J Chem Inf Model. 2025 Jan 30. doi: 10.1021/acs.jcim.4c02060. Online ahead of print.

ABSTRACT

Machine learning approaches often involve evaluating a wide range of models due to various available architectures. This standard strategy can lead to a lack of depth in exploring established methods. In this study, we concentrated our efforts on a single deep learning architecture type to assess whether a focused approach could enhance performance in fault diagnosis. We selected the benchmark Tennessee Eastman Process data set as our case study and investigated modifications on a reference convolutional neural network-based model. Results indicate a considerable improvement in the overall classification, reaching a maximum average F1-score of 89.85%, 7.47% above the baseline model, which is also a considerable improvement compared to other performances reported in the literature. These results emphasize the potential of this focused approach, indicating it could be further explored and applied to other data sets in future work.

PMID:39883649 | DOI:10.1021/acs.jcim.4c02060

J Am Med Inform Assoc. 2025 Jan 30:ocae327. doi: 10.1093/jamia/ocae327. Online ahead of print.

ABSTRACT

BACKGROUND: With the global population aging and advancements in the medical system, long-term care in healthcare institutions and home settings has become essential for older adults with disabilities. However, the diverse and scattered care requirements of these individuals make developing effective long-term care plans heavily reliant on professional nursing staff, and even experienced caregivers may make mistakes or face confusion during the care plan development process. Consequently, there is a rigid demand for intelligent systems that can recommend comprehensive long-term care plans for older adults with disabilities who have stable clinical conditions.

OBJECTIVE: This study aims to utilize deep learning methods to recommend comprehensive care plans for the older adults with disabilities.

METHODS: We model the care data of older adults with disabilities using a bipartite graph. Additionally, we employ a prediction-based graph self-supervised learning (SSL) method to mine deep representations of graph nodes. Furthermore, we propose a novel graph Transformer architecture that incorporates eigenvector centrality to augment node features and uses graph structural information as references for the self-attention mechanism. Ultimately, we present the Bipartite Graph Transformer (BiT) model to provide personalized long-term care plan recommendation.

RESULTS: We constructed a bipartite graph comprising of 1917 nodes and 195 240 edges derived from real-world care data. The proposed model demonstrates outstanding performance, achieving an overall F1 score of 0.905 for care plan recommendations. Each care service item reached an average F1 score of 0.897, indicating that the BiT model is capable of accurately selecting services and effectively balancing the trade-off between incorrect and missed selections.

DISCUSSION: The BiT model proposed in this paper demonstrates strong potential for improving long-term care plan recommendations by leveraging bipartite graph modeling and graph SSL. This approach addresses the challenges of manual care planning, such as inefficiency, bias, and errors, by offering personalized and data-driven recommendations. While the model excels in common care items, its performance on rare or complex services could be enhanced with further refinement. These findings highlight the model's ability to provide scalable, AI-driven solutions to optimize care planning, though future research should explore its applicability across diverse healthcare settings and service types.

CONCLUSIONS: Compared to previous research, the novel model proposed in this article effectively learns latent topology in bipartite graphs and achieves superior recommendation performance. Our study demonstrates the applicability of SSL and graph transformers in recommending long-term care plans for older adults with disabilities.

PMID:39883541 | DOI:10.1093/jamia/ocae327

Med Biol Eng Comput. 2025 Jan 30. doi: 10.1007/s11517-025-03295-0. Online ahead of print.

ABSTRACT

With the advancement of artificial intelligence technology, more and more effective methods are being used to identify and classify Electroencephalography (EEG) signals to address challenges in healthcare and brain-computer interface fields. The applications and major achievements of Graph Convolution Network (GCN) techniques in EEG signal analysis are reviewed in this paper. Through an exhaustive search of the published literature, a module-by-module discussion is carried out for the first time to address the current research status of GCN. An exhaustive classification of methods and a systematic analysis of key modules, such as brain map construction, node feature extraction, and GCN architecture design, are presented. In addition, we pay special attention to several key research issues related to GCN. This review enhances the understanding of the future potential of GCN in the field of EEG signal analysis. At the same time, several valuable development directions are sorted out for researchers in related fields, such as analysing the applicability of different GCN layers, building task-oriented GCN models, and improving adaptation to limited data.

