IEEE Trans Neural Netw Learn Syst. 2025 Apr 15;PP. doi: 10.1109/TNNLS.2025.3556317. Online ahead of print.

ABSTRACT

Recently, numerous deep learning algorithms have addressed wearable human activity recognition (HAR), but they often struggle with efficient knowledge transfer to lightweight models for mobile devices. Knowledge distillation (KD) is a popular technique for model compression, transferring knowledge from a complex teacher to a compact student. Most existing KD algorithms consider homogeneous architectures, hindering performance in heterogeneous setups. This is an under-explored area in wearable HAR. To bridge this gap, we propose a heterogeneous mutual KD (HMKD) framework for wearable HAR. HMKD establishes mutual learning within the intermediate and output layers of both teacher and student models. To accommodate substantial structural differences between teacher and student, we employ a weighted ensemble feature approach to merge the features from their intermediate layers, enhancing knowledge exchange within them. Experimental results on the HAPT, WISDM, and UCI_HAR datasets show HMKD outperforms ten state-of-the-art KD algorithms in terms of classification accuracy. Notably, with ResNetLSTMaN as the teacher and MLP as the student, HMKD increases by 9.19% in MLP's $F_{1}$ score on the HAPT dataset.

PMID:40232930 | DOI:10.1109/TNNLS.2025.3556317

IEEE Trans Vis Comput Graph. 2025 Apr 15;PP. doi: 10.1109/TVCG.2025.3561091. Online ahead of print.

ABSTRACT

We present FLINT (learning-based FLow estimation and temporal INTerpolation), a novel deep learning-based approach to estimate flow fields for 2D+time and 3D+time scientific ensemble data. FLINT can flexibly handle different types of scenarios with (1) a flow field being partially available for some members (e.g., omitted due to space constraints) or (2) no flow field being available at all (e.g., because it could not be acquired during an experiment). The design of our architecture allows to flexibly cater to both cases simply by adapting our modular loss functions, effectively treating the different scenarios as flow-supervised and flow-unsupervised problems, respectively (with respect to the presence or absence of ground-truth flow). To the best of our knowledge, FLINT is the first approach to perform flow estimation from scientific ensembles, generating a corresponding flow field for each discrete timestep, even in the absence of original flow information. Additionally, FLINT produces high-quality temporal interpolants between scalar fields. FLINT employs several neural blocks, each featuring several convolutional and deconvolutional layers. We demonstrate performance and accuracy for different usage scenarios with scientific ensembles from both simulations and experiments.

PMID:40232923 | DOI:10.1109/TVCG.2025.3561091

IEEE Trans Haptics. 2025 Apr 15;PP. doi: 10.1109/TOH.2025.3561049. Online ahead of print.

ABSTRACT

Tactile sensing is pivotal for enhancing robot manipulation abilities by providing crucial feedback for localized information. However, existing sensors often lack the necessary resolution and bandwidth required for intricate tasks. To address this gap, we introduce VibTac, a novel multi-modal tactile sensing finger designed to offer high-resolution and high-bandwidth tactile sensing simultaneously. VibTac seamlessly integrates vision-based and vibration-based tactile sensing modes to achieve high-resolution and high-bandwidth tactile sensing respectively, leveraging a streamlined human-inspired design for versatility in tasks. This paper outlines the key design elements of VibTac and its fabrication methods, highlighting the significance of the Elastomer Gel Pad (EGP) in its sensing mechanism. The sensor's multi-modal performance is validated through 3D reconstruction and spectral analysis to discern tactile stimuli effectively. In experimental trials, VibTac demonstrates its efficacy by achieving over 90% accuracy in insertion tasks involving objects emitting distinct sounds, such as ethernet connectors. Leveraging vision-based tactile sensing for object localization and employing a deep learning model for "click" sound classification, VibTac showcases its robustness in real-world scenarios. Video of the sensor working can be accessed at https://youtu.be/kmKIUlXGroo.

PMID:40232917 | DOI:10.1109/TOH.2025.3561049

IEEE Trans Image Process. 2025 Apr 15;PP. doi: 10.1109/TIP.2025.3558434. Online ahead of print.

