Front Neurorobot. 2025 Jan 23;19:1527908. doi: 10.3389/fnbot.2025.1527908. eCollection 2025.

ABSTRACT

Traffic forecasting is crucial for a variety of applications, including route optimization, signal management, and travel time estimation. However, many existing prediction models struggle to accurately capture the spatiotemporal patterns in traffic data due to its inherent nonlinearity, high dimensionality, and complex dependencies. To address these challenges, a short-term traffic forecasting model, Trafficformer, is proposed based on the Transformer framework. The model first uses a multilayer perceptron to extract features from historical traffic data, then enhances spatial interactions through Transformer-based encoding. By incorporating road network topology, a spatial mask filters out noise and irrelevant interactions, improving prediction accuracy. Finally, traffic speed is predicted using another multilayer perceptron. In the experiments, Trafficformer is evaluated on the Seattle Loop Detector dataset. It is compared with six baseline methods, with Mean Absolute Error, Mean Absolute Percentage Error, and Root Mean Square Error used as metrics. The results show that Trafficformer not only has higher prediction accuracy, but also can effectively identify key sections, and has great potential in intelligent traffic control optimization and refined traffic resource allocation.

PMID:39917631 | PMC:PMC11799296 | DOI:10.3389/fnbot.2025.1527908

Bioinformation. 2024 Sep 30;20(9):1512-1515. doi: 10.6026/9732063002001512. eCollection 2024.

ABSTRACT

Artificial intelligence (AI) is being used to diagnose deadly diseases such as cancer. The possible decrease in human error, fast diagnosis, and consistency of judgment are the key incentives for implementing these technologies. Therefore, it is of interest to assess the use of artificial intelligence in cancer diagnosis. Total 200 cancer cases were included with 100 cases each of Breast and lung cancer to evaluate with AI and conventional method by the radiologist. The cancer cases were identified with the application of AI-based machine learning techniques. The sensitivity and specificity check-up was used to assess the effectiveness of both approaches. The obtained data was statistically evaluated. AI has shown higher accuracy, sensitivity and specificity in cancer diagnosis compared to manual method of diagnosis by radiologist.

PMID:39917228 | PMC:PMC11795495 | DOI:10.6026/9732063002001512

Biomed Eng Online. 2025 Feb 6;24(1):11. doi: 10.1186/s12938-025-01341-4.

ABSTRACT

The degeneration of the intervertebral discs in the lumbar spine is the common cause of neurological and physical dysfunctions and chronic disability of patients, which can be stratified into single-(e.g., disc herniation, prolapse, or bulge) and comorbidity-type degeneration (e.g., simultaneous presence of two or more conditions), respectively. A sample of lumbar magnetic resonance imaging (MRI) images from multiple clinical hospitals in China was collected and used in the proposal assessment. We devised a weighted transfer learning framework WDRIV-Net by ensembling four pre-trained models including Densenet169, ResNet101, InceptionV3, and VGG19. The proposed approach was applied to the clinical data and achieved 96.25% accuracy, surpassing the benchmark ResNet101 (87.5%), DenseNet169 (82.5%), VGG19 (88.75%), InceptionV3 (93.75%), and other state-of-the-art (SOTA) ensemble deep learning models. Furthermore, improved performance was observed as well for the metric of the area under the curve (AUC), producing a ≥ 7% increase versus other SOTA ensemble learning, a ≥ 6% increase versus most-studied models, and a ≥ 2% increase versus the baselines. WDRIV-Net can serve as a guide in the initial and efficient type screening of complex degeneration of lumbar intervertebral discs (LID) and assist in the early-stage selection of clinically differentiated treatment options.

PMID:39915867 | DOI:10.1186/s12938-025-01341-4

BMC Musculoskelet Disord. 2025 Feb 7;26(1):126. doi: 10.1186/s12891-025-08356-x.

ABSTRACT

BACKGROUND: Low back pain is the leading cause of disability worldwide with a significant socioeconomic burden; artificial intelligence (AI) has proved to have a great potential in supporting clinical decisions at each stage of the healthcare process. In this article, we have systematically reviewed the available literature on the applications of AI-based Decision Support Systems (DSS) in the clinical prevention and management of Low Back Pain (LBP) due to lumbar degenerative spine disorders.