PMID:39883372 | DOI:10.1007/s11517-025-03295-0

EJNMMI Phys. 2025 Jan 30;12(1):9. doi: 10.1186/s40658-025-00724-9.

ABSTRACT

Single photon emission computed tomography (SPECT), a technique capable of capturing functional and molecular information, has been widely adopted in theranostics applications across various fields, including cardiology, neurology, and oncology. The spatial resolution of SPECT imaging is relatively poor, which poses a significant limitation, especially the visualization of small lesions. The main factors affecting the limited spatial resolution of SPECT include projection sampling techniques, hardware and software. Both hardware and software innovations have contributed substantially to improved SPECT imaging quality. The present review provides an overview of state-of-the-art methods for improving spatial resolution in clinical and pre-clinical SPECT systems. It delves into advancements in detector design and modifications, projection sampling techniques, traditional reconstruction algorithm development and optimization, and the emerging role of deep learning. Hardware enhancements can result in SPECT systems that are both lighter and more compact, while also improving spatial resolution. Software innovations can mitigate the costs of hardware modifications. This survey offers a thorough overview of the rapid advancements in resolution enhancement techniques within the field of SPECT, with the objective of identifying the most recent trends. This is anticipated to facilitate further optimization and improvement of clinical systems, enabling the visualization of small lesions in the early stages of tumor detection, thereby enhancing accurate localization and facilitating both diagnostic imaging and radionuclide therapy, ultimately benefiting both clinicians and patients.

PMID:39883257 | DOI:10.1186/s40658-025-00724-9

Adv Biotechnol (Singap). 2024 Mar 1;2(1):9. doi: 10.1007/s44307-024-00018-8.

ABSTRACT

Engineering microbial cell factories have achieved much progress in producing fuels, natural products and bulk chemicals. However, in industrial fermentation, microbial cells often face various predictable and stochastic disturbances resulting from intermediate metabolites or end product toxicity, metabolic burden and harsh environment. These perturbances can potentially decrease productivity and titer. Therefore, strain robustness is essential to ensure reliable and sustainable production efficiency. In this review, the current strategies to improve host robustness were summarized, including knowledge-based engineering approaches, such as transcription factors, membrane/transporters and stress proteins, and the traditional adaptive laboratory evolution based on natural selection. Computation-assisted (e.g. GEMs, deep learning and machine learning) design of robust industrial hosts was also introduced. Furthermore, the challenges and future perspectives on engineering microbial host robustness are proposed to promote the development of green, efficient and sustainable biomanufacturers.

PMID:39883204 | DOI:10.1007/s44307-024-00018-8

Eur Spine J. 2025 Jan 30. doi: 10.1007/s00586-025-08672-9. Online ahead of print.

ABSTRACT

PURPOSE: Lumbar spinal stenosis (LSS) is a frequently occurring condition defined by narrowing of the spinal or nerve root canal due to degenerative changes. Physicians use MRI scans to determine the severity of stenosis, occasionally complementing it with X-ray or CT scans during the diagnostic work-up. However, manual grading of stenosis is time-consuming and induces inter-reader variability as a standardized grading system is lacking. Machine Learning (ML) has the potential to aid physicians in this process by automating segmentation and classification of LSS. However, it is unclear what models currently exist to perform these tasks.

METHODS: A systematic review of literature was performed by searching the Cochrane Library, Embase, Emcare, PubMed, and Web of Science databases for studies describing an ML-based algorithm to perform segmentation or classification of the lumbar spine for LSS. Risk of bias was assessed through an adjusted version of the Newcastle-Ottawa Quality Assessment Scale that was more applicable to ML studies. Qualitative analyses were performed based on type of algorithm (conventional ML or Deep Learning (DL)) and task (segmentation or classification).