ABSTRACT

Deep learning-based person re-identification (reid) models are widely employed in surveillance systems and inevitably inherit the vulnerability of deep networks to adversarial attacks. Existing attacks merely consider cross-dataset and cross-model transferability, ignoring the cross-test capability to perturb models trained in different domains. To powerfully examine the robustness of real-world re-id models, the Meta Transferable Generative Attack (MTGA) method is proposed, which adopts meta-learning optimization to promote the generative attacker producing highly transferable adversarial examples by learning comprehensively simulated transfer-based crossmodel&dataset&test black-box meta attack tasks. Specifically, cross-model&dataset black-box attack tasks are first mimicked by selecting different re-id models and datasets for meta-train and meta-test attack processes. As different models may focus on different feature regions, the Perturbation Random Erasing module is further devised to prevent the attacker from learning to only corrupt model-specific features. To boost the attacker learning to possess cross-test transferability, the Normalization Mix strategy is introduced to imitate diverse feature embedding spaces by mixing multi-domain statistics of target models. Extensive experiments show the superiority of MTGA, especially in cross-model&dataset and cross-model&dataset&test attacks, our MTGA outperforms the SOTA methods by 20.0% and 11.3% on mean mAP drop rate, respectively. The source codes are available at https://github.com/yuanbianGit/MTGA.

PMID:40232916 | DOI:10.1109/TIP.2025.3558434

ACS Appl Mater Interfaces. 2025 Apr 15. doi: 10.1021/acsami.5c02901. Online ahead of print.

ABSTRACT

Optical phase imaging has become a pivotal tool in biomedical research, enabling label-free visualization of transparent specimens. Traditional optical phase imaging techniques, such as Zernike phase contrast and differential interference contrast microscopy, fall short of providing quantitative phase information. Digital holographic microscopy (DHM) addresses this limitation by offering precise phase measurements; however, off-axis configurations, particularly Mach-Zehnder and Michelson-based setups, are often hindered by environmental susceptibility and bulky optical components due to their separate reference and object beam paths. In this work, we have developed a meta-based interferometric quantitative phase imaging system using a common-path off-axis DHM configuration. A meta-biprism, featuring two opposite gradient phases created using GaN nanopillars selected for their low loss and durability, serves as a compact and efficient beam splitter. Our system effectively captures the complex wavefronts of samples, enabling the retrieval of quantitative phase information, which we demonstrate using standard resolution phase targets and human lung cell lines. Additionally, our system exhibits enhanced temporal phase stability compared to conventional off-axis DHM configurations, reducing phase fluctuations over extended measurement periods. These results not only underline the potential of metasurfaces in advancing the capabilities of quantitative phase imaging but also promise significant advancements in biomedical imaging and diagnostics.

PMID:40233216 | DOI:10.1021/acsami.5c02901

PLoS Genet. 2025 Apr 15;21(4):e1011661. doi: 10.1371/journal.pgen.1011661. Online ahead of print.

ABSTRACT

Most of our understanding of the fundamental processes of mutation and recombination stems from a handful of disparate model organisms and pedigree studies of mammals, with little known about other vertebrates. To gain a broader comparative perspective, we focused on the zebra finch (Taeniopygia castanotis), which, like other birds, differs from mammals in its karyotype (which includes many micro-chromosomes), in the mechanism by which recombination is directed to the genome, and in aspects of ontogenesis. We collected genome sequences from three generation pedigrees that provide information about 80 meioses, inferring 202 single-point de novo mutations, 1,088 crossovers, and 275 non-crossovers. On that basis, we estimated a sex-averaged mutation rate of 5.0 × 10-9 per base pair per generation, on par with mammals that have a similar generation time (~2-3 years). Also as in mammals, we found a paternal germline mutation bias at later stages of gametogenesis (of 1.7:1) but no discernible difference between sexes in early development. Examining recombination patterns, we found that the sex-averaged crossover rate on macro-chromosomes is 0.93 cM/Mb, with a pronounced enrichment of crossovers near telomeres. In contrast, non-crossover rates are more uniformly distributed. On micro-chromosomes, sex-averaged crossover rates are substantially higher (3.96 cM/Mb), in accordance with crossover homeostasis, and both crossover and non-crossover events are more uniformly distributed. At a finer scale, recombination events overlap CpG islands more often than expected by chance, as expected in the absence of PRDM9. Estimates of the degree of GC-biased gene conversion (59%), the mean non-crossover conversion tract length (~32 bp), and the non-crossover-to-crossover ratio (5.4:1) are all comparable to those reported in primates and mice. Therefore, properties of germline mutation and recombination resolutions remain similar over large phylogenetic distances.

