Light Sci Appl. 2025 Aug 11;14(1):267. doi: 10.1038/s41377-025-01971-2.

ABSTRACT

We present a broadband and polarization-insensitive unidirectional imager that operates at the visible part of the spectrum, where image formation occurs in one direction, while in the opposite direction, it is blocked. This approach is enabled by deep learning-driven diffractive optical design with wafer-scale nano-fabrication using high-purity fused silica to ensure optical transparency and thermal stability. Our design achieves unidirectional imaging across three visible wavelengths (covering red, green, and blue parts of the spectrum), and we experimentally validated this broadband unidirectional imager by creating high-fidelity images in the forward direction and generating weak, distorted output patterns in the backward direction, in alignment with our numerical simulations. This work demonstrates wafer-scale production of diffractive optical processors, featuring 16 levels of nanoscale phase features distributed across two axially aligned diffractive layers for visible unidirectional imaging. This approach facilitates mass-scale production of ~0.5 billion nanoscale phase features per wafer, supporting high-throughput manufacturing of hundreds to thousands of multi-layer diffractive processors suitable for large apertures and parallel processing of multiple tasks. Beyond broadband unidirectional imaging in the visible spectrum, this study establishes a pathway for artificial-intelligence-enabled diffractive optics with versatile applications, signaling a new era in optical device functionality with industrial-level, massively scalable fabrication.

PMID:40789836 | DOI:10.1038/s41377-025-01971-2

Annu Rev Biomed Data Sci. 2025 Aug;8(1):149-171. doi: 10.1146/annurev-biodatasci-103123-095814.

ABSTRACT

The rapid development of artificial intelligence (AI) has had a significant impact on medical research, introducing new possibilities for pathology studies. There is a recent trend of applying large-scale AI models to many fields, and this trend has given rise to the pathology foundation models and pathology ensemble models. Large models in pathology are not standalone innovations; they build upon a legacy where AI has consistently played a vital role in pathology studies long before their advent. Numerous pathology datasets and AI models have been developed to support advancements in the field, with these combined efforts paving the way for the emergence of large models in pathology. AI greatly enhances pathology studies, yet its widespread use in sensitive applications also raises significant ethical concerns, including privacy risks. In this review, we summarize the datasets and models that are useful to pathology studies, with a particular focus on how they illuminate the path toward large-scale applications.

PMID:40789736 | DOI:10.1146/annurev-biodatasci-103123-095814

Annu Rev Biomed Data Sci. 2025 Aug;8(1):101-124. doi: 10.1146/annurev-biodatasci-103023-050856.

ABSTRACT

Deep learning and artificial intelligence (AI) have seen explosive growth and success in biomedical applications in the last decade, largely due to the rapid development of deep neural networks and their underlying neural network (NN) architectures. Here, we explore biomedical deep learning and AI from the specific perspective of NN architectures. We discuss widely varying design principles of NN architectures, their use in particular biomedical applications, and the assumptions (often hidden) built into them. We explore neural architecture search techniques that automate the design of NN topology to optimize task performance. Advanced neural architectures are being developed for both molecular and healthcare applications, employing elements of graph networks, transformers, and interpretable NNs, and we discuss and summarize the design considerations and unique advantages of each architecture. Future advances will include the employment of multimodal language models and smaller highly focused mechanistic models that build on the success of today's large models.

PMID:40789735 | DOI:10.1146/annurev-biodatasci-103023-050856

SLAS Technol. 2025 Aug 9:100338. doi: 10.1016/j.slast.2025.100338. Online ahead of print.

ABSTRACT

Optic disc and cup segmentation is a crucial subfield of computer vision, playing a pivotal role in automated pathological image analysis. It enables precise, efficient, and automated diagnosis of ocular conditions, significantly aiding clinicians in real-world medical applications. However, due to the scarcity of medical segmentation data and the insufficient integration of global contextual information, the segmentation accuracy remains suboptimal. This issue becomes particularly pronounced in optic disc and cup cases with complex anatomical structures and ambiguous boundaries.In order to address these limitations, this paper introduces a self-supervised training strategy integrated with a newly designed network architecture to improve segmentation accuracy.Specifically,we initially propose a non-local dual deformable convolutional block,which aims to capture the irregular image patterns(i.e. boundary).Secondly,we modify the traditional vision transformer and design an adaptive K-Nearest Neighbors(KNN) transformation block to extract the global semantic context from images. Finally,an initialization strategy based on self-supervised training is proposed to reduce the burden on the network on labeled data.Comprehensive experimental evaluations demonstrate the effectiveness of our proposed method, which outperforms previous networks and achieves state-of-the-art performance,with IOU scores of 0.9577 for the optic disc and 0.8399 for the optic cup on the REFUGE dataset.