METHODS: A systematic review of Pubmed and Scopus databases was performed according to the PRISMA statement. Studies reporting the application of DSS to support the prevention and/or management of LBP due to lumbar degenerative diseases were included. The QUADAS-2 tool was utilized to assess the risk of bias in the included studies. The area under the curve (AUC) and accuracy were assessed for each study.

RESULTS: Twenty five articles met the inclusion criteria. Several different machine learning and deep learning algorithms were employed, and their predictive ability on clinical, demographic, psychosocial, and imaging data was assessed. The included studies mainly encompassed three tasks: clinical score definition, clinical assessment, and eligibility prediction and reached AUC scores of 0.93, 0.99 and 0.95, respectively.

CONCLUSIONS: AI-based DSS applications showed a high degree of accuracy in performing a wide set of different tasks. These findings lay the foundation for further research to improve the current understanding and encourage wider adoption of AI in clinical decision-making.

PMID:39915847 | DOI:10.1186/s12891-025-08356-x

J Biomed Sci. 2025 Feb 7;32(1):16. doi: 10.1186/s12929-024-01110-w.

ABSTRACT

Artificial intelligence (AI) has emerged as a transformative force in precision medicine, revolutionizing the integration and analysis of health records, genetics, and immunology data. This comprehensive review explores the clinical applications of AI-driven analytics in unlocking personalized insights for patients with autoimmune rheumatic diseases. Through the synergistic approach of integrating AI across diverse data sets, clinicians gain a holistic view of patient health and potential risks. Machine learning models excel at identifying high-risk patients, predicting disease activity, and optimizing therapeutic strategies based on clinical, genomic, and immunological profiles. Deep learning techniques have significantly advanced variant calling, pathogenicity prediction, splicing analysis, and MHC-peptide binding predictions in genetics. AI-enabled immunology data analysis, including dimensionality reduction, cell population identification, and sample classification, provides unprecedented insights into complex immune responses. The review highlights real-world examples of AI-driven precision medicine platforms and clinical decision support tools in rheumatology. Evaluation of outcomes demonstrates the clinical benefits and impact of these approaches in revolutionizing patient care. However, challenges such as data quality, privacy, and clinician trust must be navigated for successful implementation. The future of precision medicine lies in the continued research, development, and clinical integration of AI-driven strategies to unlock personalized patient care and drive innovation in rheumatology.

PMID:39915780 | DOI:10.1186/s12929-024-01110-w

BMC Med Imaging. 2025 Feb 6;25(1):41. doi: 10.1186/s12880-025-01573-9.

ABSTRACT

BACKGROUND: Vertebral compression fractures (VCFs) are prevalent in the elderly, often caused by osteoporosis or trauma. Differentiating acute from chronic VCFs is vital for treatment planning, but MRI, the gold standard, is inaccessible for some. However, CT, a more accessible alternative, lacks precision. This study aimed to enhance CT's diagnostic accuracy for VCFs using deep transfer learning (DTL) and radiomics.

METHODS: We retrospectively analyzed 218 VCF patients scanned with CT and MRI within 3 days from Oct 2022 to Feb 2024. MRI categorized VCFs. 3D regions of interest (ROIs) from CT scans underwent feature extraction and DTL modeling. Receiver operating characteristic (ROC) analysis evaluated models, with the best fused with radiomic features via LASSO. AUCs compared via Delong test, and clinical utility assessed by decision curve analysis (DCA).

RESULTS: Patients were split into training (175) and test (43) sets. Traditional radiomics with LR yielded AUCs of 0.973 (training) and 0.869 (test). Optimal DTL modeling improved to 0.992 (training) and 0.941 (test). Feature fusion further boosted AUCs to 1.000 (training) and 0.964 (test). DCA validated its clinical significance.

CONCLUSION: The feature fusion model enhances the differential diagnosis of acute and chronic VCFs, outperforming single-model approaches and offering a valuable decision-support tool for patients unable to undergo spinal MRI.

PMID:39915711 | DOI:10.1186/s12880-025-01573-9

Nat Methods. 2025 Feb 6. doi: 10.1038/s41592-025-02599-1. Online ahead of print.

