bims-aukdir Biomed News
on Automated knowledge discovery in diabetes research
Issue of 2025–04–27
fourteen papers selected by
Mott Given



  1. Int J Retina Vitreous. 2025 Apr 22. 11(1): 48
       BACKGROUND: To evaluate the efficacy of artificial intelligence (AI) in screening for diabetic retinopathy (DR) using fundus images and optical coherence tomography (OCT) in comparison to traditional screening methods.
    METHODS: This systematic review was registered with PROSPERO (ID: CRD42024560750). Systematic searches were conducted in PubMed Medline, Cochrane Central, ScienceDirect, and Web of Science using keywords such as "diabetic retinopathy," "screening," and "artificial intelligence." Only studies published in English from 2019 to July 22, 2024, were considered. We also manually reviewed the reference lists of relevant reviews. Two independent reviewers assessed the risk of bias using the QUADAS-2 tool, resolving disagreements through discussion with the principal investigator. Meta-analysis was performed using MetaDiSc software (version 1.4). To calculate combined sensitivity, specificity, summary receiver operating characteristic (SROC) plots, forest plots, and subgroup analyses were performed according to clinician type (ophthalmologists vs. retina specialists) and imaging modality (fundus images vs. fundus images + OCT).
    RESULTS: 18 studies were included. Meta-analysis showed that AI systems demonstrated superior diagnostic performance compared to doctors, with the pooled sensitivity, specificity, diagnostic odds ratio, and Cochrane Q index of the AI being 0.877, 0.906, 0.94, and 153.79 accordingly. The Fagan nomogram analysis further confirmed the strong diagnostic value of AI. Subgroup analyses revealed that factors like imaging modality, and doctor expertise can influence diagnostic performance.
    CONCLUSION: AI systems have demonstrated strong diagnostic performance in detecting diabetic retinopathy, with sensitivity and specificity comparable to or exceeding traditional clinicians.
    Keywords:  Artificial intelligence; Diabetic retinopathy; Screening
    DOI:  https://doi.org/10.1186/s40942-025-00670-9
  2. Invest Ophthalmol Vis Sci. 2025 Apr 01. 66(4): 72
       Purpose: To develop a deep learning method for vessel segmentation in fundus images, measure retinal vessels, and study the connection between retinal vascular features and systemic indicators in diabetic patients.
    Methods: We conducted a study on patients with diabetes mellitus (DM) at various stages of diabetic retinopathy (DR) using data from the Joint Asia Diabetes Evaluation (JADE) Register. All participants underwent comprehensive clinical assessments, including anthropometric measurements, laboratory tests, and fundus photography, during each follow-up visit (2.81 average follow-up visits). A custom U-Net deep learning model utilizing a variety of open-source datasets was developed for the segmentation and measurement of retinal vessels. We investigated the relationship between systemic indicators and the severity of DR, analyzing the correlation coefficients between systemic indicators and retinal vascular characteristics.
    Results: We enrolled a total of 637 patients diagnosed with DM and collected 3575 series of photographs for analysis. Some of the systemic indicators and retinal vascular metrics, including central retinal arteriolar equivalent, central retinal venular equivalent, arteriole-to-venule ratio, and fractal dimension, were significantly correlated with the severity of diabetic retinopathy (P < 0.05). Some physical characteristics, hematological parameters, renal function parameters, metabolism-related parameters, biochemical markers such as folic acid and fasting insulin, liver enzymes, and macrovascular indicators were significantly correlated with certain retinal vascular metrics (P < 0.05).
    Conclusions: Multiple systemic indicators were identified as significantly associated with the advancement of diabetic retinopathy and retinal vascular metrics. Utilizing deep learning techniques for vessel segmentation and measurement on color fundus photographs can help elucidate the connections between retinal vascular characteristics and systemic indicators.
    DOI:  https://doi.org/10.1167/iovs.66.4.72
  3. J Biomed Phys Eng. 2025 Apr;15(2): 137-158
       Background: Diabetic retinopathy (DR), a diabetes complication, causes blindness by damaging retinal blood vessels. While deep learning has advanced DR diagnosis, many models face issues like inconsistent performance, limited datasets, and poor interpretability, reducing their clinical utility.
    Objective: This research aimed to develop and evaluate a deep learning structure combining Convolutional Neural Networks (CNNs) and transformer architecture to improve the accuracy, reliability, and generalizability of DR detection and severity classification.
