-echo-time (ZTE-DL) sequences enhance image quality and bone visualization in cervical spine MRI compared to traditional zero-echo-time (ZTE) techniques, and to assess the added value of ZTE-DL sequences alongside standard cervical spine MRI for comprehensive pathology evaluation.

METHODS: In this retrospective study, 52 patients underwent cervical spine MRI using ZTE, ZTE-DL, and T2-weighted 3D sequences on a 1.5-Tesla scanner. ZTE-DL sequences were reconstructed from raw data using the AirReconDL algorithm. Three blinded readers independently evaluated image quality, artifacts, and bone delineation on a 5-point Likert scale. Cervical structures and pathologies, including soft tissue and bone components in spinal canal and neural foraminal stenosis, were analyzed. Image quality was quantitatively assessed by signal-to-noise ratio (SNR) and contrast-to-noise ratio (CNR).

RESULTS: Mean image quality scores were 2.0 ± 0.7 for ZTE and 3.2 ± 0.6 for ZTE-DL, with ZTE-DL exhibiting fewer artifacts and superior bone delineation. Significant differences were observed between T2-weighted and ZTE-DL sequences for evaluating intervertebral space, anterior osteophytes, spinal canal, and neural foraminal stenosis (p < 0.05), with ZTE-DL providing more accurate assessments. ZTE-DL also showed improved evaluation of the osseous components of neural foraminal stenosis compared to ZTE (p < 0.05).

CONCLUSIONS: ZTE-DL sequences offer superior image quality and bone visualization compared to ZTE sequences and enhance standard cervical spine MRI in assessing bone involvement in spinal canal and neural foraminal stenosis.

CRITICAL RELEVANCE STATEMENT: Deep learning-based reconstructions improve zero-echo-time sequences in cervical spine MRI by enhancing image quality and bone visualization. This advancement offers additional insights for assessing bone involvement in spinal canal and neural foraminal stenosis, advancing clinical radiology practice.

KEY POINTS: Conventional MRI encounters challenges with osseous structures due to low signal-to-noise ratio. Zero-echo-time (ZET) sequences offer CT-like images of the C-spine but with lower quality. Deep learning reconstructions improve image quality of zero-echo-time sequences. ZTE sequences with deep learning reconstructions refine cervical spine osseous pathology assessment. These sequences aid assessment of bone involvement in spinal and foraminal stenosis.

PMID:39881081 | DOI:10.1186/s13244-025-01902-0

Eur Radiol Exp. 2025 Jan 29;9(1):12. doi: 10.1186/s41747-025-00552-7.

ABSTRACT

BACKGROUND: Body composition scores allow for quantifying the volume and physical properties of specific tissues. However, their manual calculation is time-consuming and prone to human error. This study aims to develop and validate CompositIA, an automated, open-source pipeline for quantifying body composition scores from thoraco-abdominal computed tomography (CT) scans.

METHODS: A retrospective dataset of 205 contrast-enhanced thoraco-abdominal CT examinations was used for training, while 54 scans from a publicly available dataset were used for independent testing. Two radiology residents performed manual segmentation, identifying the centers of the L1 and L3 vertebrae and segmenting the corresponding axial slices. MultiResUNet was used to identify CT slices intersecting the L1 and L3 vertebrae, and its performance was evaluated using the mean absolute error (MAE). Two U-nets were used to segment the axial slices, with performance evaluated through the volumetric Dice similarity coefficient (vDSC). CompositIA's performance in quantifying body composition indices was assessed using mean percentage relative error (PRE), regression, and Bland-Altman analyses.

RESULTS: On the independent dataset, CompositIA achieved a MAE of about 5 mm in detecting slices intersecting the L1 and L3 vertebrae, with a MAE < 10 mm in at least 85% of cases and a vDSC greater than 0.85 in segmenting axial slices. Regression and Bland-Altman analyses demonstrated a strong linear relationship and good agreement between automated and manual scores (p values < 0.001 for all indices), with mean PREs ranging from 5.13% to 15.18%.