PMID:40789537 | DOI:10.1016/j.slast.2025.100338

Ann Vasc Surg. 2025 Aug 9:S0890-5096(25)00526-6. doi: 10.1016/j.avsg.2025.08.001. Online ahead of print.

ABSTRACT

INTRODUCTION: Decreased kidney volume is a sign of renal aging and/or decreased vascularization. The aim of this study was to determine whether renal volume changes 24 months after exclusion of an abdominal aortic aneurysm (AAA), and to compare fenestrated (FEVAR) and subrenal (EVAR) stents.

METHODS: Retrospective single-center study from a prospective registry, including patients between 60 and 80 years with normal preoperative renal function (eGFR≥60 ml/min/1.73 m-2) who underwent fenestrated (FEVAR) or infrarenal (EVAR) stent grafts between 2015 and 2021. Patients had to have had an CT scan at 24 months of the study to be included. Exclusion criteria were renal branches, the presence of preoperative renal insufficiency, a single kidney, embolization or coverage of an accessory renal artery, occlusion of a renal artery during follow-up and mention of AAA rupture. Renal volume was measured using sizing software (EndoSize, therenva) based on fully automatic deep-learning segmentation of several anatomical structures (arterial lumen, bone structure, thrombus, heart, etc.), including the kidneys. In the presence of renal cysts, these were manually excluded from the segmentation.

RESULTS: Forty-eight patients were included (24 EVAR vs. 24 FEVAR), 96 kidneys were segmented. There was no difference between groups in age (78.9±6.7 years vs. 69.4±6.8, p=0.89), eGFR 85.8 ± 12.4 [62-107] ml/min/1.73 m-2 vs. 81 ± 16.2 [42-107] (p=0.36), and renal volume 170.9 ± 29.7 [123-276] mL vs. 165.3 ± 37.4 [115-298] (p=0.12). At 24 months in the EVAR group, there was a non-significant reduction in eGFR 84.1 ± 17.2 [61-128] ml/min/1.73 m-2 vs. 81 ± 16.2 [42-107] (p=0.36) or renal volume 170.9 ± 29.7 [123-276] mL vs. 165.3 ± 37.4 [115-298] (p=0.12). In the FEVAR group, at 24 months there was a non-significant fall in eGFR 84.1 ± 17.2 [61-128] ml/min/1.73 m-2 vs. 73.8 ± 21.4 [40-110] (p=0.09), while renal volume decreased significantly 182 ± 37.8 [123-293] mL vs. 158.9 ± 40.2 [45-258] (p=0.007).

CONCLUSION: In this study, there appears to be a significant decrease in renal volume without a drop in eGFR 24 months after fenestrated stenting. This decrease may reflect changes in renal perfusion and could potentially be predictive of long-term renal impairment, although this cannot be confirmed within the limits of this small sample. Further studies with long-term follow-up are needed.

PMID:40789507 | DOI:10.1016/j.avsg.2025.08.001

Biol Psychiatry Cogn Neurosci Neuroimaging. 2025 Aug 9:S2451-9022(25)00247-2. doi: 10.1016/j.bpsc.2025.07.012. Online ahead of print.

ABSTRACT

BACKGROUND: Research has demonstrated a broad network of dysfunction across the brain in Attention Deficit/Hyperactivity Disorder (ADHD), suggesting the potential role of white matter (WM) organization. This study sought to estimate the developmental trajectories of brain WM myelination in children with ADHD.