PMID:40233115 | DOI:10.1371/journal.pgen.1011661

PLoS Comput Biol. 2025 Apr 15;21(4):e1012298. doi: 10.1371/journal.pcbi.1012298. Online ahead of print.

ABSTRACT

Cellular senescence is known to drive age-related pathology through the senescence-associated secretory phenotype (SASP). However, it also plays important physiological roles such as cancer suppression, embryogenesis and wound healing. Wound healing is a tightly regulated process which when disrupted results in conditions such as fibrosis and chronic wounds. Senescent cells appear during the proliferation phase of the healing process where the SASP is involved in maintaining tissue homeostasis after damage. Interestingly, SASP composition and functionality was recently found to be temporally regulated, with distinct SASP profiles involved: a fibrogenic, followed by a fibrolytic SASP, which could have important implications for the role of senescent cells in wound healing. Given the number of factors at play a full understanding requires addressing the multiple levels of complexity, pertaining to the various cell behaviours, individually followed by investigating the interactions and influence each of these elements have on each other and the system as a whole. Here, a systems biology approach was adopted whereby a multi-scale model of wound healing that includes the dynamics of senescent cell behaviour and corresponding SASP composition within the wound microenvironment was developed. The model was built using the software CompuCell3D, which is based on a Cellular Potts modelling framework. We used an existing body of data on healthy wound healing to calibrate the model and validation was done on known disease conditions. The model clearly shows how differences in the spatiotemporal dynamics of different senescent cell phenotypes lead to several distinct repair outcomes. These differences in senescent cell dynamics can be attributed to variable SASP composition, duration of senescence and temporal induction of senescence relative to the healing stage. The range of outcomes demonstrated strongly highlight the dynamic and heterogenous role of senescent cells in wound healing, fibrosis and chronic wounds, and their fine-tuned control. Further specific data to increase model confidence could be used to explore senolytic treatments in wound disorders.

PMID:40233102 | DOI:10.1371/journal.pcbi.1012298

FASEB J. 2025 Apr 30;39(8):e70541. doi: 10.1096/fj.202403146RR.

ABSTRACT

Injured liver cells undergoing chronic wound healing produce excessive amounts of extracellular matrix (ECM) components, such as collagen and fibronectin, leading to fibrosis. This process is largely mediated by transforming growth factor-β (TGFβ) signaling, which intersects with the tumor suppressor p53 pathway. However, the roles of specific p53 isoforms in this interaction remain unclear. In this study, we report the involvement of the Δ40p53α isoform, an N-terminal truncated variant of p53, in regulating ECM gene expression in TGFβ1-activated LX-2 human hepatic stellate cells. RT-PCR analysis of cirrhotic liver tissues revealed clinically relevant increases in Δ40p53 expression. Knockdown of Δ40p53 using antisense oligonucleotides in LX-2 cells attenuated TGFβ1-induced activation and significantly reduced collagen production and deposition, particularly fibrillar collagen III. Conversely, overexpression of Δ40p53α upregulated collagen III expression in concert with full-length p53 (FLp53). Co-immunoprecipitation analysis demonstrated that Δ40p53α forms a complex with FLp53, which associates with phosphorylated Smad3 following TGFβ1 stimulation. These findings suggest that Δ40p53 enhances collagen III expression by interacting with FLp53 and Smads, highlighting its role in profibrogenic ECM expression.

PMID:40232888 | DOI:10.1096/fj.202403146RR

Cent Nerv Syst Agents Med Chem. 2025 Apr 11. doi: 10.2174/0118715249350698250317041551. Online ahead of print.

