Deep learning

DeepBindi: An End-to-End Fear Detection System Optimized for Extreme-Edge Deployment

Thu, 2025-07-10 06:00

IEEE J Biomed Health Inform. 2025 Jul 10;PP. doi: 10.1109/JBHI.2025.3587961. Online ahead of print.

ABSTRACT

The growing interest in affective computing has resulted in substantial advancements in emotion recognition through the application of various machine learning and deep learning techniques. Nevertheless, existing methodologies exhibit notable limitations. Specifically, they often fail to address extreme-edge design requirements, making them unfeasible for deployment in wearable systems under real-world conditions. With this aim, this paper introduces a novel end-to-end fear recognition system based on physiological signals designed specifically for deployment in extreme-edge contexts. This solution combines advanced feature engineering techniques with optimized lightweight 1D-CNN model architecture that integrates the advantages of both hand-crafted features and advanced deep-learning convolutional techniques. An experimental validation conducted with the WEMAC dataset provides f1-scores of 80% and accuracy rates of 74%, and reveals significant performance improvements with respect to our previous model proposed: 11.6% and 26.4% in accuracy and F1-score metrics, respectively. Additionally, this research demonstrates the successful integration and validation of the model within an ultra-low-power ARM Cortex-M4 architecture, which exhibits an average power consumption of 16 mW at 5 V, with each inference requiring 496 ms. These results pave the way to a sustainable implementation of deep learning solutions in extreme-edge devices.

PMID:40638343 | DOI:10.1109/JBHI.2025.3587961

Categories: Literature Watch

Research on a deep learning-based model for measurement of X-ray imaging parameters of atlantoaxial joint

Thu, 2025-07-10 06:00

Eur Spine J. 2025 Jul 10. doi: 10.1007/s00586-025-09075-6. Online ahead of print.

ABSTRACT

PURPOSE: To construct a deep learning-based SCNet model, in order to automatically measure X-ray imaging parameters related to atlantoaxial subluxation (AAS) in cervical open-mouth view radiographs, and the accuracy and reliability of the model were evaluated.

METHODS: A total of 1973 cervical open-mouth view radiographs were collected from picture archiving and communication system (PACS) of two hospitals(Hospitals A and B). Among them, 365 images of Hospital A were randomly selected as the internal test dataset for evaluating the model's performance, and the remaining 1364 images of Hospital A were used as the training dataset and validation dataset for constructing the model and tuning the model hyperparameters, respectively. The 244 images of Hospital B were used as an external test dataset to evaluate the robustness and generalizability of our model. The model identified and marked landmarks in the images for the parameters of the lateral atlanto-dental space (LADS), atlas lateral mass inclination (ALI), lateral mass width (LW), axis spinous process deviation distance (ASDD). The measured results of landmarks on the internal test dataset and external test dataset were compared with the mean values of manual measurement by three radiologists as the reference standard. Percentage of correct key-points (PCK), intra-class correlation coefficient (ICC), mean absolute error (MAE), Pearson correlation coefficient (r), mean square error (MSE), root mean square error (RMSE) and Bland-Altman plot were used to evaluate the performance of the SCNet model.

RESULTS: (1) Within the 2 mm distance threshold, the PCK of the SCNet model predicted landmarks in internal test dataset images was 98.6-99.7%, and the PCK in the external test dataset images was 98-100%. (2) In the internal test dataset, for the parameters LADS, ALI, LW, and ASDD, there were strong correlation and consistency between the SCNet model predictions and the manual measurements (ICC = 0.80-0.96, r = 0.86-0.96, MAE = 0.47-2.39 mm/°, MSE = 0.38-8.55 mm2/°2, RMSE = 0.62-2.92 mm/°). (3) The same four parameters also showed strong correlation and consistency between SCNet and manual measurements in the external test dataset (ICC = 0.81-0.91, r = 0.82-0.91, MAE = 0.46-2.29 mm/°, MSE = 0.29-8.23mm2/°2, RMSE = 0.54-2.87 mm/°).

CONCLUSION: The SCNet model constructed based on deep learning algorithm in this study can accurately identify atlantoaxial vertebral landmarks in cervical open-mouth view radiographs and automatically measure the AAS-related imaging parameters. Furthermore, the independent external test set demonstrates that the model exhibits a certain degree of robustness and generalization capability under meet radiographic standards.