METHODS: Neuroimaging and clinical data were collected as part of a longitudinal community-based pediatric cohort (Nscans=400; 195 with ADHD; age range, 9-14 years). Brain WM myelin was examined for 71 WM tracts across 3 time points using the T1w/T2w-ratio. Tracts were defined via a deep-learning based automated tractography method, performed on participant diffusion-weighted imaging. Linear and non-linear regression was conducted to examine group differences in T1w/T2w-ratio values. In addition to this, voxel-wise analysis was undertaken at each time point.

RESULTS: Brain-wide, children with ADHD were found to exhibit the same developmental profile as those without ADHD for WM myelin. No group effects were seen at a cross-sectional or longitudinal level. In agreement with previous work, modelling suggests non-linear developmental increases with age across most tract. This non-linear relationship was characterized by a positive parabolic, or U-shaped developmental trajectory.

CONCLUSIONS: These findings indicate that there may not be distinct difference in the development of brain white matter myelination between children with and without ADHD. However, this suggests that previously reported differences in ADHD brain WM development may be attributable to properties other than myelin, such as fiber architecture and axon diameter. This further informs the understanding of brain development and highlights the need for further multi-modal longitudinal work.

PMID:40789484 | DOI:10.1016/j.bpsc.2025.07.012

AAPS J. 2025 Aug 11;27(5):129. doi: 10.1208/s12248-025-01119-5.

ABSTRACT

Nintedanib (Nint) is a potent tyrosine kinase inhibitor recently approved by the US FDA to treat idiopathic pulmonary fibrosis (IPF). Delivery of Nint through available approaches is highly challenging because of its poor solubility and rapid metabolic degradation via hydrolytic ester cleavage, thereby reflecting poor oral bioavailability (< 5%). Hence, the current study was focused on formulating a Nint-loaded microemulsion (Nint-ME) and investigating its therapeutic potential in experimental animals to overcome the constraints of available therapies. Nint-ME was prepared via low-energy O/W emulsification aqueous titration techniques and optimized using QbD approach. Optimized ME subjected to screen for globule size, polydispersity index, encapsulation efficiency, transmittance, surface charge, and viscosity and were found to be 23.8 ± 1.4 nm, 0.18 ± 0.03, 99.8 ± 2.4%, 99.4 ± 0.1%, -0.7 ± 0.01 mV, and 1.5 ± 0.3 cP, respectively. Additionally, 94.5 ± 3.1% Nint was released from Nint-ME through the dialysis cassette within 72 h, demonstrating first-order kinetics with R2 of 0.966. First-order and Higuchi release kinetic patterns support concentration-dependent release and Fickian diffusion from the matrix of Nint-ME. In-vitro permeation study of Nint across Caco2 colon epithelial cell monolayer depicted 48.1 ± 1.5 µg of cellular permeation out of 50 µg, ensuring the permeation potential of Nint-ME. Concurrently, an in-vivo pharmacokinetic study for optimized Nint-ME against Nint suspension reflected 41.0 ± 12.5% oral bioavailability, a 2-fold enhancement compared to plain Nint suspension. Existing work demonstrated the successful development of oral Nint-ME as a novel formulation for safe and effective delivery of Nint in IPF.

PMID:40789799 | DOI:10.1208/s12248-025-01119-5

NPJ Vaccines. 2025 Aug 11;10(1):188. doi: 10.1038/s41541-025-01121-0.

ABSTRACT

Given the local and systemic adverse reactions associated with whole-cell pertussis vaccines combined with diphtheria and tetanus toxoids (DTP), acellular pertussis vaccines combined with the same toxoids (DTaP) were developed in the 1990s. In comparison to DTP, DTaP vaccines demonstrated reduced reactogenicity and equivalent or improved immunogenicity and efficacy. However, there has been a resurgence of pertussis disease, particularly in DTaP-vaccinated children, suggesting that immunity wanes more quickly with DTaP vaccination. To elucidate the differences in immune responses to DTP and DTaP vaccines, we employed a systems biology-based strategy to compare global changes in gene expression following primary vaccination with either DTP or DTaP. We used RNA-Seq and ribosome profiling (RP) to identify transcriptional and translational signatures, respectively, in peripheral blood mononuclear cells (PBMCs) collected from 50 infant recipients of DTP or DTaP at two time-points (baseline (pre-vaccination at Day 1) and either Day 2 or 8 post-vaccination). We also used standard serologic methods to assess immunogenicity, and correlated these results with transcriptional and translational signatures. Here, we provide a detailed description of the rationale, experimental design, methodology, and enrollment procedures used. Given the technical complexity of our approach, our objective is to fill knowledge gaps, describe key quality metrics, and support future publications. In brief, we recovered 4-12 million PBMCs (average 8.9 million) with 99% viability per 2.5 mL blood sample, enabling excellent nucleic acid recovery yields for the preparation of high-quality sequencing libraries. In turn, these generated RNA-Seq and RP datasets with sufficient genome coverage breadth and depth to enable differential gene expression analyses, demonstrating the validity of this approach to study pertussis vaccine immunology specifically, and its utility to characterize mechanisms of the human immune response to vaccination generally.