ABSTRACT

BACKGROUND: The search for effective treatments for neurodegenerative diseases, particularly Alzheimer's disease, has been fraught with challenges. Alzheimer's disease accounts for 60-80% of dementia cases globally, affecting approximately about 50 million people. Currently, drug repurposing has emerged as a promising strategy in new drug development, attracting significant attention from regulatory agencies, such as the US FDA.

AIM: This study aimed to investigate the potential therapeutic role of dolutegravir in Alzheimer's disease (AD) treatment using a novel network pharmacology approach. Specifically, it explored the interaction of dolutegravir with key molecular targets involved in AD pathology, predicted its effects on relevant biological pathways, and evaluated its viability as a new therapeutic candidate.

OBJECTIVE: This study employed a network pharmacology framework to evaluate dolutegravir, an antiretroviral drug, as a potential treatment for Alzheimer's disease, shedding light on its possible therapeutic mechanisms.

METHOD: A network pharmacology approach was used to predict the drug targets of dolutegravir. Gene Ontology (GO) enrichment and Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway analyses were performed to identify interacting pathways. Additionally, protein- protein interaction (PPI) network analysis was conducted to assess key interactions and molecular docking studies were performed to evaluate the binding affinity of dolutegravir to the predicted targets.

RESULT: PPI network analysis revealed that dolutegravir interacted with several key targets, including BRAF, mTOR, MAPK1, MAPK3, NOS1, BACE1, CAPN1, CASP3, CASP7, CASP8, CHUK, IKBKB, PIK3CA, and PIK3CD. KEGG pathway analysis suggested that dolutegravir could influence amyloid-beta formation, amyloid precursor protein metabolism, and the cellular response to amyloid-beta. Molecular docking results showed the highest binding affinity of dolutegravir for PI3KCD (-8.5 kcal/mol) and MTOR (-8.7 kcal/mol).

CONCLUSION: The findings indicated that dolutegravir holds significant potential in modulating key pathways involved in Alzheimer's disease pathogenesis. These results provide a strong foundation for further investigations into the therapeutic efficacy and safety of dolutegravir in the treatment of Alzheimer's disease. The use of drug repurposing strategies, leveraging Dolutegravir's established pharmacological profile, offers a promising route for accelerated therapeutic development in AD.

PMID:40231534 | DOI:10.2174/0118715249350698250317041551

Protein Pept Lett. 2025 Apr 14. doi: 10.2174/0109298665364226250328084245. Online ahead of print.

ABSTRACT

BACKGROUND: Since the Coronavirus Disease (COVID-19) became a pandemic in late 2019, vaccination remains the primary approach to combating the virus. Nevertheless, the emergence of new variants poses challenges to vaccine efficacy. This study aimed to identify targets within the SARS-CoV-2 spike (S) protein to detect T-cell responses to the five variants of concern from SARS-CoV-2: Alpha, Beta, Delta, Gamma, and Omicron.

METHODS: Here was employed immunoinformatics tools to develop a peptide-based vaccine targeting the spike protein of SARS-CoV-2 and its major variants, including Alpha, Beta, Delta, Gamma, and Omicron. The peptides were screened for antigenicity, toxicity, allergenicity, and physicochemical properties to ensure their safety and efficacy.

RESULTS: The potential T-cell epitopes with high immunogenicity and IFN-γ induction, are essential for a robust immune response by a comprehensive computational analysis. Population coverage analysis revealed significant coverage across diverse geographical regions, with significant efficacy in areas heavily impacted by the pandemic. Molecular docking simulations revealed strong interactions between the selected peptides and major histocompatibility complex class I (MHC-I) molecules, indicating their potential as vaccine candidates.

CONCLUSION: Our study provides a systematic approach to the rational design of a peptide-based vaccine against COVID-19, providing insights for further experimental validation and development of effective vaccines.

PMID:40231512 | DOI:10.2174/0109298665364226250328084245

Mycoses. 2025 Apr;68(4):e70058. doi: 10.1111/myc.70058.

ABSTRACT

BACKGROUND: Pulmonary aspergillosis is a major global health concern, yet its diagnosis remains challenging. Aspergillus-specific IgG measurement is essential for identifying chronic and allergic forms.