PMID:40637839 | DOI:10.1007/s00586-025-09075-6

Categories: Literature Watch

Recurrence prediction of invasive ductal carcinoma from preoperative contrast-enhanced computed tomography using deep convolutional neural network

Thu, 2025-07-10 06:00

Biomed Phys Eng Express. 2025 Jul 10;11(4). doi: 10.1088/2057-1976/adeab5.

ABSTRACT

Predicting the risk of breast cancer recurrence is crucial for guiding therapeutic strategies, including enhanced surveillance and the consideration of additional treatment after surgery. In this study, we developed a deep convolutional neural network (DCNN) model to predict recurrence within six years after surgery using preoperative contrast-enhanced computed tomography (CECT) images, which are widely available and effective for detecting distant metastases. This retrospective study included preoperative CECT images from 133 patients with invasive ductal carcinoma. The images were classified into recurrence and no-recurrence groups using ResNet-101 and DenseNet-201. Classification performance was evaluated using the area under the receiver operating curve (AUC) with leave-one-patient-out cross-validation. At the optimal threshold, the classification accuracies for ResNet-101 and DenseNet-201 were 0.73 and 0.72, respectively. The median (interquartile range) AUC of DenseNet-201 (0.70 [0.69-0.72]) was statistically higher than that of ResNet-101 (0.68 [0.66-0.68]) (p < 0.05). These results suggest the potential of preoperative CECT-based DCNN models to predict breast cancer recurrence without the need for additional invasive procedures.

PMID:40637704 | DOI:10.1088/2057-1976/adeab5

Categories: Literature Watch

Identification of neural crest and melanoma cancer cell invasion and migration genes using high-throughput screening and deep attention networks

Thu, 2025-07-10 06:00

Dev Dyn. 2025 Jul 10. doi: 10.1002/dvdy.70059. Online ahead of print.

ABSTRACT

BACKGROUND: Cell migration and invasion are well-coordinated in development and disease but remain poorly understood. We previously showed that the neural crest (NC) cell migratory wavefront shares a 45-gene panel with other cell invasion phenomena. To rapidly and systematically identify critical genes, we performed a high-throughput siRNA screen and statistical and deep learning analyses to determine changes in NC- versus non-NC-derived human cell line behaviors.

RESULTS: We find 14 out of 45 genes significantly reduced c8161 melanoma cell migration; four of the 14 genes altered leader cell motility (BMP4, ITGB1, KCNE3, and RASGRP1). Deep learning identified marked disruptions in cell-neighbor interactions after BMP4 or RASGRP1 knockdown in c8161 cells. Recombinant proteins added to the culture media revealed five out of the 11 known secreted molecules stimulated c8161 cell migration. BMP4 knockdown severely reduced c8161 in vivo invasion in a chick embryo transplant model. Addition of BMP4 protein to the culture media of BMP4-siRNA-treated c8161 cells rescued cell migratory ability.

CONCLUSION: High-throughput screening and deep learning distilled a 45-gene panel to a small subset of genes critical to melanoma and warrant deeper in vivo functional analysis for their role and potential synergies in driving NC cell migration and invasion.

PMID:40637615 | DOI:10.1002/dvdy.70059

Categories: Literature Watch

Physics informed neural networks simulation of fingering instabilities arising during immiscible and miscible multiphase flow in oil recovery processes

Thu, 2025-07-10 06:00

Chaos. 2025 Jul 1;35(7):073123. doi: 10.1063/5.0273935.

