Circ Heart Fail. 2025 May 14:e012667. doi: 10.1161/CIRCHEARTFAILURE.124.012667. Online ahead of print.

ABSTRACT

BACKGROUND: Hypertrophic cardiomyopathy (HCM) is associated with significant morbidity and mortality, including sudden cardiac death in the young. Its prevalence is estimated to be 1 in 500, although many people are undiagnosed. The ability to screen electrocardiograms for its presence could improve detection and enable earlier diagnosis. This study evaluated the accuracy of an artificial intelligence device (viz HCM) in detecting HCM based on a 12-lead electrocardiogram.

METHODS: The device was previously trained using deep learning and provides a binary outcome (HCM suspected or not suspected). This study included 293 HCM-positive and 2912 HCM-negative cases, which were selected from 3 hospitals based on chart review incorporating billing diagnostic codes, cardiac imaging, and electrocardiogram features. The device produced an output for 291 (99.3%) HCM-positive and 2905 (99.8%) HCM-negative cases.

RESULTS: The device identified HCM with sensitivity of 68.4% (95% CI, 62.8-73.5%), specificity of 99.1% (95% CI, 98.7-99.4%), and area under the curve of 0.975 (95% CI, 0.965-0.982). With assumed population prevalence of 0.002 (1 in 500), the positive predictive value was 13.7% (95% CI, 10.1-19.9%) and the negative predictive value was 99.9% (95% CI, 99.9-99.9%). The device demonstrated broadly consistent performance across demographic and technical subgroups.

CONCLUSIONS: The device identified HCM based on a 12-lead electrocardiogram with good performance. Coupled with clinical expertise, it has the potential to augment HCM detection and diagnosis.

PMID:40365710 | DOI:10.1161/CIRCHEARTFAILURE.124.012667

Front Artif Intell. 2025 Apr 29;8:1545946. doi: 10.3389/frai.2025.1545946. eCollection 2025.

ABSTRACT

This research proposes a novel approach to cryptographic key generation using biometric data from face and finger vein modalities enhanced by deep learning techniques. Using pretrained models FaceNet and VGG19 for feature extraction and employing a Siamese Neural Network (SNN), the study demonstrates the integration of multimodal biometrics with fuzzy extractors to create secure and reproducible cryptographic keys. Feature fusion techniques, combined with preprocessing and thresholding, ensure robust feature extraction and conversion to binary formats for key generation. The model demonstrates impressive accuracy with a vector converter, achieving a sigma similarity of 93% and a sigma difference of 64.0%. Evaluation metrics, including False Acceptance Rate (FAR) and False Rejection Rate (FRR), indicate significant improvements, achieving FRR < 3.4% and FAR < 1%, outperforming previous works. Additionally, the adoption of Goppa code-based cryptographic systems ensures post-quantum security. This study not only enhances biometric cryptography's accuracy and resilience but also paves the way for future exploration of quantum-resistant and scalable systems.

PMID:40365578 | PMC:PMC12069345 | DOI:10.3389/frai.2025.1545946

Front Artif Intell. 2025 Apr 29;8:1563016. doi: 10.3389/frai.2025.1563016. eCollection 2025.

ABSTRACT

Alzheimer's disease (AD) is a progressive, incurable neurological disorder that leads to a gradual decline in cognitive abilities. Early detection is vital for alleviating symptoms and improving patient quality of life. With a shortage of medical experts, automated diagnostic systems are increasingly crucial in healthcare, reducing the burden on providers and enhancing diagnostic accuracy. AD remains a global health challenge, requiring effective early detection strategies to prevent its progression and facilitate timely intervention. In this study, a deep convolutional neural network (CNN) architecture is proposed for AD classification. The model, consisting of 6,026,324 parameters, uses three distinct convolutional branches with varying lengths and kernel sizes to improve feature extraction. The OASIS dataset used includes 80,000 MRI images sourced from Kaggle, categorized into four classes: non-demented (67,200 images), very mild demented (13,700 images), mild demented (5,200 images), and moderate demented (488 images). To address the dataset imbalance, a data augmentation technique was applied. The proposed model achieved a remarkable 99.68% accuracy in distinguishing between the four stages of Alzheimer's: Non-Dementia, Very Mild Dementia, Mild Dementia, and Moderate Dementia. This high accuracy highlights the model's potential for real-time analysis and early diagnosis of AD, offering a promising tool for healthcare professionals.