ABSTRACT

Neuroimaging has entered the era of big data. However, the advancement of preprocessing pipelines falls behind the rapid expansion of data volume, causing substantial computational challenges. Here we present DeepPrep, a pipeline empowered by deep learning and a workflow manager. Evaluated on over 55,000 scans, DeepPrep demonstrates tenfold acceleration, scalability and robustness compared to the state-of-the-art pipeline, thereby meeting the scalability requirements of neuroimaging.

PMID:39915693 | DOI:10.1038/s41592-025-02599-1

Commun Med (Lond). 2025 Feb 6;5(1):38. doi: 10.1038/s43856-024-00722-5.

ABSTRACT

BACKGROUND: Over the next 5 years, new breast cancer screening guidelines recommending magnetic resonance imaging (MRI) for certain patients will significantly increase the volume of imaging data to be analyzed. While this increase poses challenges for radiologists, artificial intelligence (AI) offers potential solutions to manage this workload. However, the development of AI models is often hindered by manual annotation requirements and strict data-sharing regulations between institutions.

METHODS: In this study, we present an integrated pipeline combining weakly supervised learning-reducing the need for detailed annotations-with local AI model training via swarm learning (SL), which circumvents centralized data sharing. We utilized three datasets comprising 1372 female bilateral breast MRI exams from institutions in three countries: the United States (US), Switzerland, and the United Kingdom (UK) to train models. These models were then validated on two external datasets consisting of 649 bilateral breast MRI exams from Germany and Greece.

RESULTS: Upon systematically benchmarking various weakly supervised two-dimensional (2D) and three-dimensional (3D) deep learning (DL) methods, we find that the 3D-ResNet-101 demonstrates superior performance. By implementing a real-world SL setup across three international centers, we observe that these collaboratively trained models outperform those trained locally. Even with a smaller dataset, we demonstrate the practical feasibility of deploying SL internationally with on-site data processing, addressing challenges such as data privacy and annotation variability.

CONCLUSIONS: Combining weakly supervised learning with SL enhances inter-institutional collaboration, improving the utility of distributed datasets for medical AI training without requiring detailed annotations or centralized data sharing.

PMID:39915630 | DOI:10.1038/s43856-024-00722-5

Sci Rep. 2025 Feb 6;15(1):4470. doi: 10.1038/s41598-025-88843-2.

ABSTRACT

With the fast growth of artificial intelligence (AI) and a novel generation of network technology, the Internet of Things (IoT) has become global. Malicious agents regularly utilize novel technical vulnerabilities to use IoT networks in essential industries, the military, defence systems, and medical diagnosis. The IoT has enabled well-known connectivity by connecting many services and objects. However, it has additionally made cloud and IoT frameworks vulnerable to cyberattacks, production cybersecurity major concerns, mainly for the growth of trustworthy IoT networks, particularly those empowering smart city systems. Federated Learning (FL) offers an encouraging solution to address these challenges by providing a privacy-preserving solution for investigating and detecting cyberattacks in IoT systems without negotiating data privacy. Nevertheless, the possibility of FL regarding IoT forensics remains mostly unexplored. Deep learning (DL) focused cyberthreat detection has developed as a powerful and effective approach to identifying abnormal patterns or behaviours in the data field. This manuscript presents an Advanced Artificial Intelligence with a Federated Learning Framework for Privacy-Preserving Cyberthreat Detection (AAIFLF-PPCD) approach in IoT-assisted sustainable smart cities. The AAIFLF-PPCD approach aims to ensure robust and scalable cyberthreat detection while preserving the privacy of IoT users in smart cities. Initially, the AAIFLF-PPCD model utilizes Harris Hawk optimization (HHO)-based feature selection to identify the most related features from the IoT data. Next, the stacked sparse auto-encoder (SSAE) classifier is employed for detecting cyberthreats. Eventually, the walrus optimization algorithm (WOA) is used for hyperparameter tuning to improve the parameters of the SSAE approach and achieve optimal performance. The simulated outcome of the AAIFLF-PPCD technique is evaluated using a benchmark dataset. The performance validation of the AAIFLF-PPCD technique exhibited a superior accuracy value of 99.47% over existing models under diverse measures.

PMID:39915579 | DOI:10.1038/s41598-025-88843-2

Sci Rep. 2025 Feb 7;15(1):4569. doi: 10.1038/s41598-025-87290-3.

