Front Cell Dev Biol. 2025 Mar 26;13:1505697. doi: 10.3389/fcell.2025.1505697. eCollection 2025.

ABSTRACT

BACKGROUND: Understanding the cell functionality during disease progression or drugs' mechanism are major challenges for precision medicine. Predictive models describing biological phenotypes can be challenging to obtain, particularly in scenarios where sample availability is limited, such as in the case of rare diseases. Here we propose a new method that reproduces the fibroadipogenic expansion that occurs in muscle wasting.

METHODS: We used immortalized fibroadipogenic progenitor cells (FAPs) and differentiated them into fibroblasts or adipocytes. The method successfully identified FAPs cell differentiation fate using accurate measurements of changes in specific proteins, which ultimately constitute a valid cellular in vitro platform for drug screening. Results were confirmed using primary FAPs differentiation as well as comparison with omics data from proteomics and genomic studies.

RESULTS: Our method allowed us to screen 508 different drugs from 2 compounds libraries. Out of these 508, we identified 4 compounds that reduced fibrogenesis and adipogenesis of ≥30% of fibrogenesis and adipogenesis using immortalized cells. After selecting the optimal dose of each compound, the inhibitory effect on FAP differentiation was confirmed by using primary FAPs from healthy subjects (n = 3) and DMD patients (n = 3). The final 4 selected hits reduced fibrogenic differentiation in healthy and DMD samples. The inhibition of adipogenesis was more evident in DMD samples than healthy samples. After creating an inhibitory map of the tested drugs, we validated the signalling pathways more involved in FAPs differentiation analysing data from proteomic and genomic studies.

CONCLUSION: We present a map of molecular targets of approved drugs that helps in predicting which therapeutic option may affect FAP differentiation. This method allows to study the potential effect of signalling circuits on FAP differentiation after drug treatment providing insights into molecular mechanism of action of muscle degeneration. The accuracy of the method is demonstrated by comparing the signal pathway activity obtained after drug treatment with proteomic and genomic data from patient-derived cells.

PMID:40206397 | PMC:PMC11979640 | DOI:10.3389/fcell.2025.1505697

OMICS. 2025 Apr 10. doi: 10.1089/omi.2024.0205. Online ahead of print.

ABSTRACT

The COVID-19 pandemic phase caused by the SARS-CoV-2 has ended, but the emergence of new variants continues to threaten public health. The public health toolbox for COVID-19 is in need of not only vaccines but also drug discovery against the SARS-CoV-2 virus, the causative agent for the ongoing COVID-19 infections. We report here an in silico molecular docking and dynamics study that uncovered the interactions of 26 flexible heteroarotinoids (FHT18), which are a class of anti-cancer compounds, as potential inhibitors against all 24 SARS-CoV-2 proteins. Of the 624 docked complexes, 69 displayed binding energies between -9.0 and -11.6 kcal/mol, indicating good to strong binding affinities. At least five of these compounds displayed excellent binding affinities against the nonstructural protein 2, papain-like protease, nonstructural protein 4 (Nsp4), proof-reading exoribonuclease, membrane protein, and nucleocapsid protein. Structure-activity relationship (SAR) analyses of these results revealed that a urea linker in place of a thiourea linker, enhanced the hydrophobic side chains attached to the chromane unit, and a CF3 or OCF3 functional group attached to the benzene ring contributed to increased binding affinities. Further, the molecular dynamics simulation study of the best-docked complex FHT18-6c with Nsp4 remained stable for at least 200 ns, leading to decreased structural fluctuations and increased compactness of the binding site. In conclusion, FHT18-6c deserves further translational research to explore its potential for repurposing as a potent drug candidate to combat COVID-19. We also call for continued drug discovery efforts to enrich the public health toolbox for COVID-19.

PMID:40205995 | DOI:10.1089/omi.2024.0205

J Clin Transl Hepatol. 2025 Apr 28;13(4):339-357. doi: 10.14218/JCTH.2024.00402. Epub 2025 Jan 17.