    Material and Methods: This computational experimental study leverages CNNs to extract local features and transformers to capture long-range dependencies in retinal images. The model classifies five types of retinal images and assesses four levels of DR severity. The training was conducted on the augmented APTOS 2019 dataset, addressing class imbalance through data augmentation techniques. Performance metrics, including accuracy, Area Under the Curve (AUC), specificity, and sensitivity, were used for metric evaluation. The model's robustness was further validated using the IDRiD dataset under diverse scenarios.
    Results: The model achieved a high accuracy of 94.28% on the APTOS 2019 dataset, demonstrating strong performance in both image classification and severity assessment. Validation on the IDRiD dataset confirmed its generalizability, achieving a consistent accuracy of 95.23%. These results indicate the model's effectiveness in accurately diagnosing and assessing DR severity across varied datasets.
    Conclusion: The proposed Artificial intelligence (AI)-powered diagnostic tool improves diabetic patient care by enabling early DR detection, preventing progression and reducing vision loss. The proposed AI-powered diagnostic tool offers high performance, reliability, and generalizability, providing significant value for clinical DR management.
    Keywords:   Artificial Intelligence; Convolutional Neural Networks; Deep Learning; Diabetic Retinopathy; Fundus Oculi
    DOI:  https://doi.org/10.31661/jbpe.v0i0.2408-1811
  4. Sci Rep. 2025 Apr 22. 15(1): 13858
      Retinal screening provides for earlier detection of diabetic retinopathy (DR) as well as prompt diagnosis. Recognizing DR utilizing color fundus imaging needs qualified specialists to know about the presence and significance of a few insignificant features that when it linked with complicated categorization structure create this as an engaging and difficult task. The automatic progression of DR detection consumes more time and cost. To conquer these gaps, a hybrid network structure for DR detection utilizing retinal fundus image named Mobile Maxout network (MM-Net). Here, MM-Net is merged with the merging of MobileNet and Deep Maxout Network (DMN). At first, the input retinal image is pre-processed by utilizing a median filter. Then, optic disk (OD) segmentation progress is done by utilizing the active contour model as well as the filtered image is also passed through blood vessel segmentation that is progressed by O-SegNet. Afterwards, the segmented and input images are allowed into the feature extraction phase. Finally, DR detection is achieved by the proposed MM-Net. The analytic metrics deployed for MM-Net, such as accuracy, sensitivity and specificity achieved 89.2%, 90.5%, and 92.0%.
    Keywords:  Deep Maxout Network (DMN); Diabetic retinopathy (DR); MobileNet; Optic disk (OD) segmentation; Retinal fundus image
    DOI:  https://doi.org/10.1038/s41598-025-97675-z
  5. Asia Pac J Public Health. 2025 Apr 18. 10105395251332798
      This study aimed to develop machine learning (ML) models to predict diabetic complications in patients with type 2 diabetes (T2D) in Malaysia. Data from the Malaysian National Diabetes Registry and Death Register were used to develop predictive models for five complications: all-cause mortality, retinopathy, nephropathy, ischemic heart disease (IHD), and cerebrovascular disease (CeVD). Accurate predictions may enable targeted preventive intervention and optimal disease management. The cohort comprised 90 933 T2D patients treated at public health clinics in southern Malaysia from 2011 to 2021. Seven ML algorithms were tested, with the Light Gradient Boosting Machine (LGBM) demonstrating the best performance. LGBM models achieved ROC-AUC scores of 0.84 for all-cause mortality, 0.71 for retinopathy, 0.71 for nephropathy, 0.66 for IHD, and 0.74 for CeVD. These findings support integrating ML models, particularly LGBM, into clinical practice for predicting diabetes complications. Further optimization and validation are necessary to enhance applicability across diverse populations.
    Keywords:  diabetes complications; diabetes registry; machine learning; predictive models; type 2 diabetes
    DOI:  https://doi.org/10.1177/10105395251332798
  6. QJM. 2025 Apr 24. pii: hcaf101. [Epub ahead of print]
       OBJECTIVE: This study aimed to develop and validate a risk prediction model for 1-year CKD progression in patients with T2DM and CKD by employing various machine learning (ML) algorithms.
    METHODS: This study included a total of 12,151 patients with T2DM and CKD with estimated glomerular filtration rate (eGFR) between 30 and 59.9 mL/min/1.73 m2 from a tertiary hospital in Wuhan, enrolled between 2012 and 2024. The cohort was divided into a training set of 5,954 patients, an internal validation set of 2,552 patients, and an external validation set of 3,645 patients. We developed 1-year CKD progression risk prediction models using 10 different machine learning algorithms. CKD progression was defined as a decline in eGFR by more than 30% from baseline and/or a reduction in eGFR to below 15 mL/min/1.73 m2. The SHAP (SHapley Additive exPlanations) method was utilized to explain the predictions of a model.