ABSTRACT

Diabetic Macular Edema (DME) is a major complication of diabetic retinopathy characterized by fluid accumulation in the macula, leading to vision impairment. The standard treatment involves anti-VEGF (Vascular Endothelial Growth Factor) therapy, but approximately 36% of patients do not respond adequately, highlighting the need for more precise predictive models to guide treatment. This study aims to develop a Hybrid Deep Learning model to predict treatment responses in DME patients undergoing anti-VEGF therapy, thereby improving the accuracy of treatment planning and minimizing the unnecessary use of costly anti-VEGF agents. The model integrates both Optical Coherence Tomography (OCT) images and clinical data from 181 patients, including key parameters such as serum VEGFR-2 concentration and the duration of DME. The architecture combines convolutional neural networks (CNNs) for image data with multi-layer perceptron (MLP) for tabular clinical data, allowing for a comprehensive analysis of both data types. These pathways converge into a unified predictive framework designed to enhance the model's accuracy. This study utilized a Hybrid Deep Learning model that achieved an 85% accuracy, with additional metrics including precision, recall, and the area under the receiver operating characteristic curve (AUC-ROC) confirming its robustness and reliability. The findings suggest that the model accurately predicts patient responses to anti-VEGF therapy, paving the way for more personalized and targeted treatment strategies. This approach has the potential to enhance patient outcomes and minimize unnecessary administration of anti-VEGF agents, thereby optimizing therapeutic interventions in ophthalmology.

PMID:39915516 | DOI:10.1038/s41598-025-87290-3

Nat Commun. 2025 Feb 7;16(1):1425. doi: 10.1038/s41467-025-56784-z.

ABSTRACT

Treatment decisions for an incidental renal mass are mostly made with pathologic uncertainty. Improving the diagnosis of benign renal masses and distinguishing aggressive cancers from indolent ones is key to better treatment selection. We analyze 13261 pre-operative computed computed tomography (CT) volumes of 4557 patients. Two multi-phase convolutional neural networks are developed to predict the malignancy and aggressiveness of renal masses. The first diagnostic model designed to predict the malignancy of renal masses achieves area under the curve (AUC) of 0.871 in the prospective test set. This model surpasses the average performance of seven seasoned radiologists. The second diagnostic model differentiating aggressive from indolent tumors has AUC of 0.783 in the prospective test set. Both models outperform corresponding radiomics models and the nephrometry score nomogram. Here we show that the deep learning models can non-invasively predict the likelihood of malignant and aggressive pathology of a renal mass based on preoperative multi-phase CT images.

PMID:39915478 | DOI:10.1038/s41467-025-56784-z

Light Sci Appl. 2025 Feb 7;14(1):70. doi: 10.1038/s41377-024-01721-w.

ABSTRACT

To reveal the fundamental aspects hidden behind a variety of transient events in mechanics, physics, and biology, the highly desired ability to acquire three-dimensional (3D) images with ultrafast temporal resolution has been long sought. As one of the most commonly employed 3D sensing techniques, fringe projection profilometry (FPP) reconstructs the depth of a scene from stereo images taken with sequentially structured illuminations. However, the imaging speed of current FPP methods is generally capped at several kHz, which is limited by the projector-camera hardware and the number of fringe patterns required for phase retrieval and unwrapping. Here we report a novel learning-based ultrafast 3D imaging technique, termed single-shot super-resolved FPP (SSSR-FPP), which enables ultrafast 3D imaging at 100,000 Hz. SSSR-FPP uses only one pair of low signal-to-noise ratio (SNR), low-resolution, and pixelated fringe patterns as input, while the high-resolution unwrapped phase and fringe orders can be deciphered with a specific trained deep neural network. Our approach exploits the significant speed gain achieved by reducing the imaging window of conventional high-speed cameras, while "regenerating" the lost spatial resolution through deep learning. To demonstrate the high spatio-temporal resolution of SSSR-FPP, we present 3D videography of several transient scenes, including rotating turbofan blades, exploding building blocks, and the reciprocating motion of a steam engine, etc., which were previously challenging or even impossible to capture with conventional methods. Experimental results establish SSSR-FPP as a significant step forward in the field of 3D optical sensing, offering new insights into a broad spectrum of dynamic processes across various scientific disciplines.

PMID:39915449 | DOI:10.1038/s41377-024-01721-w

J Med Internet Res. 2025 Feb 6;27:e66072. doi: 10.2196/66072.