ABSTRACT

Numerical simulation and experimental techniques are the primary methods for solving fluid dynamics problems. However, while numerical simulation approaches are sensitive when meshing a complex structure, experimental methods have difficulty simulating the physical challenges. Therefore, building an affordable model to solve the fluid dynamics problem is very important. Deep learning (DL) approaches have great abilities to handle strong nonlinearity and high dimensionality that attract much attention for solving fluid dynamics problems. In this paper, we used a deep learning-based framework, physics-informed neural networks (PINNs). The main idea of PINN approaches is to encode the underlying physical law (i.e., the partial differential equation) into the neural network as prior information. In the oil recovery process involving the injection of fluids and multiphase flow in porous media, fingering instability is observed if a fluid with low viscosity displaces a high viscosity fluid. This paper provides a deep learning framework to simulate the instability (fingering) phenomenon during secondary and enhanced oil recovery methods. We have considered two models, namely, the first model on immiscible multiphase flow from the secondary oil recovery method, which is analyzed both with and without deliberating mass flow rate. In the second model, instability arising during the enhanced oil recovery method involving miscible displacement of multiphase is examined, focusing on oil recovery using carbonated water injection. We solve the governing nonlinear partial differential equation using PINNs. Furthermore, we have compared results from PINNs with the semi-analytical solution from the literature. The results show that PINNs are very effective in fluid flow problems and deserve further research.

PMID:40637571 | DOI:10.1063/5.0273935

Categories: Literature Watch

DeepNanoHi-C: deep learning enables accurate single-cell nanopore long-read data analysis and 3D genome interpretation

Thu, 2025-07-10 06:00

Nucleic Acids Res. 2025 Jul 8;53(13):gkaf640. doi: 10.1093/nar/gkaf640.

ABSTRACT

Single-cell long-read concatemer sequencing (scNanoHi-C) technology provides unique insights into the higher-order chromatin structure across the genome in individual cells, crucial for understanding 3D genome organization. However, the lack of specialized analytical tools for scNanoHi-C data impedes progress, as existing methods, which primarily focus on scHi-C technologies, do not fully address the specific challenges of scNanoHi-C, such as sparsity, cell-specific variability, and complex chromatin interaction networks. Here, we introduce DeepNanoHi-C, a novel deep learning framework specifically designed for scNanoHi-C data, which leverages a multistep autoencoder and a Sparse Gated Mixture of Experts (SGMoE) to accurately predict chromatin interactions by imputing sparse contact maps, thereby capturing cell-specific structural features. DeepNanoHi-C effectively captures complex global chromatin contact patterns through the multistep autoencoder and dynamically selects the most appropriate expert from a pool of experts based on distinct chromatin contact patterns. Furthermore, DeepNanoHi-C integrates multiscale predictions through a dual-channel prediction net, refining complex interaction information and facilitating comprehensive downstream analyses of chromatin architecture. Experimental validation shows that DeepNanoHi-C outperforms existing methods in distinguishing cell types and demonstrates robust performance in data imputation tasks. Additionally, the framework identifies single-cell 3D genome features, such as cell-specific topologically associating domain (TAD) boundaries, further confirming its ability to accurately model chromatin interactions. Beyond single-cell analysis, DeepNanoHi-C also uncovers conserved genomic structures across species, providing insights into the evolutionary conservation of chromatin organization.

PMID:40637236 | DOI:10.1093/nar/gkaf640

Categories: Literature Watch

Quantum dynamics in multistate harmonic models using tensor-train thermofield dynamics and semiclassical mapping dynamics

Thu, 2025-07-10 06:00

J Chem Phys. 2025 Jul 14;163(2):024131. doi: 10.1063/5.0276946.

ABSTRACT

We present quantum dynamics of the multi-state harmonic (MSH) model using numerically exact tensor-train (TT)-based calculations. The MSH model provides a general framework for mapping a realistic system onto an effective model Hamiltonian, which is defined in extended spatial dimensions, ensuring consistent reorganization energies between all state pairs. Its analytic structure allows efficient propagation of wavepackets via a rank-adaptive TT-KSL scheme and rigorous finite-temperature dynamics via TT-thermofield dynamics. These exact results are used to benchmark various approximate semiclassical and mixed quantum-classical dynamics, including the linearized semiclassical (LSC), symmetrical quasiclassical, classical mapping models (CMMs), mean-field Ehrenfest, and fewest-switches surface hopping dynamics. We systematically explore the parameter space of the MSH model by changing electronic coupling, reorganization energy, reaction free energy, and the nuclear characteristic frequency. In the adiabatic-inverted regime, strong electronic coupling and low reorganization energy lead all approximate methods to converge with the exact TT results. In contrast, discrepancies emerge in the nonadiabatic or normal regimes, where resolution-of-identity LSC and CMMs provide the reliable predictions. This study establishes the MSH model as a powerful tool for validating nonadiabatic dynamics methods in complex condensed-phase systems.