OBJECTIVE: This study aimed to evaluate a quantitative method, the ImmunoCAP assay IgG m3 (ICAP) (Phadia-ThermoFisher Scientific, Waltham, USA), a qualitative method, the Aspergillus IgG Western blot kit (Asp-WB) (LDBio Diagnostics, Lyon, France) and a combination of both methods for the diagnosis of Aspergillus lung disease.

METHODS: A retrospective study was conducted at the University Hospital of Marseille, France, during 1 year. Patients undergoing Aspergillus serology were divided into three groups: Group 1 (G1) with ICAP ≥ 40 mgA/L and positive Asp-WB, Group 2 (G2) with ICAP ≥ 40 mgA/L and negative Asp-WB and Group 3 (G3) with ICAP < 40 mgA/L and positive Asp-WB. Data were collected on demographics, underlying diseases, imaging and biological outcomes. Patients were classified according to their Aspergillus lung disease, whether acute pulmonary aspergillosis, chronic pulmonary aspergillosis (CPA), allergic broncho-pulmonary aspergillosis (ABPA), colonisation or Aspergillus sensitisation.

RESULTS: A total of 536 patients were studied: 173 in G1, 204 in G2 and 200 in G3, with 38 patients found in several groups. The primary underlying disease was cystic fibrosis in 44.6% of patients. Twenty-two patients were diagnosed with ABPA. The number of diagnosed ABPA cases in G1 (20; 11.6%) combining positive ICAP and Asp-WB was significantly higher than that found in the groups with a single positive test result (p < 0.001). Fifteen patients were diagnosed with CPA. Isolated positive Western blot (G3) identified five cases of aspergilloma. Significantly fewer Aspergillus lung diseases were diagnosed in isolated positive ICAP G2 (8.8%) than in G1 (53.8%) and G3 (42.5%) (p < 0.001).

CONCLUSIONS: This study highlights the benefits of combining Asp-WB and ICAP for the diagnosis of Aspergillus lung disease and the relatively high false-positive rate in patients with isolated positive ICAP results.

PMID:40231710 | DOI:10.1111/myc.70058

Curr Mol Pharmacol. 2025 Apr 11. doi: 10.2174/0118761429368288250401054301. Online ahead of print.

ABSTRACT

BACKGROUND: Airway mucus hypersecretion is a prominent pathophysiological characteristic observed in chronic obstructive pulmonary disease (COPD), cystic fibrosis, and asthma. It is a significant risk factor for lung dysfunction and impaired quality of life. Therefore, it is crucial to investigate changes in the major genes expressed in the lungs during airway mucus hypersecretion. Such investigations can help to identify genetic targets for the development of effective treatments to manage airway mucus hypersecretion and improve clinical outcomes for those affected by these respiratory disorders.

OBJECTIVE: Our study aims to identify changes in the expression of key genes in the lungs during airway mucus hypersecretion in mice.

METHODS: Thirty male C57BL/6 mice were randomly allocated into two groups. The Pyocyanin (PCN) group was intranasally infected with 25 μl of pyocyanin solution (1 μg/μl), while the phosphate-buffered saline (PBS) group received 25 μl of PBS intranasally once daily. The lung tissue of mice was extracted after 21 days for the purpose of identifying causal genes through a combination of transcriptomic and proteomic analysis. Finally, we validated the differentially expressed proteins using qRT-PCR and western blot.

RESULTS: Our findings revealed significant alterations in 35,268 genes and 7,004 proteins within the lung tissue of mice treated with PCN. Pathway enrichment analysis, utilizing the Kyoto Encyclopedia of Genes and Genomes (KEGG) database, showed that the differentially expressed proteins were mainly associated with apoptosis, galactose metabolism, and asthma, among the overlapping genes and proteins. To validate the results of the transcriptomic and proteomic analyses, we used qRT-PCR to examine the expression levels of fourteen differentially expressed proteins (DEPs), namely Fpr1, Ear1, Lama3, Col19a1, Spag16, Ropn1l, Dnali1, Cfap70, Ear2, Drc1, Ifit3, Lrrc23, Slpi, and Fam166b. Subsequently, we confirmed the expression of Spag16, Dnali1, and Ropn1l by western blotting.