RESULTS: A total of 27 articles were included of which nine on segmentation, 16 on classification and 2 on both tasks. The majority of studies focused on algorithms for MRI analysis. There was wide variety among the outcome measures used to express model performance. Overall, ML algorithms are able to perform segmentation and classification tasks excellently. DL methods tend to demonstrate better performance than conventional ML models. For segmentation the best performing DL models were U-Net based. For classification U-Net and unspecified CNNs powered the models that performed the best for the majority of outcome metrics. The number of models with external validation was limited.

CONCLUSION: DL models achieve excellent performance for segmentation and classification tasks for LSS, outperforming conventional ML algorithms. However, comparisons between studies are challenging due to the variety in outcome measures and test datasets. Future studies should focus on the segmentation task using DL models and utilize a standardized set of outcome measures and publicly available test dataset to express model performance. In addition, these models need to be externally validated to assess generalizability.

PMID:39883162 | DOI:10.1007/s00586-025-08672-9

Eur Radiol. 2025 Jan 30. doi: 10.1007/s00330-025-11385-8. Online ahead of print.

ABSTRACT

OBJECTIVES: We aimed to use artificial intelligence to accurately identify molecular subgroups of medulloblastoma (MB), predict clinical outcomes, and incorporate deep learning-based imaging features into the risk stratification.

METHODS: The MRI features were extracted for molecular subgroups by a novel multi-parameter convolutional neural network (CNN) called Bi-ResNet-MB. Then, MR features were used to establish a prognosis model based on XGBoost. Finally, a novel risk stratification system to stratify the patients based on the M2R Score (Machine learning-based Medulloblastoma Risk Score) was proposed.

RESULTS: A total of 139 MB patients (36 female, average age 7.27 ± 3.62 years) were treated at Beijing Tiantan Hospital. The Bi-ResNet-MB model excelled in molecular subgroup classification, achieving an average AUC of 0.946 (95% CI: 0.899-0.993). For prognostic prediction, our models achieved AUCs of 0.840 (95% CI: 0.792-0.888), 0.949 (95% CI: 0.899-0.999), and 0.960 (95% CI: 0.915-1.000) for OS, and 0.946 (95% CI: 0.905-0.987), 0.932 (95% CI: 0.875-0.989), and 0.964 (95% CI: 0.921-1.000) for PFS at 1, 3, and 5 years. In an independent validation dataset of 108 patients (33 female, average age 7.11 ± 2.92 years), the average AUC of molecular subgroup classification reached 0.894 (95% CI: 0.797-1.000). For PFS prediction at 1, 3, and 5 years, the AUCs were 0.832 (95% CI: 0.724-0.920), 0.875 (95% CI: 0.781-0.967), and 0.907 (95% CI: 0.760-1.000), respectively.

CONCLUSIONS: Based on machine learning and MRI data, models for MB molecular subgroups and prognosis prediction and the novel risk stratification system may significantly benefit patients.

KEY POINTS: Question Medulloblastoma exhibits significant heterogeneity, leading to considerable variations in patient prognosis and there is a lack of effective risk assessment strategies. Findings We have constructed a comprehensive machine learning system that excels in subgrouping diagnosis, prognosis assessment, and risk stratification for medulloblastoma patients preoperatively. Clinical relevance The utilization of non-invasive preoperative diagnosis and assessment is advantageous for clinicians in creating personalized treatment plans, particularly for high-risk patients. Additionally, it lays a foundation for the subsequent implementation of neoadjuvant therapy for medulloblastoma.

PMID:39883158 | DOI:10.1007/s00330-025-11385-8

Eur J Nucl Med Mol Imaging. 2025 Jan 30. doi: 10.1007/s00259-025-07090-9. Online ahead of print.

ABSTRACT

BACKGROUND: Pathological grade is a critical determinant of clinical outcomes and decision-making of follicular lymphoma (FL). This study aimed to develop a deep learning model as a digital biopsy for the non-invasive identification of FL grade.

METHODS: This study retrospectively included 513 FL patients from five independent hospital centers, randomly divided into training, internal validation, and external validation cohorts. A multimodal fusion Transformer model was developed integrating 3D PET tumor images with tabular data to predict FL grade. Additionally, the model is equipped with explainable modules, including Gradient-weighted Class Activation Mapping (Grad-CAM) for PET images, SHapley Additive exPlanations analysis for tabular data, and the calculation of predictive contribution ratios for both modalities, to enhance clinical interpretability and reliability. The predictive performance was evaluated using the area under the receiver operating characteristic curve (AUC) and accuracy, and its prognostic value was also assessed.