PMID:40365577 | PMC:PMC12069281 | DOI:10.3389/frai.2025.1563016

Front Med (Lausanne). 2025 Apr 29;12:1526144. doi: 10.3389/fmed.2025.1526144. eCollection 2025.

ABSTRACT

OBJECTIVE: This research aimed to create a dataset of Iliac CTV scans for automated May-Thurner syndrome (MTS) detection using deep learning and radiomics. In addition, it sought to establish an automated segmentation model for Iliac Vein CTV scans and construct a radiomic signature for MTS diagnosis.

METHODS: We collected a dataset of 490 cases meeting specific inclusion and exclusion criteria, anonymized to comply with HIPAA regulations. Iliac Vein CTV scans were prepared with contrast agent administration, followed by image acquisition and evaluation. A deep learning-based segmentation model, UPerNet, was employed using 10-fold cross-validation. Radiomic features were extracted from the scans and used to construct a diagnostic radiomic signature. Statistical analysis, including Dice values and ROC analysis, was conducted to evaluate segmentation and diagnostic performance.

RESULTS: The dataset consisted of 201 positive cases of MTS and 289 negative cases. The UPerNet segmentation model exhibited remarkable accuracy in identifying MTS regions. A Dice coefficient of 0.925 (95% confidence interval: 0.875-0.961) was observed, indicating the precision and reliability of our segmentation model. Radiomic analysis produced a diagnostic radiomic signature with significant clinical potential. ROC analysis demonstrated promising results, underscoring the efficacy of the developed model in distinguishing MTS cases. The radiomic signature demonstrated strong diagnostic capabilities for MTS. Within the training dataset, it attained a notable area under the curve (AUC) of 0.891, with a 95% confidence interval ranging from 0.825 to 0.956, showcasing its effectiveness. This diagnostic capability extended to the validation dataset, where the AUC remained strong at 0.892 (95% confidence interval: 0.793-0.991). These results highlight the accuracy of our segmentation model and the diagnostic value of our radiomic signature in identifying MTS cases.

CONCLUSION: This study presents a comprehensive approach to automate MTS detection from Iliac CTV scans, combining deep learning and radiomics. The results suggest the potential clinical utility of the developed model in diagnosing MTS, offering a non-invasive and efficient alternative to traditional methods.

PMID:40365495 | PMC:PMC12069258 | DOI:10.3389/fmed.2025.1526144

Alzheimers Dement (Amst). 2025 May 12;17(2):e70118. doi: 10.1002/dad2.70118. eCollection 2025 Apr-Jun.

ABSTRACT

INTRODUCTION: The locus coeruleus (LC) is linked to the development and pathophysiology of neurodegenerative diseases such as Alzheimer's disease (AD). Magnetic resonance imaging-based LC features have shown potential to assess LC integrity in vivo.

METHODS: We present a deep learning-based LC segmentation and feature extraction method called Ensemble-based Locus Coeruleus Segmentation Network (ELSI-Net) and apply it to healthy aging and AD dementia datasets. Agreement to expert raters and previously published LC atlases were assessed. We aimed to reproduce previously reported differences in LC integrity in aging and AD dementia and correlate extracted features to cerebrospinal fluid (CSF) biomarkers of AD pathology.

RESULTS: ELSI-Net demonstrated high agreement to expert raters and published atlases. Previously reported group differences in LC integrity were detected and correlations to CSF biomarkers were found.

DISCUSSION: Although we found excellent performance, further evaluations on more diverse datasets from clinical cohorts are required for a conclusive assessment of ELSI-Net's general applicability.

HIGHLIGHTS: We provide a thorough evaluation of a fully automatic locus coeruleus (LC) segmentation method termed Ensemble-based Locus Coeruleus Segmentation Network (ELSI-Net) in aging and Alzheimer's disease (AD) dementia.ELSI-Net outperforms previous work and shows high agreement with manual ratings and previously published LC atlases.ELSI-Net replicates previously shown LC group differences in aging and AD.ELSI-Net's LC mask volume correlates with cerebrospinal fluid biomarkers of AD pathology.