CONCLUSIONS: Our study identified three DEPs, namely Spag16, Dnali1, and Ropn1l, which are closely associated with the movement and organization of cilia. This study provides novel insights for the development of therapeutic interventions targeting airway mucus hypersecretion.

PMID:40231527 | DOI:10.2174/0118761429368288250401054301

Med Biol Eng Comput. 2025 Apr 15. doi: 10.1007/s11517-025-03358-2. Online ahead of print.

ABSTRACT

The aim of this study was to develop a deep learning method for analyzing CT images with varying doses and qualities, aiming to categorize lung lesions into nodules and non-nodules. This study utilized the lung nodule analysis 2016 challenge dataset. Different low-dose CT (LDCT) images, including 10%, 20%, 40%, and 60% levels, were generated from the full-dose CT (FDCT) images. Five different 3D convolutional networks were developed to classify lung nodules from LDCT and reference FDCT images. The models were evaluated using 400 nodule and 400 non-nodule samples. An ensemble model was also developed to achieve a generalizable model across different dose levels. The model achieved an accuracy of 97.0% for nodule classification on FDCT images. However, the model exhibited relatively poor performance (60% accuracy) on LDCT images, indicating that dedicated models should be developed for each low-dose level. Dedicated models for handling LDCT led to dramatic increases in the accuracy of nodule classification. The dedicated low-dose models achieved a nodule classification accuracy of 90.0%, 91.1%, 92.7%, and 93.8% for 10%, 20%, 40%, and 60% of FDCT images, respectively. The accuracy of the deep learning models decreased gradually by almost 7% as LDCT images proceeded from 100 to 10%. However, the ensemble model led to an accuracy of 95.0% when tested on a combination of various dose levels. We presented an ensemble 3D CNN classifier for lesion classification, utilizing both LDCT and FDCT images. This model is able to analyze a combination of CT images with different dose levels and image qualities.

PMID:40232605 | DOI:10.1007/s11517-025-03358-2

Discov Oncol. 2025 Apr 15;16(1):529. doi: 10.1007/s12672-025-02314-8.

ABSTRACT

OBJECTIVE: This study aims to develop and evaluate an advanced deep learning framework for the detection, classification, and localization of lung tumors in computed tomography (CT) scan images.

MATERIALS AND METHODS: The research utilized a dataset of 1608 CT scan images, including 623 cancerous and 985 non-cancerous cases, all carefully labeled for accurate tumor detection, classification (benign or malignant), and localization. The preprocessing involved optimizing window settings, adjusting slice thickness, and applying advanced data augmentation techniques to enhance the model's robustness and generalizability. The proposed model incorporated innovative components such as transformer-based attention layers, adaptive anchor-free mechanisms, and an improved feature pyramid network. These features enabled the model to efficiently handle detection, classification, and localization tasks. The dataset was split into 70% for training, 15% for validation, and 15% for testing. A multi-task loss function was used to balance the three objectives and optimize the model's performance. Evaluation metrics included mean average precision (mAP), intersection over union (IoU), accuracy, precision, and recall.

RESULTS: The proposed model demonstrated outstanding performance, achieving a mAP of 96.26%, IoU of 95.76%, precision of 98.11%, and recall of 98.83% on the test dataset. It outperformed existing models, including You Only Look Once (YOLO)v9 and YOLOv10, with YOLOv10 achieving a mAP of 95.23% and YOLOv9 achieving 95.70%. The proposed model showed faster convergence, better stability, and superior detection capabilities, particularly in localizing smaller tumors. Its multi-task learning framework significantly improved diagnostic accuracy and operational efficiency.

CONCLUSION: The proposed model offers a robust and scalable solution for lung cancer detection, providing real-time inference, multi-task learning, and high accuracy. It holds significant potential for clinical integration to improve diagnostic outcomes and patient care.

PMID:40232589 | DOI:10.1007/s12672-025-02314-8

Gastric Cancer. 2025 Apr 15. doi: 10.1007/s10120-025-01612-y. Online ahead of print.