ABSTRACT

BACKGROUND: Mumps is a viral respiratory disease characterized by facial swelling and transmitted through respiratory secretions. Despite the availability of an effective vaccine, mumps outbreaks have reemerged globally, including in China, where it remains a significant public health issue. In Yunnan province, China, the incidence of mumps has fluctuated markedly and is higher than that in mainland China, underscoring the need for improved outbreak prediction methods. Traditional surveillance methods, however, may not be sufficient for timely and accurate outbreak prediction.

OBJECTIVE: Our study aims to leverage the Baidu search index, representing search volumes from China's most popular search engine, along with environmental data to develop a predictive model for mumps incidence in Yunnan province.

METHODS: We analyzed mumps incidence in Yunnan Province from 2014 to 2023, and used time series data, including mumps incidence, Baidu search index, and environmental factors, from 2016 to 2023, to develop predictive models based on long short-term memory networks. Feature selection was conducted using Pearson correlation analysis, and lag correlations were explored through a distributed nonlinear lag model (DNLM). We constructed four models with different combinations of predictors: (1) model BE, combining the Baidu index and environmental factors data; (2) model IB, combining mumps incidence and Baidu index data; (3) model IE, combining mumps incidence and environmental factors; and (4) model IBE, integrating all 3 data sources.

RESULTS: The incidence of mumps in Yunnan showed significant variability, peaking at 37.5 per 100,000 population in 2019. From 2014 to 2023, the proportion of female patients ranged from 41.3% in 2015 to 45.7% in 2020, consistently lower than that of male patients. After excluding variables with a Pearson correlation coefficient of <0.10 or P values of <.05, we included 3 Baidu index search term groups (disease name, symptoms, and treatment) and 6 environmental factors (maximum temperature, minimum temperature, sulfur dioxide, carbon monoxide, particulate matter with a diameter of 2.5 µm or less, and particulate matter with a diameter of 10 µm or less) for model development. DNLM analysis revealed that the relative risks consistently increased with rising Baidu index values, while nonlinear associations between temperature and mumps incidence were observed. Among the 4 models, model IBE exhibited the best performance, achieving the coefficient of determination of 0.72, with mean absolute error, mean absolute percentage error, and root-mean-square error values of 0.33, 15.9%, and 0.43, respectively, in the test set.

CONCLUSIONS: Our study developed model IBE to predict the incidence of mumps in Yunnan province, offering a potential tool for early detection of mumps outbreaks. The performance of model IBE underscores the potential of integrating search engine data and environmental factors to enhance mumps incidence forecasting. This approach offers a promising tool for improving public health surveillance and enabling rapid responses to mumps outbreaks.

PMID:39913179 | DOI:10.2196/66072

Transl Vis Sci Technol. 2025 Feb 3;14(2):11. doi: 10.1167/tvst.14.2.11.

ABSTRACT

PURPOSE: To evaluate the performance of various approaches of processing three-dimensional (3D) optical coherence tomography (OCT) images for deep learning models in predicting area and future growth rate of geographic atrophy (GA) lesions caused by age-related macular degeneration (AMD).

METHODS: The study used OCT volumes of GA patients/eyes from the lampalizumab clinical trials (NCT02247479, NCT02247531, NCT02479386); 1219 and 442 study eyes for model development and holdout performance evaluation, respectively. Four approaches were evaluated: (1) en-face intensity maps; (2) SLIVER-net; (3) a 3D convolutional neural network (CNN); and (4) en-face layer thickness and between-layer intensity maps from a segmentation model. The processed OCT images and maps served as input for CNN models to predict baseline GA lesion area size and annualized growth rate.

RESULTS: For the holdout dataset, the Pearson correlation coefficient squared (r2) in the GA growth rate prediction was comparable for all the evaluated approaches (0.33∼0.35). In baseline lesion size prediction, prediction performance was comparable (0.9∼0.91) except for the SLIVER-net (0.83). Prediction performance with only the thickness map of the ellipsoid zone (EZ) or retinal pigment epithelium (RPE) layer individually was inferior to using both. Addition of other layer thickness or intensity maps did not improve the prediction performance.

CONCLUSIONS: All explored approaches had comparable performance, which might have reached a plateau to predict GA growth rate. EZ and RPE layers appear to contain the majority of information related to the prediction.