PMID:40365469 | PMC:PMC12069022 | DOI:10.1002/dad2.70118

Front Endocrinol (Lausanne). 2025 Apr 29;16:1504412. doi: 10.3389/fendo.2025.1504412. eCollection 2025.

ABSTRACT

OBJECTIVES: The ultrasound characteristics of benign and malignant thyroid nodules were compared to develop a deep learning model, aiming to establish a nomogram model based on deep learning ultrasound image analysis to improve the predictive performance of thyroid nodules.

MATERIALS AND METHODS: This retrospective study analyzed the clinical and ultrasound characteristics of 2247 thyroid nodules from March 2016 to October 2023. Among them, 1573 nodules were used for training and testing the deep learning models, and 674 nodules were used for validation, and the deep learning predicted values were obtained. These 674 nodules were randomly divided into a training set and a validation set in a 7:3 ratio to construct a nomogram model.

RESULTS: The accuracy of the deep learning model in 674 thyroid nodules was 0.886, with a precision of 0.900, a recall rate of 0.889, and an F1-score of 0.895. The binary logistic analysis of the training set revealed that age, echogenic foci, and deep learning predicted values were statistically significant (P<0.05). These three indicators were used to construct the nomogram model, showing higher accuracy compared to the China thyroid imaging reports and data systems (C-TIRADS) classification and deep learning models. Moreover, the nomogram model exhibited high calibration and clinical benefits.

CONCLUSION: Age, deep learning predicted values, and echogenic foci can be used as independent predictive factors to distinguish between benign and malignant thyroid nodules. The nomogram integrates deep learning and patient clinical ultrasound characteristics, yielding higher accuracy than the application of C-TIRADS or deep learning models alone.

PMID:40365227 | PMC:PMC12069047 | DOI:10.3389/fendo.2025.1504412

RSC Adv. 2025 May 13;15(20):15781-15795. doi: 10.1039/d5ra01578b. eCollection 2025 May 12.

ABSTRACT

Low-cycle fatigue (LCF) data involve complex temporal interactions in a strain cycle series, which hinders accurate fatigue life prediction. Current studies lack reliable methods for fatigue life prediction using only initial-cycle data while simultaneously capturing both temporal dependencies and localized features. This study introduces a novel deep-learning-based prediction model designed for LCF data. The proposed approach combines long short-term memory (LSTM) and convolutional neural network (CNN) architectures with an attention mechanism to effectively capture the temporal and localized characteristics of stress-strain data from acquisition through a series of cycle strain-controlled tests. Among the models tested, the LSTM-contextual attention model demonstrated superior performance (R 2 = 0.99), outperforming the baseline LSTM and CNN models with higher R 2 values and improved statistical metrics. The analysis of attention weights further revealed the model's ability to focus on critical timesteps associated with fatigue damage, highlighting its effectiveness in learning key features from LCF data. This study underscores the potential of deep-learning-based methods for accurate fatigue life prediction in LCF applications. This study provides a foundation for future research to extend these approaches to diverse materials with varying fatigue conditions and advanced models capable of incorporating non-linear fatigue mechanisms.

PMID:40365199 | PMC:PMC12070260 | DOI:10.1039/d5ra01578b

Predict Intell Med. 2025;15155:185-195. doi: 10.1007/978-3-031-74561-4_16. Epub 2024 Oct 18.

ABSTRACT

Interpretability is essential to MRI brain studies relying on deep learning models for neuroscientific discovery. One way to facilitate the interpretability of a deep learning model is to ensure the samples are arranged in the model's latent space with respect to clinically meaningful variables. To achieve this in the context of cross-sectional brain MRI studies, we regularize the latent space of a multi-label classifier via pairwise disentanglement, so that the difference between the representation of two brain MRIs along the disentangled direction in the latent space is similar to the difference in their neuropsychological test scores. We apply our technique to classify brain MRIs of 156 controls, 165 cases diagnosed with mild cognitive impairment (MCI), 166 diagnosed with human immunodeficiency virus (HIV)-associated cognitive disorder (HAND), and 32 individuals diagnosed with HIV without HAND. The latent space is disentangled with respect to the neuropsychological z-score (NPZ), which is negatively correlated with the severity of cognitive impairment (i.e., low scores for those diagnosed with MCI or HAND). Based on cross-validation, the proposed model achieves statistically significantly higher balanced accuracy than the same model without disentanglement. Furthermore, the difference between representations along the disentangled direction significantly correlates with the difference in NPZ. Finally, the brain regions guiding the classification process aligned with the neuroscientific literature.