PMID:40637192 | DOI:10.1063/5.0276946

Categories: Literature Watch

Discovering Molecular Insights in Organic Optoelectronics with Knowledge-Informed Interpretable Deep Learning

Thu, 2025-07-10 06:00

J Chem Theory Comput. 2025 Jul 10. doi: 10.1021/acs.jctc.5c00713. Online ahead of print.

ABSTRACT

Deep learning holds significant promise for accelerating molecular screening and materials design. However, the black-box nature of current models limits their ability to generate fundamentally new chemical knowledge and insights. Here, we propose LUMIA (Learning and Understanding Molecular Insights with Artificial Intelligence), an innovative interpretable deep learning framework integrating chemistry-informed contrastive learning and Monte Carlo tree search (MCTS). LUMIA is pretrained on approximately 1.4 million organic molecules, using knowledge-informed augmentations that embed π-conjugation and substituent effects explicitly. This allows it to effectively capture hierarchical molecular representations aligned with chemical intuition. Critically, the explicit integration of chemical knowledge enables LUMIA to achieve state-of-the-art performance across multiple organic optoelectronic property prediction tasks. Leveraging its intrinsic interpretability through MCTS, LUMIA directly uncovers previously unexplored substructure patterns influencing reorganization energy, enabling rational molecular design beyond the training data set. Furthermore, LUMIA reveals novel chemical insights, including synergistic effects of substituent positions in pyrazole derivatives. This study highlights the pivotal role of knowledge embedding in interpretable deep learning, transforming molecular design, and accelerating chemical discovery.

PMID:40637148 | DOI:10.1021/acs.jctc.5c00713

Categories: Literature Watch

Explainable deep learning model WAL-net for individualized assessment of potentially reversible malnutrition in patients with cancer: a multicenter cohort study

Thu, 2025-07-10 06:00

Br J Nutr. 2025 Jul 10:1-40. doi: 10.1017/S000711452510384X. Online ahead of print.

ABSTRACT

Persistent malnutrition is associated with poor clinical outcomes in cancer. However, assessing its reversibility can be challenging. The present study aimed to utilize machine learning (ML) to predict reversible malnutrition (RM) in patients with cancer. A multicenter cohort study including hospitalized oncology patients. Malnutrition was diagnosed using an international consensus. RM was defined as a positive diagnosis of malnutrition upon patient admission which turned negative one month later. Time-series data on body weight and skeletal muscle were modeled using a long short-term memory (LSTM) architecture to predict RM. The model was named as WAL-net, and its performance, explainability, clinical relevance and generalizability were evaluated. We investigated 4254 patients with cancer-associated malnutrition (discovery set=2977, test set=1277). There were 2783 men and 1471 women (median age=61 years). RM was identified in 754 (17.7%) patients. RM/non-RM groups showed distinct patterns of weight and muscle dynamics, and RM was negatively correlated with the progressive stages of cancer cachexia (r=-0.340, P<0.001). WAL-net was the state-of-the-art model among all ML algorithms evaluated, demonstrating favorable performance to predict RM in the test set (AUC=0.924, 95%CI=0.904-0.944) and an external validation set (n=798, AUC=0.909, 95%CI=0.876-0.943). Model-predicted RM using baseline information was associated with lower future risks of underweight, sarcopenia, performance status decline and progression of malnutrition (all P<0.05). This study presents an explainable deep learning model, the WAL-net, for early identification of RM in patients with cancer. These findings might help the management of cancer-associated malnutrition to optimize patient outcomes in multidisciplinary cancer care.

PMID:40637106 | DOI:10.1017/S000711452510384X

Categories: Literature Watch

Segmentation of the Hyperdense Artery Sign on Noncontrast CT in Ischemic Stroke Using Artificial Intelligence

Thu, 2025-07-10 06:00

J Clin Neurol. 2025 Jul;21(4):305-314. doi: 10.3988/jcn.2024.0560.

ABSTRACT

BACKGROUND AND PURPOSE: We developed and validated an automated hyperdense artery sign (HAS) segmentation algorithm for the distal internal carotid artery and middle cerebral artery on noncontrast brain computed tomography (NCCT) using a multicenter dataset with independent annotation performed by two experts.