ABSTRACT

Liver injury in Stevens-Johnson syndrome (SJS) and toxic epidermal necrolysis (TEN) is a multifaceted disorder, lacking cohort homogeneity due to a variety of potential causes, including drugs, arsenic and other heavy metals, glyphosate, infections, and ultraviolet radiation. The goals of this review were (1) to analyze the role of diagnostic algorithms in assessing causality for potential culprits involved in the development of liver injury associated with immune-mediated SJS and TEN, which represent immune-based variant disorders within a continuous spectrum. Milder forms are classified as SJS or SJS/TEN overlap, while TEN is known as the most serious form; and (2) to interpret the findings that allow for the characterization of the different types of these disorders. The manuscript is based on an extensive literature search for single case reports, case cohorts, and review articles. Search terms included: Stevens-Johnson Syndrome, Toxic Epidermal Necrolysis, and specific diagnostic algorithms such as the Roussel Uclaf Causality Assessment Method (RUCAM) and the Algorithm of Drug Causality for Epidermal Necrolysis (ALDEN). For the purpose of basic feature description, the uniform term SJS/TEN is used in the current analysis. SJS/TEN presents with five different cohort types: SJS/TEN type (1), which refers to a cohort of SJS/TEN caused by drugs, as assessed by both ALDEN and RUCAM; type (2), representing SJS/TEN due to drugs and assessed by ALDEN only, but not by RUCAM; type (3), which includes a cohort of SJS/TEN caused by drugs, assessed by non-ALDEN and non-RUCAM tools; type (4), which focuses on a cohort of SJS/TEN caused by non-drug culprits, assessed by various tools; and type (5), which considers a cohort of SJS/TEN caused by unknown culprits. Using this new SJS/TEN typology will help better characterize individual features, personalize treatment, and clarify pathogenetic specifics for each of the five disease types. This new SJS/TEN typology provides clarity by replacing issues of inhomogeneity with cohort homogeneity.

PMID:40206276 | PMC:PMC11976437 | DOI:10.14218/JCTH.2024.00402

Pharm Stat. 2025 May-Jun;24(3):e70011. doi: 10.1002/pst.70011.

ABSTRACT

Patient enrollment can be a substantial burden in rare disease trials. One potential approach is to incorporate external control (EC) into concurrent randomized trials, or EC borrowing, to reduce such burden. Extensive research has been conducted to explore statistical methodologies. As in all designs, type I error control is essential. Conditional type I error rate has been used in the literature as the de facto metrics for type I error rate. However, research has shown that controlling the conditional type I error rate at the alpha level will disallow EC borrowing. Therefore, EC borrowing is practically at an impasse. Kopp-Schneider et al. concluded that a more appropriate metrics for type I error is necessary. We show that a trial with EC borrowing can be considered as a two-stage adaptive design. With this perspective, we propose to define type I error as the weighted averages of conditional type I error rate in trials with EC borrowing. Dynamic borrowing methods for controlling type I error are proposed.

PMID:40205746 | DOI:10.1002/pst.70011

Front Artif Intell. 2025 Mar 26;8:1451865. doi: 10.3389/frai.2025.1451865. eCollection 2025.

ABSTRACT

As we become increasingly dependent on technology in our daily lives, the usability of HCIs is a key driver of individual empowerment for us all. A primary focus of AI systems has been to make HCIs easier to use by identifying what users need and agentively taking over some of the cognitive work users would have otherwise performed, as such, they are becoming our delegates. To become effective and reliable delegates, AI agents need to understand all relevant situational semantic context surrounding a user's need and how the tools of the HCI can be leveraged. Current ML systems have fundamental semantic gaps in bespoke human context, real-time world knowledge, and how those relate to HCI tooling. These challenges are difficult to close due factors such as privacy, continual learning, access to real-time context, and how deeply integrated the semantics are with in-context learning. As such, we need to research and explore new ways to safely capture, compactly model, and incrementally evolve semantics in ways that can efficiently integrate into how AI systems act on our behalf. This article presents a thought experiment called the Game of Delegation as a lens to view the effectiveness of delegation and the semantic efficiency with which the delegation was achieved.