PMID:40789865 | DOI:10.1038/s41541-025-01121-0

Annu Rev Biomed Data Sci. 2025 Aug;8(1):605-632. doi: 10.1146/annurev-biodatasci-120924-091033.

ABSTRACT

DNA methylation, a covalent modification, fundamentally shapes mammalian gene regulation and cellular identity. This review examines methylation's biochemical underpinnings, genomic distribution patterns, and analytical approaches. We highlight three distinctive aspects that separate methylation from other epigenetic marks: its remarkable stability as a silencing mechanism, its capacity to maintain distinct states independently of DNA sequence, and its effectiveness as a quantitative trait linking genotype to disease risk. We also explore the phenomenon of methylation clocks and their biological significance. The review addresses technical considerations across major assay types-both array-based technologies and sequencing approaches-with emphasis on data normalization, quality control, cell proportion inference, and the specialized statistical models required for next-generation sequencing analysis.

PMID:40789737 | DOI:10.1146/annurev-biodatasci-120924-091033

Annu Rev Biomed Data Sci. 2025 Aug;8(1):227-250. doi: 10.1146/annurev-biodatasci-111224-124530.

ABSTRACT

Adverse drug events (ADEs) in pediatric populations pose significant public health challenges, yet research on their detection and monitoring remains limited. This scoping review evaluates the use of unstructured data from electronic health records (EHRs) to identify ADEs in children. We searched six databases, including MEDLINE, Embase, and IEEE Xplore, in September 2024. From 984 records, only nine studies met our inclusion criteria, indicating a significant gap in research toward identifying ADEs in children. We found that unstructured data in EHRs can indeed be of value and enhance pediatric pharmacovigilance, although their use has been so far very limited. Traditional natural language processing methods have been employed to extract ADEs, but the approaches utilized face challenges in generalizability and context interpretation. These challenges could be addressed with recent advances in transformer-based models and large language models, unlocking the use of EHR data at scale for pediatric pharmacovigilance.

PMID:40789734 | DOI:10.1146/annurev-biodatasci-111224-124530

Clin Transl Sci. 2025 Aug;18(8):e70300. doi: 10.1111/cts.70300.

ABSTRACT

Pharmacogenomics (PGx) can potentially tailor medication prescriptions to the genetic profiles of individuals, enhancing treatment outcomes and minimizing adverse drug reactions. This study assessed cardiovascular disease (CVD) patients' knowledge and views toward PGx testing in the United Arab Emirates (UAE). A cross-sectional study was conducted among CVD patients attending multiple clinics using a validated, culturally adapted, and piloted bilingual questionnaire. Participants were invited via phone calls or in-person contact at clinics. Data analysis was conducted using SPSS V.29, incorporating descriptive statistics and multivariable logistic regression. A total of 425 responses were analyzed; 67.5% were over 50 years old, and 67.5% held a bachelor's degree. Chronic diseases, excluding CVD, affected 65.2%, with 58.1% reporting medication side effects and 36.5% was hospitalized due to these effects. Knowledge varied, with 55.3% demonstrating good knowledge; 75.3% recognized DNA as gene-based, while 47.5% understood PGx for predicting medication responses. Participants were grouped into three PGx perception clusters: Cluster 1 (33.17%) concerned about risks but valued PGx, Cluster 2 (40.23%) worried about privacy/costs, and Cluster 3 (26.58%) confident in PGx benefits. Safety was the top priority for 60.2% of respondents, 34.8% would not pay for PGx tets, and 35.3% preferred preemptive testing. Regression linked higher PGx knowledge to females, non-healthcare workers, those with genetic diseases, and those hospitalized for side effects (p < 0.05). The study highlights a need for educational initiatives in the UAE to improve PGx literacy among CVD patients. The findings suggest that targeted awareness campaigns, policy interventions addressing privacy, and financial support could promote PGx wider adoption.