TRANSLATIONAL RELEVANCE: Our study provides important insights on the utility of 3D OCT images for GA disease progression predictions.

PMID:39913124 | DOI:10.1167/tvst.14.2.11

Eur Radiol Exp. 2025 Feb 6;9(1):15. doi: 10.1186/s41747-025-00554-5.

ABSTRACT

BACKGROUND: Gadolinium-enhanced "sampling perfection with application-optimized contrasts using different flip angle evolution" (SPACE) sequence allows better visualization of brain metastases (BMs) compared to "magnetization-prepared rapid acquisition gradient echo" (MPRAGE). We hypothesize that this better conspicuity leads to high-quality annotation (HAQ), enhancing deep learning (DL) algorithm detection of BMs on MPRAGE images.

METHODS: Retrospective contrast-enhanced (gadobutrol 0.1 mmol/kg) SPACE and MPRAGE data of 157 patients with BM were used, either annotated on MPRAGE resulting in normal annotation quality (NAQ) or on coregistered SPACE resulting in HAQ. Multiple DL methods were developed with NAQ or HAQ using either SPACE or MRPAGE images and evaluated on their detection performance using positive predictive value (PPV), sensitivity, and F1 score and on their delineation performance using volumetric Dice similarity coefficient, PPV, and sensitivity on one internal and four additional test datasets (660 patients).

RESULTS: The SPACE-HAQ model reached 0.978 PPV, 0.882 sensitivity, and 0.916 F1-score. The MPRAGE-HAQ reached 0.867, 0.839, and 0.840, the MPRAGE NAQ 0.964, 0.667, and 0.798, respectively (p ≥ 0.157). Relative to MPRAGE-NAQ, the MPRAGE-HAQ F1-score detection increased on all additional test datasets by 2.5-9.6 points (p < 0.016) and sensitivity improved on three datasets by 4.6-8.5 points (p < 0.001). Moreover, volumetric instance sensitivity improved by 3.6-7.6 points (p < 0.001).

CONCLUSION: HAQ improves DL methods without specialized imaging during application time. HAQ alone achieves about 40% of the performance improvements seen with SPACE images as input, allowing for fast and accurate, fully automated detection of small (< 1 cm) BMs.

RELEVANCE STATEMENT: Training with higher-quality annotations, created using the SPACE sequence, improves the detection and delineation sensitivity of DL methods for the detection of brain metastases (BMs)on MPRAGE images. This MRI cross-technique transfer learning is a promising way to increase diagnostic performance.

KEY POINTS: Delineating small BMs on SPACE MRI sequence results in higher quality annotations than on MPRAGE sequence due to enhanced conspicuity. Leveraging cross-technique ground truth annotations during training improved the accuracy of DL models in detecting and segmenting BMs. Cross-technique annotation may enhance DL models by integrating benefits from specialized, time-intensive MRI sequences while not relying on them. Further validation in prospective studies is needed.

PMID:39913077 | DOI:10.1186/s41747-025-00554-5

J Imaging Inform Med. 2025 Feb 6. doi: 10.1007/s10278-025-01429-2. Online ahead of print.

ABSTRACT

Prostate Cancer (PCa) is the second most common cancer in men and affects more than a million people each year. Grading prostate cancer is based on the Gleason grading system, a subjective and labor-intensive method for evaluating prostate tissue samples. The variability in diagnostic approaches underscores the urgent need for more reliable methods. By integrating deep learning technologies and developing automated systems, diagnostic precision can be improved, and human error minimized. The present work introduces a three-stage framework-based innovative deep-learning system for assessing PCa severity using the PANDA challenge dataset. After a meticulous selection process, 2699 usable cases were narrowed down from the initial 5160 cases after extensive data cleaning. There are three stages in the proposed framework: classification of PCa grades using deep neural networks (DNNs), segmentation of PCa grades, and computation of International Society for Urological Pathology (ISUP) grades using machine learning classifiers. Four classes of patches were classified and segmented (benign, Gleason 3, Gleason 4, and Gleason 5). Patch sampling at different sizes (500 × 500 and 1000 × 1000 pixels) was used to optimize the classification and segmentation processes. The segmentation performance of the proposed network is enhanced by a Self-organized operational neural network (Self-ONN) based DeepLabV3 architecture. Based on these predictions, the distribution percentages of each cancer grade within the whole slide images (WSI) were calculated. These features were then concatenated into machine learning classifiers to predict the final ISUP PCa grade. EfficientNet_b0 achieved the highest F1-score of 83.83% for classification, while DeepLabV3 + architecture based on self-ONN and EfficientNet encoder achieved the highest Dice Similarity Coefficient (DSC) score of 84.9% for segmentation. Using the RandomForest (RF) classifier, the proposed framework achieved a quadratic weighted kappa (QWK) score of 0.9215. Deep learning frameworks are being developed to grade PCa automatically and have shown promising results. In addition, it provides a prospective approach to a prognostic tool that can produce clinically significant results efficiently and reliably. Further investigations are needed to evaluate the framework's adaptability and effectiveness across various clinical scenarios.