PMID:40206708 | PMC:PMC11979188 | DOI:10.3389/frai.2025.1451865

Med J Aust. 2025 Apr 10. doi: 10.5694/mja2.52650. Online ahead of print.

ABSTRACT

Paediatric cancer is the leading cause of disease-related death in Australian children. Limited research focuses on cancer in Aboriginal and Torres Strait Islander children. Although there appears to be a lower incidence of cancer overall in Aboriginal and Torres Strait Islander children compared with non-Indigenous children, a high proportion of Aboriginal and Torres Strait Islander children are diagnosed with acute myeloid leukaemia. Five-year overall survival is lower for many cancer types in Aboriginal and Torres Strait Islander children. There is a need for Indigenous-specific research focused on molecular and genetic profiles, pharmacogenomics and survivorship, both within Australia and globally. Future research in this space should be co-designed and led by Aboriginal and Torres Strait Islander communities; alongside clinicians, researchers and services to ensure that the priorities of Aboriginal and Torres Strait Islander people are met.

PMID:40207417 | DOI:10.5694/mja2.52650

Mol Ther Oncol. 2025 Mar 8;33(2):200965. doi: 10.1016/j.omton.2025.200965. eCollection 2025 Jun 18.

ABSTRACT

Breast cancer patients with estrogen receptor-negative (ERneg) status, encompassing triple negative breast cancer (TNBC) and human epidermal growth factor receptor 2 positive breast cancer, are confronted with a heightened risk of drug resistance, often leading to early recurrence; the biomarkers and biological processes associated with recurrence is still unclear. In this study, we analyzed bulk RNA sequencing (RNA-seq) data from 285 cancer and paracancerous samples from 155 TNBC patients, along with transcriptome data from 11 independent public cohorts comprising 7,449 breast cancer patients and 26 single-cell RNA-seq datasets. Our results revealed differential enrichment of nerve-related pathways between TNBC patients with and without 10-year recurrence-free survival. We developed an early recurrence index (ERI) using a machine learning model and constructed a nomogram that accurately predicts the 10-year survival of ERneg patients (area under the curve [AUC]Training = 0.79; AUCTest = 0.796). Further analysis linked ERI to enhanced neural function and immunosuppression. Additionally, we identified EN1, the most significant ERI gene, as a potential biomarker that may regulate the tumor microenvironment and sensitize patients to immunotherapy.

PMID:40207200 | PMC:PMC11981748 | DOI:10.1016/j.omton.2025.200965

Front Pharmacol. 2025 Mar 26;16:1558897. doi: 10.3389/fphar.2025.1558897. eCollection 2025.

ABSTRACT

BACKGROUND: Despite the emergence of numerous innovative targeted therapies for the management of pediatric inflammatory bowel disease (IBD), azathioprine continues to be a pivotal first-line therapeutic agent. Nonetheless, the considerable frequency of myelosuppression associated with its use warrants careful consideration and further investigation. This study aims to investigate the application of pharmacogenomics in Chinese pediatric IBD treated with azathioprine, and to elucidate its association with the occurrence of myelosuppression.

METHODS: We conducted a retrospective analysis to determine the prevalence of pharmacogenetic abnormalities and thiopurine-induced myelosuppression in Chinese pediatric patients with IBD.