ABSTRACT

BACKGROUND: Accurate diagnosis of ESD specimens is crucial for managing early gastric cancer. Identifying tumor areas in serially sectioned ESD specimens requires experience and is time-consuming. This study aimed to develop and evaluate a deep learning model for diagnosing ESD specimens.

METHODS: Whole-slide images of 366 ESD specimens of adenocarcinoma were analyzed, with 2257 annotated regions of interest (tumor and muscularis mucosa) and 83,839 patch images. The development set was divided into training and internal validation sets. Tissue segmentation performance was evaluated using the internal validation set. A detection algorithm for tumor and submucosal invasion at the whole-slide image level was developed, and its performance was evaluated using a test set.

RESULTS: The model achieved Dice coefficients of 0.85 and 0.79 for segmentation of tumor and muscularis mucosa, respectively. In the test set, the diagnostic performance of tumor detection, measured by the AUROC, was 0.995, with a specificity of 1.000 and a sensitivity of 0.947. For detecting submucosal invasion, the model achieved an AUROC of 0.981, with a specificity of 0.956 and a sensitivity of 0.907. Pathologists' performance in diagnosing ESD specimens was evaluated with and without assistance from the deep learning model, and the model significantly reduced the mean diagnosis time (747 s without assistance vs. 478 s with assistance, P < 0.001).

CONCLUSION: The deep learning model demonstrated satisfactory performance in tissue segmentation and high accuracy in detecting tumors and submucosal invasion. This model can potentially serve as a screening tool in the histopathological diagnosis of ESD specimens.

PMID:40232558 | DOI:10.1007/s10120-025-01612-y

Gastric Cancer. 2025 Apr 15. doi: 10.1007/s10120-025-01614-w. Online ahead of print.

ABSTRACT

BACKGROUND: We developed and evaluated a skeletal muscle deep-learning (SMDL) model using skeletal muscle computed tomography (CT) imaging to predict the survival of patients with gastric cancer (GC).

METHODS: This multicenter retrospective study included patients who underwent curative resection of GC between April 2008 and December 2020. Preoperative CT images at the third lumbar vertebra were used to develop a Transformer-based SMDL model for predicting recurrence-free survival (RFS) and disease-specific survival (DSS). The predictive performance of the SMDL model was assessed using the area under the curve (AUC) and benchmarked against both alternative artificial intelligence models and conventional body composition parameters. The association between the model score and survival was assessed using Cox regression analysis. An integrated model combining SMDL signature with clinical variables was constructed, and its discrimination and fairness were evaluated.

RESULTS: A total of 1242, 311, and 94 patients were assigned to the training, internal, and external validation cohorts, respectively. The Transformer-based SMDL model yielded AUCs of 0.791-0.943 for predicting RFS and DSS across all three cohorts and significantly outperformed other models and body composition parameters. The model score was a strong independent prognostic factor for survival. Incorporating the SMDL signature into the clinical model resulted in better prognostic prediction performance. The false-negative and false-positive rates of the integrated model were similar across sex and age subgroups, indicating robust fairness.

CONCLUSIONS: The Transformer-based SMDL model could accurately predict survival of GC and identify patients at high risk of recurrence or death, thereby assisting clinical decision-making.

PMID:40232557 | DOI:10.1007/s10120-025-01614-w

Opt Lett. 2025 Apr 15;50(8):2642-2645. doi: 10.1364/OL.557452.

ABSTRACT

Current camera-array-based integral imaging lacks tampering protection, making images vulnerable to falsification and requiring high computational costs. This Letter proposes an alternative 3D integral imaging scheme that ensures clear light field display while enabling tampering detection and self-recovery. Pixel mapping and deep learning co-extract depth and angular data pixel-wisely, regulating the region of interest of 3D light field for initial verification. Multi-viewpoint recovery information is embedded to reconstruct a complete elemental image array. When tampered with, the altered region can be identified and double-recovered. Experiments demonstrate remarkable parallax effects and effective tampering detection with recovery from multiple perspectives.

PMID:40232459 | DOI:10.1364/OL.557452

Opt Lett. 2025 Apr 15;50(8):2558-2561. doi: 10.1364/OL.557170.