    RESULTS: Among the 10 ML models, the XGBoost model achieved the best predictive performance for 1-year progression of kidney function with an AUC of 0.906 in the internal validation set and 0.768 in the external validation set. The final predictive model incorporating only nine variables has been implemented into a web application to enhance its usability in clinical settings.
    CONCLUSION: Our findings suggest that the XGBoost model may serve as a valuable decision-support tool for predicting kidney function decline in patients with T2DM and CKD.
    Keywords:  CKD; Kidney function progression; Machine learning; Predictive model; T2DM
    DOI:  https://doi.org/10.1093/qjmed/hcaf101
  7. Stud Health Technol Inform. 2025 Apr 24. 324 84-89
       BACKGROUND: Type 2 diabetes (T2D) continues to present a global public health challenge due to its increasing prevalence. Early diagnosis is critical for preventing complications, but current screening methods often fail to detect early diabetic conditions.
    OBJECTIVES: This study aimed to classify T2D patients from healthy individuals using high-resolution N-glycan profiling.
    METHODS: Glycan profiling was performed on serum samples from 161 individuals using capillary electrophoresis with laser-induced fluorescence detection. Different classification methods were fine-tuned using hyperparameter optimization and feature selection techniques, and their performance was comprehensively evaluated based on quality metrics.
    RESULTS: The Extra Trees Classifier outperformed the other models with the highest median AUC, demonstrating robust accuracy (0.8982), sensitivity (0.8966), and specificity (0.9000).
    CONCLUSION: N-glycan profiling combined with machine learning provides a promising approach for early T2D detection. The Extra Trees Classifier showed exceptional predictive performance, warranting further investigation with larger datasets to validate its clinical applicability.
    Keywords:  Classification; Hyperparameter optimization; Machine learning; N-glycan; Type 2 Diabetes
    DOI:  https://doi.org/10.3233/SHTI250166
  8. J Med Internet Res. 2025 Apr 23. 27 e60367
       BACKGROUND: Retinopathy of prematurity (ROP) is the leading preventable cause of childhood blindness. A timely intravitreal injection of antivascular endothelial growth factor (anti-VEGF) is required to prevent retinal detachment with consequent vision impairment and loss. However, anti-VEGF has been reported to be associated with ROP reactivation. Therefore, an accurate prediction of reactivation after treatment is urgently needed.
    OBJECTIVE: To develop and validate prediction models for reactivation after anti-VEGF intravitreal injection in infants with ROP using multimodal machine learning algorithms.
    METHODS: Infants with ROP undergoing anti-VEGF treatment were recruited from 3 hospitals, and conventional machine learning, deep learning, and fusion models were constructed. The areas under the curve (AUCs), accuracy, sensitivity, and specificity were used to show the performances of the prediction models.
    RESULTS: A total of 239 cases with anti-VEGF treatment were recruited, including 90 (37.66%) with reactivation and 149 (62.34%) nonreactivation cases. The AUCs for the conventional machine learning model were 0.806 and 0.805 in the internal validation and test groups, respectively. The average AUC, sensitivity, and specificity in the test for the deep learning model were 0.787, 0.800, and 0.570, respectively. The specificity, AUC, and sensitivity for the fusion model were 0.686, 0.822, and 0.800 in a test, separately.
    CONCLUSIONS: We constructed 3 prediction models for ROP reactivation. The fusion model achieved the best performance. Using this prediction model, we could optimize strategies for treating ROP in infants and develop better screening plans after treatment.
    Keywords:  anti-VEGF; deep learning; machine learning; prediction; reactivation; retinopathy of prematurity
    DOI:  https://doi.org/10.2196/60367
  9. Front Endocrinol (Lausanne). 2025 ;16 1543192
       Objective: This study aims to conduct an in-depth analysis of diabetic foot ulcer (DFU) images using deep learning models, achieving automated segmentation and classification of the wounds, with the goal of exploring the application of artificial intelligence in the field of diabetic foot care.
    Methods: A total of 671 images of DFU were selected for manual annotation of the periwound erythema, ulcer boundaries, and various components within the wounds (granulation tissue, necrotic tissue, tendons, bone tissue, and gangrene). Three instance segmentation models (Mask2former, Deeplabv3plus, and Swin-Transformer) were constructed to identify DFU, and the segmentation and classification results of the three models were compared.