METHODS: For training and external validation, we included patients with ischemic stroke who underwent concurrent NCCT and CT angiography between May 2011 and December 2022 at six hospitals and one hospital, respectively. For clinical validation, nonoverlapping patients admitted within 24 hours of onset were consecutively included between December 2020 and April 2023 from six hospitals. The model was trained using the 2D U-Net deep-learning architecture with manual annotation by two experts. We constructed models trained on datasets annotated individually by each expert, and an ensemble model using shuffled annotations by both experts. The performance of the models was compared using the area under the receiver operating characteristic curve (AUROC), sensitivity, and specificity.

RESULTS: This study included 673, 365, and 774 patients in the training/internal validation, external validation, and clinical validation datasets, respectively, who were aged 68.8±13.2, 67.8±13.4, and 68.8±13.6 years (mean±standard deviation) and comprised 55.0%, 59.5%, and 57.6% males. The ensemble model achieved higher AUROC and sensitivity than the models trained on annotations by a single expert in the external validation. For the clinical validation dataset, the ensemble model exhibited an AUROC of 0.846 (95% confidence interval [CI], 0.819-0.871), sensitivity of 76.8% (95% CI, 65.1%-86.1%), and specificity of 88.5% (95% CI, 85.9%-90.8%). The predicted volume of the clot was correlated with the infarct volume in follow-up diffusion-weighted imaging (r=0.42, p<0.001).

CONCLUSIONS: Our new algorithm can rapidly and accurately identify the HAS, and so can facilitate the screening of potential patients requiring intervention.

PMID:40635535 | DOI:10.3988/jcn.2024.0560

Categories: Literature Watch

An overview of reliable and representative DVC measurements for musculoskeletal tissues

Thu, 2025-07-10 06:00

J Microsc. 2025 Jul 10. doi: 10.1111/jmi.70008. Online ahead of print.

ABSTRACT

Musculoskeletal tissues present complex hierarchical structures and mechanical heterogeneity across multiple length scales, making them difficult to characterise accurately. Digital volume correlation (DVC) is a non-destructive imaging technique that enables quantification of internal 3D strain fields under realistic loading conditions, offering a unique tool to investigate the biomechanics of biological tissues and biomaterials. This review highlights recent advancements in DVC, focusing on its applications across scales ranging from organ- to tissue-level mechanics in both mineralised and soft tissues. Instead of a traditional systematic review, we identify key technical challenges including the treatment of tissue interfaces, border effects, and the quantification of uncertainty in DVC outputs. Strategies for improving measurement accuracy and reliability are discussed. We also report on the increasing use of DVC in in vivo applications, its coupling with computational modelling to inform and validate biomechanical simulations, and its recent integration with data-driven methods such as deep learning to directly predict displacement and strain fields. Additionally, we examine its application in tissue engineering and implant-tissue interface assessment. By addressing such areas, we outline current limitations and emerging opportunities for future research. These include advancing precision, enabling clinical translation, and leveraging machine learning to create more robust, automated, and predictive DVC workflows for musculoskeletal health and tissue engineering.

PMID:40636996 | DOI:10.1111/jmi.70008

Categories: Literature Watch

Hybrid AI Framework for the Early Detection of Heart Failure: Integrating Traditional Machine Learning and Generative Language Models With Clinical Data

Thu, 2025-07-10 06:00

Cureus. 2025 Jun 9;17(6):e85638. doi: 10.7759/cureus.85638. eCollection 2025 Jun.

ABSTRACT

Cardiovascular disease (CVD) remains the leading cause of mortality globally, necessitating innovative approaches for early detection and risk stratification. This study introduces a hybrid artificial intelligence (AI) model that synergistically combines Convolutional Neural Networks (CNNs) and Large Language Models (LLMs) to enhance the accuracy of heart failure (HF) prediction. The CNN component effectively captures spatial patterns from structured clinical data, while the LLM component interprets complex, unstructured information, enabling a comprehensive analysis of patient health records. Our hybrid model achieved a superior accuracy of 95.1%, outperforming standalone models and demonstrating high precision, recall, F1-score, and area under the receiver operating characteristic curve (AUC-ROC) metrics. Key predictive features (risk factors, symptoms, signs, and electrocardiogram (ECG) investigations) identified include Chest Pain Type, Maximum Heart Rate (maxHR), and Exercise-Induced Angina, aligning with established clinical indicators of cardiac risk. Integrating explainable AI (XAI) techniques, such as Shapley Additive exPlanations (SHAP), provides transparency into the model's decision-making process, fostering trust and facilitating clinical adoption. These findings underscore the potential of hybrid AI models to revolutionize cardiovascular diagnostics by providing accurate, interpretable, and clinically relevant predictions, thereby supporting healthcare professionals in making informed decisions and improving patient outcomes.