PMID:40788819 | DOI:10.1111/cts.70300

Am J Respir Crit Care Med. 2025 Aug 11. doi: 10.1164/rccm.202411-2312OC. Online ahead of print.

ABSTRACT

RATIONALE: Abnormal glucose metabolism is a common complication in people with cystic fibrosis (CF), and those with impaired glucose tolerance (IGT) or CF-related diabetes (CFRD) have increased disease burden. Elexacaftor/tezacaftor/ivacaftor (ELX/TEZ/IVA) is safe and effective for people with CF aged ≥2 years with ELX/TEZ/IVA-responsive CFTR mutations; however, efficacy on glycemic control has not been studied.

OBJECTIVES: To evaluate the impact of ELX/TEZ/IVA on glucose tolerance in people with CF who have IGT or CFRD.

METHODS: This phase 3b, open-label study, ELX/TEZ/IVA was administered for 48weeks to participants aged ≥12 years, heterozygous for F508del and a minimal function CFTR mutation, and with either IGT or CFRD.

MEASUREMENTS AND MAIN RESULTS: Sixty-nine participants received ELX/TEZ/IVA. The primary endpoint was change in blood glucose levels following 2-hour oral glucose tolerance test (OGTT) from baseline to the average of Week36 and Week48; participants had a mean change of -35.0 mg/dL (95%CI:-49.2,-20.7;P<0.0001) (-1.9 mmol/L [95%CI: 2.7, 1.2]). Secondary endpoints were the proportion of participants with improvement in dysglycemia categorization (CFRD, IGT, normal glucose tolerance) at Week48 and safety. Among participants with abnormal glucose tolerance at baseline, 37.7% (95%CI:24.8,52.1) had improvements in dysglycemia categorization at Week48. Overall, 35.5% of participants had normal glucose tolerance at Week48 compared to 13.0% at baseline. Safety was consistent with the established safety profile of ELX/TEZ/IVA.

CONCLUSIONS: ELX/TEZ/IVA treatment led to clinically meaningful improvements in blood glucose regulation with significant within-group decreases in blood glucose levels following OGTT and improved dysglycemia categorization in people with CF with early IGT or CFRD. Clinical trial registration available at www.

CLINICALTRIALS: gov, ID: NCT04599465.

PMID:40788823 | DOI:10.1164/rccm.202411-2312OC

Comput Biol Med. 2025 Aug 10;196(Pt C):110774. doi: 10.1016/j.compbiomed.2025.110774. Online ahead of print.

ABSTRACT

Cardiac anomalies are severe and life-threatening, making early detection essential to reducing health risks and mortality. According to the European Society of Cardiology, over 13 million people suffer from heart valve diseases annually, often identified by heartbeat anomalies. Traditional diagnostic methods depend on specialized expertise and advanced equipment. This paper proposes SpectroNet-LSTM, an automated framework for detecting cardiac anomalies from a comprehensive dataset of heartbeat sound recordings. Mel-frequency cepstral coefficients (MFCCs) and spectrogram analysis are used to capture critical acoustic features. These features are then extracted and employed to train state-of-the-art deep learning models, including ResNet101, VGG16, and Inception V3. The core architecture is trained on the extracted features and optimized for improved performance. The model outperforms benchmarks on various evaluation metrics for detecting heart anomalies. To ensure system interpretability, the study integrates two Explainable AI (XAI) techniques, namely SHAP and LIME. These techniques enable clinicians and patients to visualize and understand the model's decision-making process. The novelty of SpectroNet-LSTM lies in its integrated use of advanced feature extraction, deep learning fusion and explainable AI to create a fully automated and interpretable cardiac anomaly detection system. This research underscores the potential of automation in transforming cardiovascular diagnostics, paving the way for accessible healthcare solutions and efficient patient outcomes worldwide.