RESULTS: Among the 227 patients underwent pharmacogenetic testing, abnormal genetypes occurred in 66 patients, among which 7 patients exhibited aberrant TPMT and 59 had aberrant NUDT15. Of the 58 patients who were treated with azathioprine, 23 cases experienced myelosuppression. All three children with heterozygous mutations in NUDT15 developed leukopenia following azathioprine treatment. Among patients with normal pharmacogenetic results, 20 cases (36.4%) developed myelosuppression, while 35 cases (63.6%) did not. The dose of azathioprine was below the recommended level in guidelines. The mean dose of azathioprine (mg/kg/day) in the myelosuppression group was 1.22 ± 0.32, compared to 1.42 ± 0.42 in the non-myelosuppression group, which represented a statistically significant difference (p < 0.05). Age, gender, and the use of concomitant biologics, mesalazine, or glucocorticoids did not show significant differences between the groups (p > 0.05).

CONCLUSION: NUDT15 C415T is prevalent in China and is associated with an increased risk of azathioprine-induced myelosuppression. A reduced dose of azathioprine should be considered for Chinese pediatric patients with IBD, even in those with normal pharmacogenetic profiles.

PMID:40206080 | PMC:PMC11979209 | DOI:10.3389/fphar.2025.1558897

Front Pharmacol. 2025 Mar 26;16:1590955. doi: 10.3389/fphar.2025.1590955. eCollection 2025.

ABSTRACT

[This corrects the article DOI: 10.3389/fphar.2024.1526101.].

PMID:40206068 | PMC:PMC11979611 | DOI:10.3389/fphar.2025.1590955

Pharmacogenomics. 2025 Apr 9:1-6. doi: 10.1080/14622416.2025.2489920. Online ahead of print.

ABSTRACT

INTRODUCTION: Genotype-guided tacrolimus management is not routine in clinical practice despite the availability of Clinical Pharmacogenetics Implementation Consortium dosing guidelines. Prior surveys have evaluated patient and provider perspectives of pharmacogenetics (PGx) in transplant, but limited recent data exists on tacrolimus PGx implementation across United States transplant centers.

METHODS: An electronic survey was distributed to transplant pharmacists regarding utilization of tacrolimus PGx, methods of implementing PGx, and barriers to clinical implementation. A survey response was requested for each organ program within the transplant center.

RESULTS: A total of 90 programs from 69 transplant centers (28.1% of active U.S. transplant centers) responded to the survey. Tacrolimus PGx was utilized for patient care in 14 programs (15.6%). There was substantial variability in the implementation methods and application of tacrolimus PGx results among transplant programs. In programs that had not implemented tacrolimus PGx, common barriers for implementation included PGx testing cost and availability and lack of evidence for clinical utility.

CONCLUSION: Implementation of PGx guided tacrolimus in solid organ transplant centers remains limited with heterogeneity in the implementation approach. Additional research is needed to establish the clinical utility of PGx guided tacrolimus and education on reimbursement and testing resources may help to increase uptake.

PMID:40205800 | DOI:10.1080/14622416.2025.2489920

MethodsX. 2025 Mar 22;14:103267. doi: 10.1016/j.mex.2025.103267. eCollection 2025 Jun.

ABSTRACT

In this study, we present a detailed workflow for the isolation, quantitation, and evaluation of mucin proteins. These methods are applicable to a variety of biological, mucin-containing samples from the airways and other mucosal organ systems. While this report focuses on the salivary MUC5B protein from the respiratory system, the presented methodologies can be applied to other mucins, contributing to a broader application of these techniques. We used a simplified isopycnic centrifugation to purify and enrich MUC5B from human saliva. Isolated MUC5B was then subjected to a Bradford protein assay using a bovine submaxillary mucin (BSM) standard, which more accurately reflects the mucin concentration in our samples compared to a bovine serum albumin (BSA) standard. Additionally, we compare the mucin levels following quantitation using agarose polyacrylamide gel electrophoresis. Our findings show a near 2-fold increase in quantitation from the more representative, BSM standard, suggesting its importance for mucin studies. These methods support a wide range of experimental applications looking to assess mucins, thereby contributing to the broader field of mucin studies and advancing our understanding of the implications of mucins in health and disease.•A streamlined, one-step isopycnic ultracentrifugation to isolate MUC5B from human saliva•A Mucin Bradford assay that is modified from existing Bradford assay techniques to better quantitate mucin for mucin studies•An agarose-polyacrylamide gel electrophoresis method used to visualize and confirm the isolation and quantitation of mucin.