CONCLUSION: CompositIA facilitated the automated quantification of body composition scores, achieving high precision in independent testing.

RELEVANCE STATEMENT: CompositIA is an automated, open-source pipeline for quantifying body composition indices from CT scans, simplifying clinical assessments, and expanding their applicability.

KEY POINTS: Manual body composition assessment from CTs is time-consuming and prone to errors. CompositIA was trained on 205 CT scans and tested on 54 scans. CompositIA demonstrated mean percentage relative errors under 15% compared to manual indices. CompositIA simplifies body composition assessment through an artificial intelligence-driven and open-source pipeline.

PMID:39881078 | DOI:10.1186/s41747-025-00552-7

Appl Neuropsychol Adult. 2025 Jan 29:1-10. doi: 10.1080/23279095.2025.2457999. Online ahead of print.

ABSTRACT

In the fields of engineering, science, technology, and medicine, artificial intelligence (AI) has made significant advancements. In particular, the application of AI techniques in medicine, such as machine learning (ML) and deep learning (DL), is rapidly growing and offers great potential for aiding physicians in the early diagnosis of illnesses. Depression, one of the most prevalent and debilitating mental illnesses, is projected to become the leading cause of disability worldwide by 2040. For early diagnosis, a patient-friendly, cost-effective approach based on readily observable and objective indicators is essential. The objective of this research is to develop machine learning and deep learning techniques that utilize electroencephalogram (EEG) signals to diagnose depression. Different statistical features were extracted from the EEG signals and fed into the models. Three classifiers were constructed: 1D Convolutional Neural Network (1DCNN), Support Vector Machine (SVM), and Logistic Regression (LR). The methods were tested on a dataset comprising EEG signals from 34 patients with Major Depressive Disorder (MDD) and 30 healthy subjects. The signals were collected under three distinct conditions: TASK, when the subject was performing a task; Eye Close (EC), when the subject's eyes were closed; and Eye Open (EO), when the subject's eyes were open. All three classifiers were applied to each of the three types of signals, resulting in nine (3 × 3) experiments. The results showed that TASK signals yielded the highest accuracies of 88.4%, 89.3%, and 90.21% for LR, SVM, and 1DCNN, respectively, compared to EC and EO signals. Additionally, the proposed methods outperformed some state-of-the-art approaches. These findings highlight the potential of EEG-based approaches for the clinical diagnosis of depression and provide promising avenues for further research. Additionally, the proposed methodology demonstrated statistically significant improvements in classification accuracy, with p-values < 0.05, ensuring robustness and reliability.

PMID:39879638 | DOI:10.1080/23279095.2025.2457999

J Med Internet Res. 2025 Jan 29;27:e57723. doi: 10.2196/57723.

ABSTRACT

BACKGROUND: Neuroimaging segmentation is increasingly important for diagnosing and planning treatments for neurological diseases. Manual segmentation is time-consuming, apart from being prone to human error and variability. Transformers are a promising deep learning approach for automated medical image segmentation.

OBJECTIVE: This scoping review will synthesize current literature and assess the use of various transformer models for neuroimaging segmentation.

METHODS: A systematic search in major databases, including Scopus, IEEE Xplore, PubMed, and ACM Digital Library, was carried out for studies applying transformers to neuroimaging segmentation problems from 2019 through 2023. The inclusion criteria allow only for peer-reviewed journal papers and conference papers focused on transformer-based segmentation of human brain imaging data. Excluded are the studies dealing with nonneuroimaging data or raw brain signals and electroencephalogram data. Data extraction was performed to identify key study details, including image modalities, datasets, neurological conditions, transformer models, and evaluation metrics. Results were synthesized using a narrative approach.