RESULTS: The Transformer model demonstrated high accuracy in grading FL, with AUCs of 0.964-0.985 and accuracies of 90.2-96.7% in the training cohort, and similar performance in the validation cohorts (AUCs: 0.936-0.971, accuracies: 86.4-97.0%). Ablation studies confirmed that the fusion model outperformed single-modality models (AUCs: 0.974 - 0.956, accuracies: 89.8%-85.8%). Interpretability analysis revealed that PET images contributed 81-89% of the predictive value. Grad-CAM highlighted the tumor and peri-tumor regions. The model also effectively stratified patients by survival risk (P < 0.05), highlighting its prognostic value.

CONCLUSIONS: Our study developed an explainable multimodal fusion Transformer model for accurate grading and prognosis of FL, with the potential to aid clinical decision-making.

PMID:39883138 | DOI:10.1007/s00259-025-07090-9

Arch Gynecol Obstet. 2025 Jan 30. doi: 10.1007/s00404-025-07960-5. Online ahead of print.

ABSTRACT

OBJECTIVE: To assess and compare the diagnostic accuracy of radiologist, MR findings, and radiomics-clinical models in the diagnosis of placental implantation disorders.

METHODS: Retrospective collection of MR images from patients suspected of having placenta accreta spectrum (PAS) was conducted across three institutions: Institution I (n = 505), Institution II (n = 67), and Institution III (n = 58). Data from Institution I were utilized to form a training set, while data from Institutions II and III served as an external test set. Radiologist diagnosis was performed by radiologists of varying levels of experience. The interpretation of MR findings was conducted by two radiologists with 10-15 years of experience in pelvic MR diagnosis, following the guidelines for diagnosis. Radiomics analysis extracted features from sagittal T2-weighted images and combined them with prenatal clinical features to construct predictive models. These models were then evaluated for discrimination and calibration to assess their performance.

RESULTS: As measured by the area under the receiver operating characteristic curve (AUC), the diagnostic efficacy was 0.587 (0.542-0.630) for junior radiologists from Institution I, 0.568 (0.441-0.689) from Institution II, and 0.507 (0.373-0.641) from Institution III. The AUC was 0.623 (0.580-0.666) for senior radiologists from Institution I, 0.635 (0.508-0.749) from Institution II, and 0.632 (0.495-0.755) from Institution III. The diagnostic efficacy of MR findings was 0.648 (0.601-0.695) for Institution I, 0.569 (0.429-0.709) for Institution II, and 0.588 (0.442-0.735) for Institution III. The diagnostic efficacy of the radiomics-clinical model was significantly higher, with an AUC of 0.794 (0.754-0.833) for Institution I, 0.783 (0.664-0.903) for Institution II, and 0.816 (0.704-0.927) for Institution III. The diagnostic efficacy of the Fusion model was significantly higher, with an AUC of 0.867 (0.836-0.899) for Institution I, 0.849 (0.753-0.944) for Institution II, and 0.823 (0.708-0.939) for Institution III.

CONCLUSION: The fusion models demonstrated superior diagnostic efficacy compared to radiologists, MR findings, and the radiomics-clinical models. Furthermore, the diagnostic accuracy of PAS was notably higher when utilizing the radiomics-clinical models than when relying solely on radiologist diagnosis or MR findings.

ADVANCES IN KNOWLEDGE: Radiomics analysis substantially augments the diagnostic precision in PAS, providing a significant enhancement over conventional radiologist and MRI findings. The diagnostic efficacy of the fusion model is notably superior to that of individual diagnostic modalities.

PMID:39883136 | DOI:10.1007/s00404-025-07960-5

Clin Nucl Med. 2025 Jan 28. doi: 10.1097/RLU.0000000000005685. Online ahead of print.