PMID:40365134 | PMC:PMC12068855 | DOI:10.1007/978-3-031-74561-4_16

Proc SPIE Int Soc Opt Eng. 2025 Feb;13405:134051K. doi: 10.1117/12.3047080. Epub 2025 Apr 8.

ABSTRACT

Emerging deep-learning-based CT denoising techniques have the potential to improve diagnostic image quality in low-dose CT exams. However, aggressive radiation dose reduction and the intrinsic uncertainty in convolutional neural network (CNN) outputs are detrimental to detecting critical lesions (e.g., liver metastases) in CNN-denoised images. To tackle these issues, we characterized CNN output distribution via total uncertainty (i.e., data + model uncertainties) and predictive mean. Local mean-uncertainty-ratio (MUR) was calculated to detect highly unreliable regions in the denoised images. A MUR-driven adaptive local fusion (ALF) process was developed to adaptively merge local predictive means with the original noisy images, thereby improving image robustness. This process was incorporated into a previously validated deep-learning model observer to quantify liver metastasis detectability, using area under localization receiver operating characteristic curve (LAUC) as the figure-of-merit. For proof-of-concept, the proposed method was established and validated for a ResNet-based CT denoising method. A recent patient abdominal CT dataset was used in validation, involving 3 lesion sizes (7, 9, and 11 mm), 3 lesion contrasts (15, 20, and 25 HU), and 3 dose levels (25%, 50%, and 100% dose). Visual inspection and quantitative analyses were conducted. Statistical significance was tested. Total uncertainty at lesions and liver background generally increased as radiation dose decreased. With fixed dose, lesion-wise MUR showed no dependency on lesion size or contrast, but exhibited large variance across lesion locations (MUR range ~0.7 to 19). Compared to original ResNet-based denoising, the MUR-driven ALF consistently improved lesion detectability in challenging conditions such as lower dose, smaller lesion size, or lower contrast (range of absolute gain in LAUC: 0.04 to 0.1; P-value 0.008). The proposed method has the potential to improve reliability of deep-learning CT denoising and enhance lesion detection.

PMID:40365118 | PMC:PMC12070600 | DOI:10.1117/12.3047080

Proc SPIE Int Soc Opt Eng. 2025 Feb;13405:134052J. doi: 10.1117/12.3047909. Epub 2025 Apr 8.

ABSTRACT

Low-dose CT simulation is needed to assess reconstruction/denoising techniques and optimize dose. Projection-domain noise-insertion methods require manufacturers' proprietary tools. Image-domain noise-insertion methods face various challenges that affect generalizability, and few have been systematically validated for 3D noise synthesis. To improve generalizability, we presented a physics-informed model-based generative neural network for simulating scanner- and algorithm-specific low-dose CT exams (PALETTE). PALETTE included a noise-prior-generation process, a Noise2Noisier sub-network, and a noise-texture-synthesis sub-network. Custom regularization terms were developed to enforce 3D noise texture quality. Using PALETTE, one 2D and two 2.5D models (denoted as 2.5D N-N and N-1) were developed to conduct 2D and effective 3D noise modeling, respectively (input/output images: 2D - 1/1, 2.5D N-N - 3/3, 2.5D N-1 - 5/1). These models were trained and tested with an open-access abdominal CT dataset, including 20 testing cases reconstructed with two kernels and various field-of-view. In visual inspection, the 2D and 2.5D N-N models generated realistic local and global noise texture, while 2.5D N-1 showed more perceptual difference using the sharper kernel and coronal reformat. In quantitative evaluation, local noise level was compared using mean-absolute-percent-difference (MAPD), and global spectral similarity was assessed using spectral correlation mapper (SCM) and spectral angle mapper (SAM). The 2D model provided equivalent or relatively better performance than 2.5D models, showing well-matched local noise levels and high spectral similarity compared to the reference (sharper/smoother kernels): MAPD - 2D 1.5%/5.6% (p>0.05), 2.5D N-N 8.5%/7.9% (p<0.05), 2.5D N-1 12.3%/10.9% (p<0.05); mean SCM - 2D 0.97/0.97, 2.5D N-N 0.96/0.97, 2.5D N-1 0.85/0.97; mean SAM - 2D 0.12/0.12, 2.5D N-N 0.14/0.12, 2.5D N-1 0.37/0.12. With tripled model width, the 2.5D N-N outperformed N-1. This indicated 2.5D models need more learning capacity to further enhance 3D noise modeling. Using physics-based prior information, PALETTE can provide high-quality low-dose CT simulation to resemble scanner- and algorithm-specific 3D noise characteristics.