PMID:39913023 | DOI:10.1007/s10278-025-01429-2

Eur J Nucl Med Mol Imaging. 2025 Feb 6. doi: 10.1007/s00259-025-07122-4. Online ahead of print.

ABSTRACT

PURPOSE: Whole-body PET imaging plays an essential role in cancer diagnosis and treatment but suffers from low image quality. Traditional deep learning-based denoising methods work well for a specific acquisition but are less effective in handling diverse PET protocols. In this study, we proposed and validated a 3D Denoising Diffusion Probabilistic Model (3D DDPM) as a robust and universal solution for whole-body PET image denoising.

METHODS: The proposed 3D DDPM gradually injected noise into the images during the forward diffusion phase, allowing the model to learn to reconstruct the clean data during the reverse diffusion process. A 3D convolutional network was trained using high-quality data from the Biograph Vision Quadra PET/CT scanner to generate the score function, enabling the model to capture accurate PET distribution information extracted from the total-body datasets. The trained 3D DDPM was evaluated on datasets from four scanners, four tracer types, and six dose levels representing a broad spectrum of clinical scenarios.

RESULTS: The proposed 3D DDPM consistently outperformed 2D DDPM, 3D UNet, and 3D GAN, demonstrating its superior denoising performance across all tested conditions. Additionally, the model's uncertainty maps exhibited lower variance, reflecting its higher confidence in its outputs.

CONCLUSIONS: The proposed 3D DDPM can effectively handle various clinical settings, including variations in dose levels, scanners, and tracers, establishing it as a promising foundational model for PET image denoising. The trained 3D DDPM model of this work can be utilized off the shelf by researchers as a whole-body PET image denoising solution. The code and model are available at https://github.com/Miche11eU/PET-Image-Denoising-Using-3D-Diffusion-Model .

PMID:39912940 | DOI:10.1007/s00259-025-07122-4

Eur J Nucl Med Mol Imaging. 2025 Feb 6. doi: 10.1007/s00259-025-07086-5. Online ahead of print.

ABSTRACT

PURPOSE: To address the challenges of verifying MR-based attenuation correction (MRAC) in PET/MR due to CT positional mismatches and alignment issues, this study utilized a flatbed insert and arms-down positioning during PET/CT scans to achieve precise MR-CT matching for accurate MRAC evaluation.

METHODS: A validation dataset of 21 patients underwent whole-body [18F]FDG PET/CT followed by [18F]FDG PET/MR. A flatbed insert ensured consistent positioning, allowing direct comparison of four MRAC methods-four-tissue and five-tissue models with discrete and continuous μ-maps-against CT-based attenuation correction (CTAC). A deep learning-based framework, trained on a dataset of 300 patients, was used to generate synthesized-CTs from MR images, forming the basis for all MRAC methods. Quantitative analyses were conducted at the whole-body, region of interest, and lesion levels, with lesion-distance analysis evaluating the impact of bone proximity on standardized uptake value (SUV) quantification.

RESULTS: Distinct differences were observed among MRAC methods in spine and femur regions. Joint histogram analysis showed MRAC-4 (continuous μ-map) closely aligned with CTAC. Lesion-distance analysis revealed MRAC-4 minimized bone-induced SUV interference (r = 0.01, p = 0.8643). However, tissues prone to bone segmentation interference, such as the spine and liver, exhibited greater SUV variability and lower reproducibility in MRAC-4 compared to MRAC-2 (2D bone segmentation, discrete μ-map) and MRAC-3 (3D bone segmentation, discrete μ-map).