ABSTRACT

PURPOSE: The common approach for organ segmentation in hybrid imaging relies on coregistered CT (CTAC) images. This method, however, presents several limitations in real clinical workflows where mismatch between PET and CT images are very common. Moreover, low-dose CTAC images have poor quality, thus challenging the segmentation task. Recent advances in CT-less PET imaging further highlight the necessity for an effective PET organ segmentation pipeline that does not rely on CT images. Therefore, the goal of this study was to develop a CT-less multitracer PET segmentation framework.

PATIENTS AND METHODS: We collected 2062 PET/CT images from multiple scanners. The patients were injected with either 18F-FDG (1487) or 68Ga-PSMA (575). PET/CT images with any kind of mismatch between PET and CT images were detected through visual assessment and excluded from our study. Multiple organs were delineated on CT components using previously trained in-house developed nnU-Net models. The segmentation masks were resampled to coregistered PET images and used to train 4 different deep learning models using different images as input, including noncorrected PET (PET-NC) and attenuation and scatter-corrected PET (PET-ASC) for 18F-FDG (tasks 1 and 2, respectively using 22 organs) and PET-NC and PET-ASC for 68Ga tracers (tasks 3 and 4, respectively, using 15 organs). The models' performance was evaluated in terms of Dice coefficient, Jaccard index, and segment volume difference.

RESULTS: The average Dice coefficient over all organs was 0.81 ± 0.15, 0.82 ± 0.14, 0.77 ± 0.17, and 0.79 ± 0.16 for tasks 1, 2, 3, and 4, respectively. PET-ASC models outperformed PET-NC models (P < 0.05) for most of organs. The highest Dice values were achieved for the brain (0.93 to 0.96 in all 4 tasks), whereas the lowest values were achieved for small organs, such as the adrenal glands. The trained models showed robust performance on dynamic noisy images as well.

CONCLUSIONS: Deep learning models allow high-performance multiorgan segmentation for 2 popular PET tracers without the use of CT information. These models may tackle the limitations of using CT segmentation in PET/CT image quantification, kinetic modeling, radiomics analysis, dosimetry, or any other tasks that require organ segmentation masks.

PMID:39883026 | DOI:10.1097/RLU.0000000000005685

Anal Chem. 2025 Jan 30. doi: 10.1021/acs.analchem.4c06057. Online ahead of print.

ABSTRACT

Off-axis integrated cavity output spectroscopy (OA-ICOS) allows the laser to be reflected multiple times inside the cavity, increasing the effective absorption path length and thus improving sensitivity. However, OA-ICOS systems are affected by various types of noise, and traditional filtering methods offer low processing efficiency and perform limited feature extraction. Deep learning models enable us to extract important features from large-scale, complex spectral data and analyze them efficiently and accurately. We propose a carbon dioxide (CO2) sensor operating in the near-infrared spectral region (1.602 μm) based on OA-ICOS and deep learning models. A radiofrequency (RF) noise source is employed to reduce the cavity-mode noise in OA-ICOS and thus improve the signal-to-noise ratio (SNR). A time-series-based neural network, known as the informer, is employed for filtering CO2 spectral time series. After filtering, spectral features are directly extracted from the filtered spectral data and CO2 concentrations are predicted using a multilayer perceptron (MLP) model. Our results showed that the SNR attained using informer filtering approximately double those obtained using traditional filtering methods (Savitzky-Golay filtering, Kalman filtering, and wavelet threshold). The linear correlation coefficient (R2) between measured concentrations and standard concentrations was increased from 79.74% (obtained by using the absorption-peak-fitting method) to 98.52% (obtained by using the proposed MLP model). Moreover, the detection limit of the CO2 sensor using the MLP model reached 1.38 ppm at 224.4 s, a 3.79-fold improvement compared to that obtained by using the absorption-peak-fitting method. Our results demonstrate the feasibility of integrating deep learning methods in the field of spectroscopy-based sensing and provide a promising approach for spectral data processing.

PMID:39882837 | DOI:10.1021/acs.analchem.4c06057

J Chem Theory Comput. 2025 Jan 30. doi: 10.1021/acs.jctc.4c01602. Online ahead of print.