PMID:40636608 | PMC:PMC12240570 | DOI:10.7759/cureus.85638

Categories: Literature Watch

Applications of neural networks in liver transplantation

Thu, 2025-07-10 06:00

ILIVER. 2022 Aug 9;1(2):101-110. doi: 10.1016/j.iliver.2022.07.002. eCollection 2022 Jun.

ABSTRACT

The use of neural networks (NNs) as a cutting-edge technique in the medical field has drawn considerable attention. NN models "learn" from a large amount of data and then find corresponding clinical patterns that are challenging for clinicians to recognize. In this study, we focus on liver transplantation (LT), which is an effective treatment for end-stage liver diseases. The management before and after LT produces a massive quantity of medical data, which can be fully processed by NNs. We describe recent progress in the clinical application of NNs to LT in five respects: pre-transplantation evaluation of the donor and recipient, recipient outcome prediction, allocation system development, operation monitoring, and post-transplantation complication prediction. This review provides clinicians and researchers with a description of forefront applications of NNs in the field of LT and discusses prospects and pitfalls.

PMID:40636422 | PMC:PMC12212597 | DOI:10.1016/j.iliver.2022.07.002

Categories: Literature Watch

When liver disease diagnosis encounters deep learning: Analysis, challenges, and prospects

Thu, 2025-07-10 06:00

ILIVER. 2023 Mar 4;2(1):73-87. doi: 10.1016/j.iliver.2023.02.002. eCollection 2023 Mar.

ABSTRACT

The liver is the second-largest organ in the human body and is essential for digesting food and removing toxic substances. Viruses, obesity, alcohol use, and other factors can damage the liver and cause liver disease. The diagnosis of liver disease used to depend on the clinical experience of doctors, which made it subjective, difficult, and time-consuming. Deep learning has made breakthroughs in various fields; thus, there is a growing interest in using deep learning methods to solve problems in liver research to assist doctors in diagnosis and treatment. In this paper, we provide an overview of deep learning in liver research using 139 papers from the last 5 years. We also show the relationship between data modalities, liver topics, and applications in liver research using Sankey diagrams and summarize the deep learning methods used for each liver topic, in addition to the relations and trends between these methods. Finally, we discuss the challenges of and expectations for deep learning in liver research.

PMID:40636411 | PMC:PMC12212720 | DOI:10.1016/j.iliver.2023.02.002

Categories: Literature Watch

Application of biological big data and radiomics in hepatocellular carcinoma

Thu, 2025-07-10 06:00

ILIVER. 2023 Feb 4;2(1):41-49. doi: 10.1016/j.iliver.2023.01.003. eCollection 2023 Mar.

ABSTRACT

Hepatocellular carcinoma (HCC), one of the most common gastrointestinal cancers, has been considered a worldwide threat due to its high incidence and poor prognosis. In recent years, with the continuous emergence and promotion of new sequencing technologies in omics, genomics, transcriptomics, proteomics, and liquid biopsy are used to assess HCC heterogeneity from different perspectives and become a hotspot in the field of tumor precision medicine. In addition, with the continuous improvement of machine learning algorithms and deep learning algorithms, radiomics has made great progress in the field of ultrasound, CT and MRI for HCC. This article mainly reviews the research progress of biological big data and radiomics in HCC, and it provides new methods and ideas for the diagnosis, prognosis, and therapy of HCC.

PMID:40636408 | PMC:PMC12212726 | DOI:10.1016/j.iliver.2023.01.003

Categories: Literature Watch

Fourier convolutional decoder: reconstructing solar flare images via deep learning

Thu, 2025-07-10 06:00

Neural Comput Appl. 2025;37(20):15573-15604. doi: 10.1007/s00521-025-11283-6. Epub 2025 May 27.