PMID:40789236 | DOI:10.1016/j.compbiomed.2025.110774

Comput Biol Med. 2025 Aug 10;196(Pt C):110902. doi: 10.1016/j.compbiomed.2025.110902. Online ahead of print.

ABSTRACT

Cervical cancer is a significant global health issue, and high-dose-rate brachytherapy (HDR-BT) is crucial for its treatment. However, manually creating HDR-BT plans is time-consuming and heavily relies on the planner's expertise, making standardization difficult. This study introduces two advanced deep learning models to address this need: Bi-branch Cross-Attention UNet (BiCA-UNet) and Dendrite Cross-Attention UNet (DCA-UNet). BiCA-UNet enhances the correlation between the CT scan and segmentation maps of the clinical target volume (CTV), applicator, bladder, and rectum. It uses two branches: one processes the stacked input of CT scans and segmentations, and the other focuses on the CTV segmentation. A cross-attention mechanism integrates these branches, improving the model's understanding of the CTV region for accurate dose predictions. Building on BiCA-UNet, DCA-UNet further introduces a primary branch of stacked inputs and three secondary branches for CTV, bladder, and rectum segmentations forming a dendritic structure. Cross attention with bladder and rectum segmentation helps the model understand the regions of organs at risk (OAR), refining dose prediction. Evaluation of these models using multiple metrics indicates that both BiCA-UNet and DCA-UNet significantly improve HDR-BT dose prediction accuracy for various applicator types. The cross-attention mechanisms enhance the feature representation of critical anatomical regions, leading to precise and reliable treatment plans. This research highlights the potential of BiCA-UNet and DCA-UNet in advancing HDR-BT planning, contributing to the standardization of treatment plans, and offering promising directions for future research to improve patient outcomes in the source data.

PMID:40789235 | DOI:10.1016/j.compbiomed.2025.110902

J Med Internet Res. 2025 Aug 11;27:e71786. doi: 10.2196/71786.

ABSTRACT

BACKGROUND: The seasonal influenza epidemic poses a persistent and severe threat to global public health. Web-based search data are recognized as a valuable source for forecasting influenza or other respiratory tract infection epidemics. Current influenza prediction studies typically focus on seasonal trends in traditional monitoring data, neglecting the sensitivity of different web-based search terms to seasonal changes, thereby increasing prediction challenges.

OBJECTIVE: The aim of this study was to propose a deep learning framework for different influenza epidemic states based on Baidu index and percentage of influenza-like illness (ILI%).

METHODS: Official weekly ILI% data from 2013 to 2024 were extracted from the Chinese National Notifiable Infectious Disease Reporting System (NIDRIS). Based on the Baidu index, influenza-related search indexes were acquired for the corresponding time periods. To explore the association between influenza-related search queries and ILI%, the study conducted a cross-correlation analysis. The study period was divided into influenza epidemic and nonepidemic period. The study finally used the convolutional long short-term memory (CLSTM) network framework to predict influenza epidemics with 1-3 weeks ahead for the all-time period and epidemic + nonepidemic period. The evaluation metrics included model stability metric, accuracy metrics, and explanatory power metric.

RESULTS: The ILI% presented a regular seasonal high incidence in China. Meanwhile, the prediction of ILI% after dividing the epidemic and nonepidemic seasons (mean absolute percentage error [MAPE]=10.730%, mean square error [MSE]=0.884, mean absolute error [MAE]=0.649, root-mean-square error [RMSE]=0.940, and R2=0.877) was better than that of the all-time period (MAPE=12.784%, MSE=1.513, MAE=0.744, RMSE=1.230, and R2=0.786). In addition, we found that the ILI% + Baidu search index predicts better than only the ILI% regardless of the time period and lag time of the study. Comparative analysis with long short-term memory (LSTM) and transformer models demonstrated that CLSTM achieved superior performance in 1 week-ahead ILI% predictions using ILI% + Baidu index data in epidemic + nonepidemic period (MAPE=11.824%, MSE=1.243, MAE=0.723, RMSE=1.115, and R2=0.827). Furthermore, CLSTM comprehensively surpasses LSTM in computational efficiency, complexity, extrapolation capability, and stability while partially outperforming transformer models.