RESULTS: Of the 1246 publications identified, 67 (5.38%) met the inclusion criteria. Half of all included studies were published in 2022, and more than two-thirds used transformers for segmenting brain tumors. The most common imaging modality was magnetic resonance imaging (n=59, 88.06%), while the most frequently used dataset was brain tumor segmentation dataset (n=39, 58.21%). 3D transformer models (n=42, 62.69%) were more prevalent than their 2D counterparts. The most developed were those of hybrid convolutional neural network-transformer architectures (n=57, 85.07%), where the vision transformer is the most frequently used type of transformer (n=37, 55.22%). The most frequent evaluation metric was the Dice score (n=63, 94.03%). Studies generally reported increased segmentation accuracy and the ability to model both local and global features in brain images.

CONCLUSIONS: This review represents the recent increase in the adoption of transformers for neuroimaging segmentation, particularly for brain tumor detection. Currently, hybrid convolutional neural network-transformer architectures achieve state-of-the-art performances on benchmark datasets over standalone models. Nevertheless, their applicability remains highly limited by high computational costs and potential overfitting on small datasets. The heavy reliance of the field on the brain tumor segmentation dataset hints at the use of a more diverse set of datasets to validate the performances of models on a variety of neurological diseases. Further research is needed to define the optimal transformer architectures and training methods for clinical applications. Continuing development may make transformers the state-of-the-art for fast, accurate, and reliable brain magnetic resonance imaging segmentation, which could lead to improved clinical tools for diagnosing and evaluating neurological disorders.

PMID:39879621 | DOI:10.2196/57723

JMIR Public Health Surveill. 2025 Jan 29;11:e63809. doi: 10.2196/63809.

ABSTRACT

BACKGROUND: Suicide represents a critical public health concern, and machine learning (ML) models offer the potential for identifying at-risk individuals. Recent studies using benchmark datasets and real-world social media data have demonstrated the capability of pretrained large language models in predicting suicidal ideation and behaviors (SIB) in speech and text.

OBJECTIVE: This study aimed to (1) develop and implement ML methods for predicting SIBs in a real-world crisis helpline dataset, using transformer-based pretrained models as a foundation; (2) evaluate, cross-validate, and benchmark the model against traditional text classification approaches; and (3) train an explainable model to highlight relevant risk-associated features.

METHODS: We analyzed chat protocols from adolescents and young adults (aged 14-25 years) seeking assistance from a German crisis helpline. An ML model was developed using a transformer-based language model architecture with pretrained weights and long short-term memory layers. The model predicted suicidal ideation (SI) and advanced suicidal engagement (ASE), as indicated by composite Columbia-Suicide Severity Rating Scale scores. We compared model performance against a classical word-vector-based ML model. We subsequently computed discrimination, calibration, clinical utility, and explainability information using a Shapley Additive Explanations value-based post hoc estimation model.

RESULTS: The dataset comprised 1348 help-seeking encounters (1011 for training and 337 for testing). The transformer-based classifier achieved a macroaveraged area under the curve (AUC) receiver operating characteristic (ROC) of 0.89 (95% CI 0.81-0.91) and an overall accuracy of 0.79 (95% CI 0.73-0.99). This performance surpassed the word-vector-based baseline model (AUC-ROC=0.77, 95% CI 0.64-0.90; accuracy=0.61, 95% CI 0.61-0.80). The transformer model demonstrated excellent prediction for nonsuicidal sessions (AUC-ROC=0.96, 95% CI 0.96-0.99) and good prediction for SI and ASE, with AUC-ROCs of 0.85 (95% CI 0.97-0.86) and 0.87 (95% CI 0.81-0.88), respectively. The Brier Skill Score indicated a 44% improvement in classification performance over the baseline model. The Shapley Additive Explanations model identified language features predictive of SIBs, including self-reference, negation, expressions of low self-esteem, and absolutist language.

CONCLUSIONS: Neural networks using large language model-based transfer learning can accurately identify SI and ASE. The post hoc explainer model revealed language features associated with SI and ASE. Such models may potentially support clinical decision-making in suicide prevention services. Future research should explore multimodal input features and temporal aspects of suicide risk.

PMID:39879608 | DOI:10.2196/63809