PMID:40365117 | PMC:PMC12070530 | DOI:10.1117/12.3047909

Virology. 2025 Apr 29;609:110559. doi: 10.1016/j.virol.2025.110559. Online ahead of print.

ABSTRACT

Bacteriophages, or bacterial viruses, play diverse ecological roles by shaping bacterial populations and also hold significant biotechnological and medical potential, including the treatment of infections caused by multidrug-resistant bacteria. The discovery of novel bacteriophages using large-scale metagenomic data has been accelerated by the accessibility of deep learning (Artificial Intelligence), the increased computing power of graphical processing units (GPUs), and new bioinformatics tools. This review addresses the recent revolution in bacteriophage research, ranging from the adoption of neural network algorithms applied to metagenomic data to the use of pre-trained language models, such as BERT, which have improved the reconstruction of viral metagenome-assembled genomes (vMAGs). This article also discusses the main aspects of bacteriophage biology using deep learning, highlighting the advances and limitations of this approach. Finally, prospects of deep-learning-based metagenomic algorithms and recommendations for future investigations are described.

PMID:40359589 | DOI:10.1016/j.virol.2025.110559

J Invasive Cardiol. 2025 May 13. doi: 10.25270/jic/25.00089. Online ahead of print.

ABSTRACT

Artificial intelligence (AI) has become pivotal in advancing medical care, particularly in interventional cardiology. Recent AI developments have proven effective in guiding advanced procedures and complex decisions. The authors review the latest AI-based innovations in the diagnosis of chronic total occlusions (CTO) and in determining the probability of success of CTO percutaneous coronary intervention (PCI). Neural networks and deep learning strategies were the most commonly used algorithms, and the models were trained and deployed using a variety of data types, such as clinical parameters and imaging. AI holds great promise in facilitating CTO PCI.

PMID:40359582 | DOI:10.25270/jic/25.00089

J Chem Theory Comput. 2025 May 13. doi: 10.1021/acs.jctc.5c00256. Online ahead of print.

ABSTRACT

Proteins are inherently dynamic molecules, and their conformational transitions among various states are essential for numerous biological processes, which are often modulated by their interactions with surrounding environments. Although molecular dynamics (MD) simulations are widely used to investigate these transitions, all-atom (AA) methods are often limited by short time scales and high computational costs, and coarse-grained (CG) implicit-solvent Go̅-like models are usually incapable of studying the interactions between proteins and their environments. Here, we present an approach called Multiple-basin Go̅-Martini, which combines the recent Go̅-Martini model with an exponential mixing scheme to facilitate the simulation of spontaneous protein conformational transitions in explicit environments. We demonstrate the versatility of our method through five diverse case studies: GlnBP, Arc, Hinge, SemiSWEET, and TRAAK, representing ligand-binding proteins, fold-switching proteins, de novo designed proteins, transporters, and mechanosensitive ion channels, respectively. Multiple-basin Go̅-Martini offers a new computational tool for investigating protein conformational transitions, identifying key intermediate states, and elucidating essential interactions between proteins and their environments, particularly protein-membrane interactions. In addition, this approach can efficiently generate thermodynamically meaningful data sets of protein conformational space, which may enhance deep learning-based models for predicting protein conformation distributions.

PMID:40359486 | DOI:10.1021/acs.jctc.5c00256

PLoS One. 2025 May 13;20(5):e0321530. doi: 10.1371/journal.pone.0321530. eCollection 2025.