CONCLUSIONS: This study shows strong potential for influenza prediction by combining Baidu index data with traditional surveillance and specific keywords for epidemic and nonepidemic seasons. It provides a new perspective for public health preparedness. This research is expected to support early warning systems for influenza and other diseases. Future work will further optimize these models for more timely and accurate predictions, enhancing public health responses.

PMID:40789146 | DOI:10.2196/71786

J Chem Inf Model. 2025 Aug 11. doi: 10.1021/acs.jcim.5c01030. Online ahead of print.

ABSTRACT

Noncanonical amino acids (NCAAs) occupy an important place, both in natural biology and in synthetic applications. However, modeling these amino acids still lies outside the capabilities of most deep learning methods due to sparse training data sets for this task. Instead, biophysical methods such as Rosetta can excel in modeling NCAAs. We discuss the various aspects of parametrizing an NCAA for use in Rosetta, identifying rotamer distribution modeling as one of the most impactful factors of NCAA parametrization on Rosetta performance. To this end, we also present FakeRotLib, a method that uses statistical fitting of small-molecule conformers to create rotamer distributions. We find that FakeRotLib outperforms existing methods in a fraction of the time and is able to parametrize NCAA types previously unmodeled by Rosetta.

PMID:40789114 | DOI:10.1021/acs.jcim.5c01030

Med Ultrason. 2025 Jul 29. doi: 10.11152/mu-4535. Online ahead of print.

ABSTRACT

AIMS: A computer-aided diagnosis (CAD) system for automated evaluation of developmental dysplasia of the hip (DDH) via ultrasound, integrating Deep Learning (DL) for anatomical segmentation and performing α&β angle calculations utilizing the Graf Method is presented. A custom image processing method excludes the inferior ilium's curvature during the baseline definition, enhancing accuracy and replicating radiologists' real-world workflow.

MATERIALS AND METHODS: Our dataset comprised 452 raw images from 370 newborns. For {'validation'+"test"}, {'nv=91'+"nte=45"}≡136 images were reserved (never augmented). Remaining 316 images were augmented to ntr=632 with (0%↔25%) random brightness manipulation for training. Totally (632+136)=768 images were annotated and split with the following true numbers and percentage: {'train',"validation",test}≡{'632',"91",45}≡{'82%',"12%",6%}. U-Net, MaskR-CNN, YOLOv8 and YOLOv11 were used for segmentation. α&β were measured using Method-I (centroid/orientation) and Method-II (Hough transform). An extended set of performance metrics-Precision, Recall, IoU, Dice, mAP-was calculated. Bland-Altman and Intraclass Correlation Coefficient (ICC) analyses compared CAD outputs with expert measurements.

RESULTS: YOLOv11 showed the best segmentation performance (Precision:0.990, Recall:0.993, IoU:0.983, Dice:0.990, mAP:0.991). {ICCα, ICCβ} calculated using Method-I and Method-II were {0.895, 0.907} and {0.929, 0.952}, respectively, with Method-II outperforming Method-I.

CONCLUSION: A clinically-aligned-CAD-system that integrates anatomical segmentation and α&β measurement-a combination rarely addressed in literature is introduced. By providing a comprehensive and standardized set of metrics, this work overcomes a common bottleneck in DL studies, namely heterogeneity in metric reporting, enabling better cross-study comparisons. Following curvature exclusion, obtained ICCs outperformed previous studies, demonstrating improved inter-rater reliability and strong agreement with expert radiologists, offering both technical robustness and clinical applicability in DDH assessment.

PMID:40789016 | DOI:10.11152/mu-4535

PLoS One. 2025 Aug 11;20(8):e0327782. doi: 10.1371/journal.pone.0327782. eCollection 2025.

ABSTRACT

BACKGROUND: Accurate classification of meningioma brain tumors is crucial for determining the appropriate treatment plan and improving patient outcomes. However, this task is challenging due to the slow-growing nature of these tumors and the potential for misdiagnosis. Additionally, deep learning models for tumor classification often require large amounts of labeled data, which can be costly and time-consuming to obtain, especially in the medical domain.

OBJECTIVE: Our main aim is to enhance Meningioma Tumor Classification Accuracy.