ABSTRACT

Despite its importance in understanding biology and computer-aided drug discovery, the accurate prediction of protein ionization states remains a formidable challenge. Physics-based approaches struggle to capture the small, competing contributions in the complex protein environment, while machine learning (ML) is hampered by the scarcity of experimental data. Here, we report the development of pKa ML (KaML) models based on decision trees and graph attention networks (GAT), exploiting physicochemical understanding and a new experiment pKa database (PKAD-3) enriched with highly shifted pKa's. KaML-CBtree significantly outperforms the current state of the art in predicting pKa values and ionization states across all six titratable amino acids, notably achieving accurate predictions for deprotonated cysteines and lysines─a blind spot in previous models. The superior performance of KaMLs is achieved in part through several innovations, including the separate treatment of acid and base, data augmentation using AlphaFold structures, and model pretraining on a theoretical pKa database. We also introduce the classification of protonation states as a metric for evaluating pKa prediction models. A meta-feature analysis suggests a possible reason for the lightweight tree model to outperform the more complex deep learning GAT. We release an end-to-end pKa predictor based on KaML-CBtree and the new PKAD-3 database, which facilitates a variety of applications and provides the foundation for further advances in protein electrostatic research.

PMID:39882632 | DOI:10.1021/acs.jctc.4c01602

Front Med (Lausanne). 2025 Jan 15;11:1505692. doi: 10.3389/fmed.2024.1505692. eCollection 2024.

ABSTRACT

INTRODUCTION: In recent years, the development of artificial intelligence (AI) technologies, including machine learning, deep learning, and large language models, has significantly supported clinical work. Concurrently, the integration of artificial intelligence with the medical field has garnered increasing attention from medical experts. This study undertakes a dynamic and longitudinal bibliometric analysis of AI publications within the healthcare sector over the past three decades to investigate the current status and trends of the fusion between medicine and artificial intelligence.

METHODS: Following a search on the Web of Science, researchers retrieved all reviews and original articles concerning artificial intelligence in healthcare published between January 1993 and December 2023. The analysis employed Bibliometrix, Biblioshiny, and Microsoft Excel, incorporating the bibliometrix R package for data mining and analysis, and visualized the observed trends in bibliometrics.

RESULTS: A total of 22,950 documents were collected in this study. From 1993 to 2023, there was a discernible upward trajectory in scientific output within bibliometrics. The United States and China emerged as primary contributors to medical artificial intelligence research, with Harvard University leading in publication volume among institutions. Notably, the rapid expansion of emerging topics such as COVID-19 and new drug discovery in recent years is noteworthy. Furthermore, the top five most cited papers in 2023 were all pertinent to the theme of ChatGPT.

CONCLUSION: This study reveals a sustained explosive growth trend in AI technologies within the healthcare sector in recent years, with increasingly profound applications in medicine. Additionally, medical artificial intelligence research is dynamically evolving with the advent of new technologies. Moving forward, concerted efforts to bolster international collaboration and enhance comprehension and utilization of AI technologies are imperative for fostering novel innovations in healthcare.

PMID:39882522 | PMC:PMC11775008 | DOI:10.3389/fmed.2024.1505692

Heliyon. 2025 Jan 8;11(2):e41802. doi: 10.1016/j.heliyon.2025.e41802. eCollection 2025 Jan 30.

ABSTRACT

Deep Learning (DL) has significantly contributed to the field of medical imaging in recent years, leading to advancements in disease diagnosis and treatment. In the case of Diabetic Retinopathy (DR), DL models have shown high efficacy in tasks such as classification, segmentation, detection, and prediction. However, DL model's opacity and complexity lead to errors in decision-making, particularly in complex cases, making it necessary to estimate the model's uncertainty in predictions. Therefore, there is a need to estimate uncertainty in the model's predictions, which cannot be estimated by classical DL models alone. To address this issue, Bayesian DL methods have been proposed, and their use is increasing in the field. In this paper, we developed a straightforward architecture for the classification of DR using a Convolutional Neural Network (CNN) model. We then applied the Bayesian CNN twice, once using Variational Inference (VI) and once using Monte Carlo dropout (MC-dropout) methods, to the same CNN architecture. This allowed us to gain the posterior predictive distributions for each of them. The performance of the proposed models was evaluated on two benchmark datasets, namely APTOS 2019 and Messidor-2. Experimental findings demonstrated that the proposed models surpassed other state-of-the-art models, achieving a test accuracy of 94.70 % and 77.00 % for CNN, 94.00 % and 86.00 % for BCNN-VI, and 93.30 % and 81.00 % for BCNN-MC-dropout on the APTOS dataset and Messidor-2 dataset, respectively. Finally, we computed the entropy and standard deviation on the obtained predictive distribution to quantify the model uncertainty. This research highlights the potential benefits of using Bayesian DL methods in medical image analysis to improve the accuracy and reliability of diagnosing disease and treatment.