CONCLUSION: Using a flatbed insert, this study validated MRAC with high precision. Continuous μ-value MRAC method (MRAC-4) demonstrated superior accuracy and minimized bone-related SUV errors but faced challenges in reproducibility, particularly in bone-rich regions.

PMID:39912939 | DOI:10.1007/s00259-025-07086-5

Rev Sci Instrum. 2025 Feb 1;96(2):025105. doi: 10.1063/5.0233552.

ABSTRACT

Electromagnetic tomography (EMT) has great application potential in fields such as industrial inspection. However, currently, EMT image reconstruction has problems of being highly nonlinear and ill-posed, resulting in artifacts in reconstructed images and uncertainties in quality, detail accuracy, and robustness. To address these challenges, a deep learning model STDBOGAN (generative adversarial network based on spectral normalization, two timescale update rule, and improved dung beetle optimization algorithm) based on generative adversarial networks is proposed. STDBOGAN introduces spectral normalization and two timescale update rules to stabilize the training process and avoid training instability and gradient problems. The improved dung beetle optimization algorithm automatically adjusts network hyper-parameters to improve image reconstruction accuracy. A dataset is established through simulation software. Ablation studies are conducted on the network before and after improvement, and simulations and metal physical experiments are carried out to compare STDBOGAN to UNet3+, DeepLabv3+, PSPNet, Segmenter, and SegRefiner networks. Experiments show that STDBOGAN has the best performance, anti-noise, and generalization abilities, and has made contributions to improving the quality of electromagnetic tomography image reconstruction.

PMID:39912879 | DOI:10.1063/5.0233552

JMIR AI. 2025 Feb 6;4:e60847. doi: 10.2196/60847.

ABSTRACT

BACKGROUND: The rapid advancement of deep learning in health care presents significant opportunities for automating complex medical tasks and improving clinical workflows. However, widespread adoption is impeded by data privacy concerns and the necessity for large, diverse datasets across multiple institutions. Federated learning (FL) has emerged as a viable solution, enabling collaborative artificial intelligence model development without sharing individual patient data. To effectively implement FL in health care, robust and secure infrastructures are essential. Developing such federated deep learning frameworks is crucial to harnessing the full potential of artificial intelligence while ensuring patient data privacy and regulatory compliance.

OBJECTIVE: The objective is to introduce an innovative FL infrastructure called the Personal Health Train (PHT) that includes the procedural, technical, and governance components needed to implement FL on real-world health care data, including training deep learning neural networks. The study aims to apply this federated deep learning infrastructure to the use case of gross tumor volume segmentation on chest computed tomography images of patients with lung cancer and present the results from a proof-of-concept experiment.

METHODS: The PHT framework addresses the challenges of data privacy when sharing data, by keeping data close to the source and instead bringing the analysis to the data. Technologically, PHT requires 3 interdependent components: "tracks" (protected communication channels), "trains" (containerized software apps), and "stations" (institutional data repositories), which are supported by the open source "Vantage6" software. The study applies this federated deep learning infrastructure to the use case of gross tumor volume segmentation on chest computed tomography images of patients with lung cancer, with the introduction of an additional component called the secure aggregation server, where the model averaging is done in a trusted and inaccessible environment.

RESULTS: We demonstrated the feasibility of executing deep learning algorithms in a federated manner using PHT and presented the results from a proof-of-concept study. The infrastructure linked 12 hospitals across 8 nations, covering 4 continents, demonstrating the scalability and global reach of the proposed approach. During the execution and training of the deep learning algorithm, no data were shared outside the hospital.

CONCLUSIONS: The findings of the proof-of-concept study, as well as the implications and limitations of the infrastructure and the results, are discussed. The application of federated deep learning to unstructured medical imaging data, facilitated by the PHT framework and Vantage6 platform, represents a significant advancement in the field. The proposed infrastructure addresses the challenges of data privacy and enables collaborative model development, paving the way for the widespread adoption of deep learning-based tools in the medical domain and beyond. The introduction of the secure aggregation server implied that data leakage problems in FL can be prevented by careful design decisions of the infrastructure.

TRIAL REGISTRATION: ClinicalTrials.gov NCT05775068; https://clinicaltrials.gov/study/NCT05775068.

PMID:39912580 | DOI:10.2196/60847