PMID:40207064 | PMC:PMC11981757 | DOI:10.1016/j.mex.2025.103267

Pharmazie. 2025 Mar 31;80(1):10-16. doi: 10.1691/ph.2025.4544.

ABSTRACT

Transmembrane 16A (TMEM16A) is highly expressed in interstitial cells of Cajal (ICC) and participates in ICC-mediated rhythmic contractile activity of intestinal smooth muscle. TMEM16A is also expressed in epithelium of intestine with a minor contributor to transepithelial fluid secretion, while other unidentified Ca2+ -activated Cl - channels (unCaCCs) are mainly responsible for this physiological process. TMEM16A/CaCCs dysfunction can lead to disorders of intestinal motility and transepithelial fluid secretion. TMEM16A/CaCCs regulators are important tools to identify unCaCCs and study the physiopathological functions related to TMEM16A/CaCCs. In the present study, coumarins were identified as TMEM16A inhibitors in a concentration- and time-dependent manner in TMEM16A-expressed Fischer rat thyroid (FRT) epithelial cells. Coumarins attenuated intestinal motility by inhibiting TMEM16A in vivo and ex vivo. Coumarins inhibited CaCCs-mediated Cl- currents induced by ATP in T84 and HT-29 cells or by carbachol (CCh) in mouse colonic mucosa with reduction of ATP-induced increase of cytoplasmic Ca2+ concentration in HT-29 cells. Coumarins inhibited basolateral Ca2+ -activated K+ channels without affecting Na + /K + -ATPase activity in mouse colonic mucosa. Coumarins did not show inhibition of cystic fibrosis transmembrane conductance regulator (CFTR), but mild activation of CFTR-mediated Cl - currents under the low concentration forskolin (FSK) in CFTR-expressed FRT cells, while coumarins did not activate CFTR-mediated Cl- currents in mouse colonic mucosa. This study was the first to demonstrate that coumarins attenuate intestinal motility by inhibiting TMEM16A, which may provide a strategy for clinical drug intervention aimed at reducing secretory diarrhea.

PMID:40205671 | DOI:10.1691/ph.2025.4544

Br J Nutr. 2025 Apr 10:1-13. doi: 10.1017/S0007114525000522. Online ahead of print.

ABSTRACT

One of the most significant challenges in research related to nutritional epidemiology is the achievement of high accuracy and validity of dietary data to establish an adequate link between dietary exposure and health outcomes. Recently, the emergence of artificial intelligence (AI) in various fields has filled this gap with advanced statistical models and techniques for nutrient and food analysis. We aimed to systematically review available evidence regarding the validity and accuracy of AI-based dietary intake assessment methods (AI-DIA). In accordance with PRISMA guidelines, an exhaustive search of the EMBASE, PubMed, Scopus and Web of Science databases was conducted to identify relevant publications from their inception to 1 December 2024. Thirteen studies that met the inclusion criteria were included in this analysis. Of the studies identified, 61·5 % were conducted in preclinical settings. Likewise, 46·2 % used AI techniques based on deep learning and 15·3 % on machine learning. Correlation coefficients of over 0·7 were reported in six articles concerning the estimation of calories between the AI and traditional assessment methods. Similarly, six studies obtained a correlation above 0·7 for macronutrients. In the case of micronutrients, four studies achieved the correlation mentioned above. A moderate risk of bias was observed in 61·5 % (n 8) of the articles analysed, with confounding bias being the most frequently observed. AI-DIA methods are promising, reliable and valid alternatives for nutrient and food estimations. However, more research comparing different populations is needed, as well as larger sample sizes, to ensure the validity of the experimental designs.