ABSTRACT

The Consumer Price Index (CPI) is a key economic indicator used by policymakers worldwide to monitor inflation and guide monetary policy decisions. In Korea, the CPI significantly impacts decisions on interest rates, fiscal policy frameworks, and the Bank of Korea's strategies for economic stability. Given its importance, accurately forecasting the Total CPI is crucial for informed decision-making. Achieving accurate estimation, however, presents several challenges. First, the Korean Total CPI is calculated as a weighted sum of 462 items grouped into 12 categories of goods and services. This heterogeneity makes it difficult to account for all variations in consumer behavior and price dynamics. Second, the monthly frequency of CPI data results in a relatively sparse time series, limiting the performance of the analysis. Furthermore, external factors such as policy changes and pandemics add further volatility to the CPI. To address these challenges, we propose a novel framework consisting of four key components: (1) a hybrid Convolutional Neural Network-Long Short-Term Memory mechanism designed to capture complex patterns in CPI data, enhancing estimation accuracy; (2) multivariate inputs that incorporate CPI component indices alongside auxiliary variables for richer contextual information; (3) data augmentation through linear interpolation to convert monthly data into daily data, optimizing it for highly parametrized deep learning models; and (4) sentiment index derived from Korean CPI-related news articles, providing insights into external factors influencing CPI fluctuations. Experimental results demonstrate that the proposed model outperforms existing approaches in CPI prediction, as evidenced by lower RMSE values. This improved accuracy has the potential to support the development of more timely and effective economic policies.

PMID:40359407 | DOI:10.1371/journal.pone.0321530

Retina. 2025 May 6. doi: 10.1097/IAE.0000000000004511. Online ahead of print.

ABSTRACT

PURPOSE: To compare the prognosis of neovascularization of the disc (NVD) after panretinal photocoagulation (PRP) and/or ranibizumab treatment, based on OCT angiography (OCTA) patterns.

METHODS: In this prospective study, treatment-naive patients with stage IV diabetic retinopathy (DR) and NVD were imaged with 6x6 mm2 OCTA scans. NVD was classified according to OCTA morphological features: different sources (retinal arteries and veins), different activities (exuberant vascular proliferation (EVP)+ and EVP-) and different configurations (type I&II, III and IV). All patients were treated with PRP or in combination with ranibizumab. Patients were monitored monthly to detect the occurrence of vitreous haemorrhage and/or retinal detachment (VH&RD), as well as changes in best corrected visual acuity (BCVA) and NVD.

RESULTS: Among 114 eyes, 35 developed VH&RD (mean onset 6.1 months). Different configurations and EVP status (+/-) significantly affected VH&RD occurrence (p<0.05). NVD regression occurred in 52 eyes, with EVP status significantly influencing resolution (p=0.022). No significant effect was observed on visual acuity (p>0.05).

CONCLUSION: NVD can be classified into different patterns based on morphological features in OCTA, which play a crucial role in the prognosis of NVD patients after PRP and/or ranibizumab treatment.

PMID:40359330 | DOI:10.1097/IAE.0000000000004511

J Phys Chem B. 2025 May 13. doi: 10.1021/acs.jpcb.5c03007. Online ahead of print.

NO ABSTRACT

PMID:40359295 | DOI:10.1021/acs.jpcb.5c03007

Clin Nucl Med. 2025 May 13. doi: 10.1097/RLU.0000000000005967. Online ahead of print.

ABSTRACT

PURPOSE: Sincalide-stimulated cholescintigraphy (SSC) calculates the gallbladder ejection fraction (GBEF) to diagnose functional gallbladder disorder. Currently, artificial intelligence (AI)-driven workflows that integrate real-time image processing and organ function calculation remain unexplored in nuclear medicine practice. This pilot study explored an AI-based application for gallbladder radioactivity tracking.

METHODS: We retrospectively analyzed 20 SSC exams, categorized into 10 easy and 10 challenging cases. Two human operators (H1 and H2) independently annotated the gallbladder regions of interest manually over the course of the 60-minute SSC. A U-Net-based deep learning model was developed to automatically segment gallbladder masks, and a 10-fold cross-validation was performed for both easy and challenging cases. The AI-generated masks were compared with human-annotated ones, with Dice similarity coefficients (DICE) used to assess agreement.

RESULTS: AI achieved an average DICE of 0.746 against H1 and 0.676 against H2, performing better in easy cases (0.781) than in challenging ones (0.641). Visual inspection showed AI was prone to errors with patient motion or low-count activity.