ABSTRACT

Reconstructing images from observational data is a complex and time-consuming process, particularly in astronomy, where traditional algorithms like CLEAN require extensive computational resources and expert interpretation to distinguish genuine features from artifacts, especially without ground truth data. To address these challenges, we developed the Fourier convolutional decoder (FCD), a custom-made overcomplete autoencoder trained on simulated data with available ground truth. This enables the network to generate outputs that closely approximate expected ground truth. The model's versatility was demonstrated on both simulated and observational datasets, with a specific application to data from the spectrometer/telescope for imaging X-rays (STIX) on the solar orbiter. In the simulated environment, FCD's performance was evaluated using multiple-image reconstruction metrics, demonstrating its ability to produce accurate images with minimal artifacts. For observational data, FCD was compared with benchmark algorithms, focusing on reconstruction metrics related to Fourier components. Our evaluation found that FCD is the fastest imaging method, with runtimes on the order of milliseconds. Its computational cost is comparable to the most efficient reconstruction algorithm and 280 × faster than the slowest imaging method for single-image reconstruction on a CPU. Additionally, its runtime can be reduced by an order of magnitude for multiple-image reconstruction on a GPU. FCD outperforms or matches state-of-the-art methods on simulated data, achieving a mean MS-SSIM of 0.97, LPIPS of 0.04, PSNR of 35.70 dB, a Dice coefficient of 0.83, and a Hausdorff distance of 5.08. Finally, on experimental STIX observations, FCD remains competitive with top methods despite reduced performance compared to simulated data.

PMID:40636402 | PMC:PMC12234595 | DOI:10.1007/s00521-025-11283-6

Categories: Literature Watch

Deep learning-based feature selection for detection of autism spectrum disorder

Thu, 2025-07-10 06:00

Front Artif Intell. 2025 Jun 25;8:1594372. doi: 10.3389/frai.2025.1594372. eCollection 2025.

ABSTRACT

INTRODUCTION: Autism Spectrum Disorder (ASD) is a neurodevelopmental condition characterized by challenges in communication, social interactions, and repetitive behaviors. The heterogeneity of symptoms across individuals complicates diagnosis. Neuroimaging techniques, particularly resting-state functional MRI (rs-fMRI), have shown potential for identifying neural signatures of ASD, though challenges such as high dimensionality, noise, and small sample sizes hinder their clinical application.

METHODS: This study proposes a novel approach for ASD detection utilizing deep learning and advanced feature selection techniques. A hybrid model combining Stacked Sparse Denoising Autoencoder (SSDAE) and Multi-Layer Perceptron (MLP) is employed to extract relevant features from rs-fMRI data in the ABIDE I dataset, which was preprocessed using the CPAC pipeline. Feature selection is enhanced through an optimized Hiking Optimization Algorithm (HOA) that integrates DynamicOpposites Learning (DOL) and Double Attractors to improve convergence toward the optimal subset of features.

RESULTS: The proposed model is evaluated using multiple ASD datasets. The performance metrics include an average accuracy of 0.735, sensitivity of 0.765, and specificity of 0.752, surpassing the results of existing state-of-the-art methods.

DISCUSSION: The findings demonstrate the effectiveness of the hybrid deep learning approach for ASD detection. The enhanced feature selection process, coupled with the hybrid model, addresses limitations in current neuroimaging analyses and offers a promising direction for more accurate and clinically applicable ASD detection models.

PMID:40636395 | PMC:PMC12237974 | DOI:10.3389/frai.2025.1594372

Categories: Literature Watch

Artificial intelligence in the diagnosis of temporomandibular joint disorders using cone-beam computed tomography (CBCT)

Thu, 2025-07-10 06:00

Bioinformation. 2025 Apr 30;21(4):805-808. doi: 10.6026/973206300210805. eCollection 2025.