PMID:40207441 | DOI:10.1017/S0007114525000522

Front Comput Neurosci. 2025 Mar 26;19:1545971. doi: 10.3389/fncom.2025.1545971. eCollection 2025.

ABSTRACT

In recent years, the integration of machine vision and neuroscience has provided a new perspective for deeply understanding visual information. This paper proposes an innovative deep learning model, NeuroFusionNet, designed to enhance the understanding of visual information by integrating fMRI signals with image features. Specifically, images are processed by a visual model to extract region-of-interest (ROI) features and contextual information, which are then encoded through fully connected layers. The fMRI signals are passed through 1D convolutional layers to extract features, effectively preserving spatial information and improving computational efficiency. Subsequently, the fMRI features are embedded into a 3D voxel representation to capture the brain's activity patterns in both spatial and temporal dimensions. To accurately model the brain's response to visual stimuli, this paper introduces a Mutli-scale fMRI Timeformer module, which processes fMRI signals at different scales to extract both fine details and global responses. To further optimize the model's performance, we introduce a novel loss function called the fMRI-guided loss. Experimental results show that NeuroFusionNet effectively integrates image and brain activity information, providing more precise and richer visual representations for machine vision systems, with broad potential applications.

PMID:40207297 | PMC:PMC11978827 | DOI:10.3389/fncom.2025.1545971

Med Image Comput Comput Assist Interv. 2024 Oct;15011:37-47. doi: 10.1007/978-3-031-72120-5_4. Epub 2024 Oct 3.

ABSTRACT

The robustness of supervised deep learning-based medical image classification is significantly undermined by label noise in the training data. Although several methods have been proposed to enhance classification performance in the presence of noisy labels, they face some challenges: 1) a struggle with class-imbalanced datasets, leading to the frequent overlooking of minority classes as noisy samples; 2) a singular focus on maximizing performance using noisy datasets, without incorporating experts-in-the-loop for actively cleaning the noisy labels. To mitigate these challenges, we propose a two-phase approach that combines Learning with Noisy Labels (LNL) and active learning. This approach not only improves the robustness of medical image classification in the presence of noisy labels but also iteratively improves the quality of the dataset by relabeling the important incorrect labels, under a limited annotation budget. Furthermore, we introduce a novel Variance of Gradients approach in the LNL phase, which complements the loss-based sample selection by also sampling under-represented examples. Using two imbalanced noisy medical classification datasets, we demonstrate that our proposed technique is superior to its predecessors at handling class imbalance by not misidentifying clean samples from minority classes as mostly noisy samples. Code available at: https://github.com/Bidur-Khanal/imbalanced-medical-active-label-cleaning.git.

PMID:40207034 | PMC:PMC11981598 | DOI:10.1007/978-3-031-72120-5_4

Mayo Clin Proc Digit Health. 2024 Nov 27;3(1):100183. doi: 10.1016/j.mcpdig.2024.11.004. eCollection 2025 Mar.

ABSTRACT

OBJECTIVE: To develop a deep learning framework to predict lower-limb joint kinematics from inertial measurement unit (IMU) data across multiple gait tasks (walking, jogging, and running) and evaluate the impact of dynamic time warping (DTW) on reducing prediction errors.

PATIENTS AND METHODS: Data were collected from 18 participants fitted with IMUs and an optical motion capture system between May 25, 2023, and May 30, 2023. A long short-term memory autoencoder supervised regression model was developed. The model consisted of multiple long short-term memory and convolution layers. Acceleration and gyroscope data from the IMUs in 3 axes and their magnitude for the proximal and distal sensors of each joint (hip, knee, and ankle) were inputs to the model. Optical motion capture kinematics were considered ground truth and used as an output to train the prediction model.

RESULTS: The deep learning models achieved a root-mean-square error of less than 6° for hip, knee, and ankle joint sagittal plane angles, with the ankle showing the lowest error (5.1°). Task-specific models reported enhanced performance during certain gait phases, such as knee flexion during running. The application of DTW significantly reduced root-mean-square error across all tasks by at least 3° to 4°. External validation of independent data confirmed the model's generalizability.