PMID:39882466 | PMC:PMC11774831 | DOI:10.1016/j.heliyon.2025.e41802

Front Neurol. 2025 Jan 15;15:1510522. doi: 10.3389/fneur.2024.1510522. eCollection 2024.

ABSTRACT

BACKGROUND: Alzheimer's disease (AD) is a common neurodegenerative disorder worldwide and the using of magnetic resonance imaging (MRI) in the management of AD is increasing. The present study aims to summarize MRI in AD researches via bibliometric analysis and predict future research hotspots.

METHODS: We searched for records related to MRI studies in AD patients from 2004 to 2023 in the Web of Science Core Collection (WoSCC) database. CiteSpace was applied to analyze institutions, references and keywords. VOSviewer was used for the analysis of countries, authors and journals.

RESULTS: A total of 13,659 articles were obtained in this study. The number of published articles showed overall exponential growth from 2004 to 2023. The top country and institution were the United States and the University of California System, accounting for 40.30% and 9.88% of the total studies, respectively. Jack CR from the United States was the most productive author. The most productive journal was the Journal of Alzheimers Disease. Keyword burst analysis revealed that "machine learning" and "deep learning" were the keywords that frequently appeared in the past 6 years. Timeline views of the references revealed that "#0 tau pathology" and "#1 deep learning" are currently the latest research focuses.

CONCLUSION: This study provides an in-depth overview of publications on MRI studies in AD. The United States is the leading country in this field with a concentration of highly productive researchers and high-level institutions. The current research hotspot is deep learning, which is being applied to develop noninvasive diagnosis and safer treatment of AD.

PMID:39882364 | PMC:PMC11774745 | DOI:10.3389/fneur.2024.1510522

Digit Discov. 2025 Jan 24. doi: 10.1039/d4dd00219a. Online ahead of print.

ABSTRACT

Plastic pollution, particularly microplastics (MPs), poses a significant global threat to ecosystems and human health, necessitating innovative remediation strategies. Biocompatible and biodegradable plastic-binding peptides (PBPs) offer a potential solution through targeted adsorption and subsequent MP detection or removal from the environment. A challenge in discovering plastic-binding peptides is the vast combinatorial space of possible peptides (i.e., over 1015 for 12-mer peptides), which far exceeds the sample sizes typically reachable by experiments or biophysics-based computational methods. One step towards addressing this issue is to train deep learning models on experimental or biophysical datasets, permitting faster and cheaper evaluations of peptides. However, deep learning predictions are not always accurate, which could waste time and money due to synthesizing and evaluating false positives. Here, we resolve this issue by combining biophysical modeling data from Peptide Binder Design (PepBD) algorithm, the predictive power and uncertainty quantification of evidential deep learning, and metaheuristic search methods to identify high-affinity PBPs for several common plastics. Molecular dynamics simulations show that the discovered PBPs have greater median adsorption free energies for polyethylene (5%), polypropylene (18%), and polystyrene (34%) relative to PBPs previously designed by PepBD. The impact of including uncertainty quantification in peptide design is demonstrated by the increasing improvement in the median adsorption free energy with decreasing uncertainty. This robust framework accelerates peptide discovery, paving the way for effective, bio-inspired solutions to MP remediation.

PMID:39882101 | PMC:PMC11771220 | DOI:10.1039/d4dd00219a

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