CONCLUSIONS: This study highlights AI's potential in real-time gallbladder tracking and GBEF calculation during SSC. AI-enabled real-time evaluation of nuclear imaging data holds promise for advancing clinical workflows by providing instantaneous organ function assessments and feedback to technologists. This AI-enabled workflow could enhance diagnostic efficiency, reduce scan duration, and improve patient comfort by alleviating symptoms associated with SSC, such as abdominal discomfort due to sincalide administration.

PMID:40359029 | DOI:10.1097/RLU.0000000000005967

Angew Chem Int Ed Engl. 2025 May 13:e202507835. doi: 10.1002/anie.202507835. Online ahead of print.

ABSTRACT

Approximately 40% of marketed drugs exhibit suboptimal pharmacokinetic profiles. Co-crystallization, where pairs of molecules form a multicomponent crystal, constitutes a promising strategy to enhance physicochemical properties without compromising pharmacological activity. However, finding promising co-crystal pairs is resource-intensive, due to the large and diverse range of possible molecular combinations. We present DeepCocrystal, a novel deep learning approach designed to predict co-crystal formation by processing the "chemical language" from a supramolecular vantage point. Rigorous validation of DeepCocrystal showed a balanced accuracy of 78% in realistic scenarios, outperforming existing models. Explainable AI approaches uncovered the decision-making process of DeepCocrystal, showing its capability to learn chemically relevant aspects of the "supramolecular language" that match experimental co-crystallization patterns. By leveraging properties of molecular string representations, DeepCocrystal can also estimate the uncertainty of its predictions. We harness this capability in a challenging prospective study and successfully discovered two novel co-crystals of diflunisal, an anti-inflammatory drug. This study underscores the potential of deep learning - and in particular of chemical language processing - to accelerate co-crystallization and ultimately drug development, in both academic and industrial contexts. DeepCocrystal is available as an easy-to-use web application at https://deepcocrystal.streamlit.app/.

PMID:40358977 | DOI:10.1002/anie.202507835

Interdiscip Sci. 2025 May 13. doi: 10.1007/s12539-025-00706-6. Online ahead of print.

ABSTRACT

The exploration of associations between circular RNAs (circRNAs) and diseases contributes to a deeper understanding of the pathogenesis of diseases. Many computational methods have been proposed for circRNA-disease associations identification. However, these methods still exhibit some limitations such as ignoring the effect of noise. In this paper, we proposed a new knowledge graph attribute mining attention network (KAATCDA) to predict circRNA-disease associations based on knowledge graph attribute network (KGA) and attribute mining attention network (AMA). Firstly, KGA is used to learn the feature representation of diseases. Then, the features of circRNAs are obtained using AMA, which are similar to disease feature representations. Finally, the scores of circRNA-disease associations are predicted based on circRNA feature representation and disease feature representation. Experiments of five-fold cross-validation on two datasets demonstrate that KAATCDA outperforms other state-of-the-art methods. In addition, the case study shows our method can effectively predict unknown circRNA-disease associations.

PMID:40358837 | DOI:10.1007/s12539-025-00706-6

Phys Eng Sci Med. 2025 May 13. doi: 10.1007/s13246-025-01542-0. Online ahead of print.

ABSTRACT

Benign Paroxysmal Positional Vertigo (BPPV) is a prevalent vestibular disorder encountered in clinical settings. Diagnosis of this condition primarily relies on the observation of nystagmus, which involves monitoring the eye movements of patients. However, existing medical equipment for collecting and analyzing nystagmus data has notable limitations and deficiencies. To address this challenge, a comprehensive BPPV nystagmus data collection and intelligent analysis framework has been developed. Our framework leverages a neural network model, Egeunet, in conjunction with mathematical statistical techniques like Fast Fourier Transform (FFT), enabling precise segmentation of eye structures and accurate analysis of eye movement data. Furthermore, an eye movement analysis method has been introduced, designed to enhance clinical decision-making, resulting in more intuitive and clear analysis outcomes. Benefiting from the high sensitivity of our eye movement capture and its robustness in the face of environmental conditions and noise, our BPPV nystagmus data collection and intelligent analysis framework has demonstrated outstanding performance in BPPV detection.

PMID:40358819 | DOI:10.1007/s13246-025-01542-0