CONCLUSION: Our findings underscore the potential of IMU-based deep learning models for joint kinematic predictions, offering a practical solution for remote and continuous biomechanical assessments in health care and sports science.

PMID:40207006 | PMC:PMC11975825 | DOI:10.1016/j.mcpdig.2024.11.004

Mayo Clin Proc Digit Health. 2025 Jan 10;3(1):100194. doi: 10.1016/j.mcpdig.2025.100194. eCollection 2025 Mar.

ABSTRACT

OBJECTIVE: To develop a fully end-to-end deep learning framework capable of estimating left ventricular ejection fraction (LVEF), estimating patient age, and classifying patient sex from echocardiographic videos, including videos collected using handheld cardiac ultrasound (HCU).

PATIENTS AND METHODS: Deep learning models were trained using retrospective transthoracic echocardiography (TTE) data collected in Mayo Clinic Rochester and surrounding Mayo Clinic Health System sites (training: 6432 studies and internal validation: 1369 studies). Models were then evaluated using retrospective TTE data from the 3 Mayo Clinic sites (Rochester, n=1970; Arizona, n=1367; Florida, n=1562) before being applied to a prospective dataset of handheld ultrasound and TTE videos collected from 625 patients. Study data were collected between January 1, 2018 and February 29, 2024.

RESULTS: Models showed strong performance on the retrospective TTE datasets (LVEF regression: root mean squared error (RMSE)=6.83%, 6.53%, and 6.95% for Rochester, Arizona, and Florida cohorts, respectively; classification of LVEF ≤40% versus LVEF > 40%: area under curve (AUC)=0.962, 0.967, and 0.980 for Rochester, Arizona, and Florida, respectively; age: RMSE=9.44% for Rochester; sex: AUC=0.882 for Rochester), and performed comparably for prospective HCU versus TTE data (LVEF regression: RMSE=6.37% for HCU vs 5.57% for TTE; LVEF classification: AUC=0.974 vs 0.981; age: RMSE=10.35% vs 9.32%; sex: AUC=0.896 vs 0.933).

CONCLUSION: Robust TTE datasets can be used to effectively power HCU deep learning models, which in turn demonstrates focused diagnostic images can be obtained with handheld devices.

PMID:40207004 | PMC:PMC11975991 | DOI:10.1016/j.mcpdig.2025.100194

Mayo Clin Proc Digit Health. 2025 Jan 21;3(1):100195. doi: 10.1016/j.mcpdig.2025.100195. eCollection 2025 Mar.

ABSTRACT

OBJECTIVE: To develop an efficient and automated method for selecting appropriate coronary angiography videos for training deep learning models, thereby improving the accuracy and efficiency of medical image analysis.

PATIENTS AND METHODS: We developed deep learning models using 232 coronary angiographic studies from the Mayo Clinic. We utilized 2 state-of-the-art convolutional neural networks (CNN: ResNet and X3D) to identify low-quality angiograms through binary classification (satisfactory/unsatisfactory). Ground truth for the quality of the input angiogram was determined by 2 experienced cardiologists. We validated the developed model in an independent dataset of 3208 procedures from 3 Mayo sites.

RESULTS: The 3D-CNN models outperformed their 2D counterparts, with the X3D-L model achieving superior performance across all metrics (AUC 0.98, accuracy 0.96, precision 0.87, and F1 score 0.92). Compared with 3D models, 2D architectures are smaller and less computationally complex. Despite having a 3D architecture, the X3D-L model had lower computational demand (19.34 Giga Multiply Accumulate Operation) and parameter count (5.34 M) than 2D models. When validating models on the independent dataset, slight decreases in all metrics were observed, but AUC and accuracy remained robust (0.95 and 0.92, respectively, for the X3D-L model).