ABSTRACT

In this Letter, we report what we believe to be a new type of abruptly autofocusing beams, termed symmetric butterfly Gaussian beams (SBGBs). The proposed beams appear to have a high degree of tunability for their focal position, focal length, focal intensity, and propagation trajectory. In addition, we propose a deep learning-based model for quick and accurate predictions of the propagation properties of SBGBs, achieving an average relative error of no more than 2.1% and being 8000 times faster than that of split-Fourier transform algorithms. This work may open a new platform for optical manipulation, optical communication, and biomedical applications.

PMID:40232438 | DOI:10.1364/OL.557170

Abdom Radiol (NY). 2025 Apr 15. doi: 10.1007/s00261-025-04942-8. Online ahead of print.

ABSTRACT

BACKGROUND AND PURPOSE: Endometriosis affects 5-10% of women of reproductive age. Despite its prevalence, diagnosing endometriosis through imaging remains challenging. Advances in deep learning (DL) are revolutionizing the diagnosis and management of complex medical conditions. This study aims to evaluate DL tools in enhancing the accuracy of multi-sequence MRI-based detection of endometriosis.

METHOD: We gathered a patient cohort from our institutional database, composed of patients with pathologically confirmed endometriosis from 2015 to 2024. We created an age-matched control group that underwent a similar MR protocol without an endometriosis diagnosis. We used sagittal fat-saturated T1-weighted (T1W FS) pre- and post-contrast and T2-weighted (T2W) MRIs. Our dataset was split at the patient level, allocating 12.5% for testing and conducting seven-fold cross-validation on the remainder. Seven abdominal radiologists with experience in endometriosis MRI and complex surgical planning and one women's imaging fellow with specific training in endometriosis MRI reviewed a random selection of images and documented their endometriosis detection.

RESULTS: 395 and 356 patients were included in the case and control groups respectively. The final 3D-DenseNet-121 classifier model demonstrated robust performance. Our findings indicated the most accurate predictions were obtained using T2W, T1W FS pre-, and post-contrast images. Using an ensemble technique on the test set resulted in an F1 Score of 0.881, AUROCC of 0.911, sensitivity of 0.976, and specificity of 0.720. Radiologists achieved 84.48% and 87.93% sensitivity without and with AI assistance in detecting endometriosis. The agreement among radiologists in predicting labels for endometriosis was measured as a Fleiss' kappa of 0.5718 without AI assistance and 0.6839 with AI assistance.

CONCLUSION: This study introduced the first DL model to use multi-sequence MRI on a large cohort, showing results equivalent to human detection by trained readers in identifying endometriosis.

PMID:40232413 | DOI:10.1007/s00261-025-04942-8

Acta Radiol. 2025 Apr 15:2841851251333974. doi: 10.1177/02841851251333974. Online ahead of print.

ABSTRACT

BackgroundAutism spectrum disorder (ASD) is a neurodevelopmental disease marked by a variety of repetitive behaviors and social communication difficulties.PurposeTo develop a generalizable machine learning (ML) classifier that can accurately and effectively predict ASD in children.Material and MethodsThis paper makes use of neuroimaging data from the Autism Brain Imaging Data Exchange (ABIDE I and II) datasets through a combination of structural and functional magnetic resonance imaging data. Several ML models, such as Support Vector Machines (SVM), CatBoost, random forest (RF), and stack classifiers, were tested to demonstrate which model performs the best in ASD classification when used alongside a deep convolutional neural network.ResultsResults showed that stack classifier performed the best among the models, with the highest accuracy of 81.68%, sensitivity of 85.08%, and specificity of 79.13% for ABIDE I, and 81.34%, 83.61%, and 82.21% for ABIDE II, showing its superior ability to identify complex patterns in neuroimaging data. SVM performed poorly across all metrics, showing its limitations in dealing with high-dimensional neuroimaging data.ConclusionThe results show that the application of ML models, especially ensemble approaches like stack classifier, holds significant promise in improving the accuracy with which ASD is detected using neuroimaging and thus shows their potential for use in clinical applications and early intervention strategies.

PMID:40232228 | DOI:10.1177/02841851251333974