METHOD: This study proposes a multi-task learning (MTL) approach to enhance the accuracy of meningioma tumor classification while mitigating the need for excessive labeled data. The primary task involves classifying meningioma tumors based on MRI imaging data, while auxiliary tasks leverage patient demographic information, such as age and gender. By incorporating these additional data sources into the learning process, the proposed MTL framework leverages the interdependencies among multiple tasks to improve overall prediction accuracy. The study evaluates the performance of the MTL approach using a dataset of 2218 brain MRI images from 34 patients diagnosed with meningioma, obtained from the Mahdia Imaging Center in Hamadan, Iran.

RESULTS: Results demonstrate that the MTL model significantly outperforms single-task learning baselines, achieving 99.6% ± 0.2 accuracy on the test data in 95% confidence interval.

DISCUSSION: This highlights the efficacy of the proposed approach in enhancing meningioma tumor classification and its potential for aiding clinical decision-making and personalized treatment planning.

CONCLUSION: Our proposed method can be used in computer-aided diagnosis systems.

PMID:40788922 | DOI:10.1371/journal.pone.0327782

PLoS One. 2025 Aug 11;20(8):e0327577. doi: 10.1371/journal.pone.0327577. eCollection 2025.

ABSTRACT

Assessing the escalating biodiversity crisis, driven by climate change, habitat destruction, and exploitation, necessitates efficient monitoring strategies to assess species presence and abundance across diverse habitats. Video-based surveys using remote cameras are a promising, non-invasive way to collect valuable data in various environments. Yet, the analysis of recorded videos remains challenging due to time and expertise constraints. Recent advances in deep learning models have enhanced image processing capabilities in both object detection and classification. However, the impacts on models' performances and usage on assessment of biodiversity metrics on videos is yet to be assessed. This study evaluates the impacts of video processing rates, detection and identification model performance, and post-processing algorithms on the accuracy of biodiversity metrics, using simulated remote videos of fish communities and 14,406 simulated automated processing pipelines. We found that a processing rate of one image per second minimizes errors while ensuring detection of all species. However, even near-perfect detection (both recall and precision of 0.99) and identification (accuracy of 0.99) models resulted in overestimation of total abundance, species richness and species diversity due to false positives. We reveal that post-processing model outputs using a confidence threshold approach (i.e., to discard most erroneous predictions while also discarding a smaller proportion of correct predictions) is the most efficient method to accurately estimate biodiversity from videos.

PMID:40788894 | DOI:10.1371/journal.pone.0327577

IEEE J Biomed Health Inform. 2025 Aug 11;PP. doi: 10.1109/JBHI.2025.3597589. Online ahead of print.

ABSTRACT

Long non-coding RNAs (lncRNAs) play essential roles in various biological processes, including chromatin modification, cell cycle regulation, transcription, and translation. Recent studies have revealed that the biological functions of lncRNAs are closely associated with their subcellular localizations, making accurate localization prediction critical for understanding their biological roles in cellular regulation and disease mechanisms. However, most existing methods mainly rely on sequence features while neglecting structural information, and they are often limited to single-label predictions covering only a small number of subcellular compartments. In this study, we proposed an efficient deep learning framework, LncTracker, for multi-label prediction of lncRNA subcellular localizations across seven distinct compartments. LncTracker adopts a multi-channel architecture that integrates diverse input features into model training, including both primary sequence and secondary structure information. Secondary structures are converted into attributed graphs to capture spatial relationships among nucleotides, including adjacency and base-pairing connections. These structural features are then combined with sequence-based features to predict subcellular localization probabilities. Such a design enables LncTracker to learn joint representations of sequences and structures, thereby enhancing predictive performance and robustness. Benchmarking experiments demonstrated the superiority of LncTracker over state-of-the-art approaches, particularly in handling imbalanced localization scenarios. Furthermore, we leveraged LncTracker to identify sequence motifs critical for each subcellular localization and analysed key sub-structures contributing to predictions. The codes are provided on GitHub https://github.com/ABILiLab/LncTracker. To enhance the usability of LncTracker, we developed a web server that is publicly accessible at http://lnctracker.biotools.bio.

PMID:40788811 | DOI:10.1109/JBHI.2025.3597589