CONCLUSION: We developed a rapid and effective method for automating the selection of coronary angiogram video clips using 3D-CNNs, potentially improving model accuracy and efficiency in clinical applications. The X3D-L model reports a balanced trade-off between computational efficiency and complexity, making it suitable for real-life clinical applications.

PMID:40206993 | PMC:PMC11975815 | DOI:10.1016/j.mcpdig.2025.100195

Comput Struct Biotechnol J. 2025 Mar 7;28:94-105. doi: 10.1016/j.csbj.2025.02.039. eCollection 2025.

ABSTRACT

Corneal diseases present a significant challenge to global health. Given the uneven distribution of ophthalmic resources, the development of a system to facilitate remote diagnosis of corneal diseases is particularly crucial. In this study, we developed an artificial intelligence system named Gancor, based on a large-scale clinical dataset comprising 9669 anterior segment (AS) images and corresponding corneal fluorescein staining (CFS) images from the Affiliated Eye Hospital of Nanchang University, as well as 967 pairs of AS-CFS images captured via smartphone from the Jiangxi Province Division of National Clinical Research Center for Ocular Diseases. The system utilizes Generative Adversarial Networks (GANs) to convert AS images into CFS images for the screening of 11 common corneal diseases. Objective assessments of the generated CFS images were conducted using Mean Absolute Error (MAE), Peak Signal-to-Noise Ratio (PSNR), and Structural Similarity Index (SSIM), along with subjective evaluations by three experienced ophthalmologists, confirming the high quality and diagnostic relevance of the synthesized images. In terms of diagnostic performance for corneal diseases, the accuracy rate exceeded 75 %, and the Area Under the Curve (AUC) value reached above 0.90. This innovative approach not only provides images with greater diagnostic value for telemedicine but also enhances the efficiency of remote diagnosis, offering an effective tool for achieving the goal of comprehensive, equitable, and accessible eye care services.

PMID:40206787 | PMC:PMC11981786 | DOI:10.1016/j.csbj.2025.02.039

J Hepatocell Carcinoma. 2025 Apr 5;12:685-704. doi: 10.2147/JHC.S505752. eCollection 2025.

ABSTRACT

BACKGROUND: With the global epidemic of obesity and diabetes, non-alcoholic fatty liver disease (NAFLD) is becoming the most common chronic liver disease, and NASH is increasingly becoming a major risk factor for hepatocellular carcinoma. Therefore, it is essential to explore novel biomarkers in NASH-related HCC.

METHODS: Deep Learning (DL) methods are a promising and encouraging tool widely used in genomics by automatically applying neural networks (NNs). Therefore, DL, "limma package", weighted gene co-expression network analysis (WGCNA), and Protein-Protein Interaction Networks (PPI) were used to screen feature genes. Real-time quantitative PCR was used to validate the expression of feature genes in the NAFLD mice model. Enrichment and single-cell sequencing analyses of single genes were performed to investigate the role of feature genes in NASH-related HCC.

RESULTS: Combined core genes screened by DL in NAFLD with important genes in metabolic syndrome, six feature genes (FDFT1, TNFSF10, DNAJC16, RDH11, PGRMC1, and MYC) were obtained. ROC analysis demonstrates the model's superiority with the AUC was 0.983 (0.9241-0.98885). Animal experiments based on NAFLD mouse models have also shown that FDFT1, TNFSF10, DNAJC16, RDH11, and PGRMC1 have a higher expression in NAFLD livers. Among the feature genes, FDFT1 and PGRMC1 showed significant expression trends and outstanding diagnosis value in NASH-HCC.

CONCLUSION: In conclusion, FDFT1 and PGRMC1 are key enzymes in the cholesterol synthesis pathway, our study validates the important role of cholesterol metabolism in NAFLD from another perspective, implying they may be new prognostic and diagnostic markers for NASH-HCC.

PMID:40206734 | PMC:PMC11980943 | DOI:10.2147/JHC.S505752