ABSTRACT

Temporomandibular joint disorders represent disorders which hinder the proper functioning of TMJ alongside causing pain-related problems. Therefore, it is of interest to analyse 150 CBCT scans using AI integration methods applied for TMD diagnosis. The AI-generated model displayed 92.4% accurate results and 90.8% sensitivity together with 93.7% specificity at a 0.95 AUC that matched radiologist agreement at κ = 0.89. The availability of AI diagnostics cut down diagnostic assessment time to deliver higher efficiency together with greater consistency. The future application of AI-assisted CBCT analysis appears promising yet needs additional verification steps before it becomes clinically available for broader medical use.

PMID:40636185 | PMC:PMC12236576 | DOI:10.6026/973206300210805

Categories: Literature Watch

Advances in artificial intelligence techniques drive the application of radiomics in the clinical research of hepatocellular carcinoma

Thu, 2025-07-10 06:00

ILIVER. 2022 Mar 10;1(1):49-54. doi: 10.1016/j.iliver.2022.02.005. eCollection 2022 Mar.

ABSTRACT

Hepatocellular carcinoma (HCC) remains the most common malignancy to threaten public health globally. With advances in artificial intelligence techniques, radiomics for HCC management provides a novel perspective to solve unmet needs in clinical settings, and reveals pixel-level radiological information for medical imaging big data, correlating the radiological phenotype with targeted clinical issues. Conventional radiomics pipelines depend on handcrafted engineering features, and further deep learning-based radiomics pipelines are supplemented with deep features calculated via self-learning strategies. During the past decade, radiomics has been widely applied in accurate diagnoses and pathological or biological behavior evaluation, as well as in prognosis prediction. In this review, we systematically introduce the main pipelines of artificial intelligence-based radiomics and their efficacy in the clinical studies of HCC.

PMID:40636134 | PMC:PMC12212591 | DOI:10.1016/j.iliver.2022.02.005

Categories: Literature Watch

Accurate classification of benign and malignant breast tumors in ultrasound imaging with an enhanced deep learning model

Thu, 2025-07-10 06:00

Front Bioeng Biotechnol. 2025 Jun 25;13:1526260. doi: 10.3389/fbioe.2025.1526260. eCollection 2025.

ABSTRACT

BACKGROUND: Breast cancer is the most common malignant tumor in women worldwide, and early detection is crucial to improving patient prognosis. However, traditional ultrasound examinations rely heavily on physician judgment, and diagnostic results are easily influenced by individual experience, leading to frequent misdiagnosis or missed diagnosis. Therefore, there is a pressing need for an automated, highly accurate diagnostic method to support the detection and classification of breast cancer. This study aims to build a reliable breast ultrasound image benign and malignant classification model through deep learning technology to improve the accuracy and consistency of diagnosis.

METHODS: This study proposed an innovative deep learning model RcdNet. RcdNet combines deep separable convolution and Convolutional Block Attention Module (CBAM) attention modules to enhance the ability to identify key lesion areas in ultrasound images. The model was internally validated and externally independently tested, and compared with commonly used models such as ResNet, MobileNet, RegNet, ViT and ResNeXt to verify its performance advantage in benign and malignant classification tasks. In addition, the model's attention area was analyzed by heat map visualization to evaluate its clinical interpretability.

RESULTS: The experimental results show that RcdNet outperforms other mainstream deep learning models, including ResNet, MobileNet, and ResNeXt, across all key evaluation metrics. On the external test set, RcdNet achieved an accuracy of 0.9351, a precision of 0.9168, a recall of 0.9495, and an F1-score of 0.9290, demonstrating superior classification performance and strong generalization ability. Furthermore, heat map visualizations confirm that RcdNet accurately attends to clinically relevant features such as tumor edges and irregular structures, aligning well with radiologists' diagnostic focus and enhancing the interpretability and credibility of the model in clinical applications.

CONCLUSION: The RcdNet model proposed in this study performs well in the classification of benign and malignant breast ultrasound images, with high classification accuracy, strong generalization ability and good interpretability. RcdNet can be used as an auxiliary diagnostic tool to help physicians quickly and accurately screen breast cancer, improve the consistency and reliability of diagnosis, and provide strong support for early detection and precise diagnosis and treatment of breast cancer. Future work will focus on integrating RcdNet into real-time ultrasound diagnostic systems and exploring its potential in multi-modal imaging workflows.

PMID:40635689 | PMC:PMC12237964 | DOI:10.3389/fbioe.2025.1526260

Categories: Literature Watch

Pages