    Results: Among the three models, Mask2former exhibited the best recognition performance, with a mean Intersection over Union of 65%, surpassing Deeplabv3's 62% and Swin-Transformer's 52%. The Intersection over Union value of Mask2former for wound recognition reached 85.9%, with IoU values of 80%, 78%, 62%, 61%, 47%, and 39% for granulation tissue, gangrene, bone tissue, necrotic tissue, tendons, and periwound erythema, respectively. In the wound classification task, the Mask2former model achieved an accuracy of 0.9185 and an Area Under the Curve of 0.9429 for the classification of Wagner grade 1-2, grade 3, and grade 4 wounds.
    Conclusion: Among the three deep learning models, the Mask2former model demonstrated the best overall performance. This method can effectively assist clinicians in recognizing DFU and segmenting the tissues within the wounds.
    Keywords:  artificial intelligence; classification; deep learning; diabetic foot ulcers; segmentation
    DOI:  https://doi.org/10.3389/fendo.2025.1543192
  10. J Imaging. 2025 Apr 21. pii: 123. [Epub ahead of print]11(4):
      Early detection of diabetic retinopathy is critical for preserving vision in diabetic patients. The classification of lesions in Retinal fundus images, particularly macular edema, is an essential diagnostic tool, yet it presents a significant learning curve for both novice and experienced ophthalmologists. To address this challenge, a novel Convolutional Deep Belief Network (CDBN) is proposed to classify image patches into three distinct categories: two types of macular edema-microhemorrhages and hard exudates-and a healthy category. The method leverages high-level feature extraction to mitigate issues arising from the high similarity of low-level features in noisy images. Additionally, a Real-Coded Genetic Algorithm optimizes the parameters of Gabor filters and the network, ensuring optimal feature extraction and classification performance. Experimental results demonstrate that the proposed CDBN outperforms comparative models, achieving an F1 score of 0.9258. These results indicate that the architecture effectively overcomes the challenges of lesion classification in retinal images, offering a robust tool for clinical application and paving the way for advanced clinical decision support systems in diabetic retinopathy management.
    Keywords:  automatic classification; convolutional deep belief network; genetic algorithm; hard exudates; macular edema; microhemorrhages
    DOI:  https://doi.org/10.3390/jimaging11040123
  11. Sci Rep. 2025 Apr 23. 15(1): 14215
      Diabetic Retinopathy (DR) is a leading cause of vision impairment globally, necessitating regular screenings to prevent its progression to severe stages. Manual diagnosis is labor-intensive and prone to inaccuracies, highlighting the need for automated, accurate detection methods. This study proposes a novel approach for early DR detection by integrating advanced machine learning techniques. The proposed system employs a three-phase methodology: initial image preprocessing, blood vessel segmentation using a Hopfield Neural Network (HNN), and feature extraction through an Attention Mechanism-based Capsule Network (AM-CapsuleNet). The features are optimized using a Taylor-based African Vulture Optimization Algorithm (AVOA) and classified using a Bilinear Convolutional Attention Network (BCAN). To enhance classification accuracy, the system introduces a hybrid Electric Fish Optimization Arithmetic Algorithm (EFAOA), which refines the exploration phase, ensuring rapid convergence. The model was evaluated on a balanced dataset from the APTOS 2019 Blindness Detection challenge, demonstrating superior performance in terms of accuracy and efficiency. The proposed system offers a robust solution for the early detection and classification of DR, potentially improving patient outcomes through timely and precise diagnosis.
    Keywords:  African vulture optimization; Attention mechanism based capsule network; Bilinear convolutional attention network; Diabetic retinopathy; Electric fish optimization; Hopfield neural network
    DOI:  https://doi.org/10.1038/s41598-025-99228-w
  12. Cureus. 2025 Mar;17(3): e80933
       INTRODUCTION: With its rising prevalence and serious complications, type 2 diabetes mellitus (T2DM) is a major worldwide health burden that calls for early detection using non-invasive screening techniques. Existing screening techniques, including OGTT, HbA1c, and fasting plasma glucose, have drawbacks in terms of accessibility, expense, and invasiveness. Recent developments in heart rate variability (HRV) analysis and machine learning (ML) offer a possible non-invasive substitute for diabetes screening. Previous research on HRV-based ML models in the classification of diabetes has issues with generalizability. The objective of this study is to develop and validate ML models using HRV features: time-domain, frequency-domain, and nonlinear HRV, to improve the prediction of T2DM. The study also evaluates the developed ML model's effectiveness against existing ML models.
    METHOD: A retrospective dataset comprising 519 individuals (261 T2DM patients and 258 non-diabetic controls) was collected from the Autonomic Function Testing (AFT) laboratory repositories. To ensure comparability of age, gender, height, and weight among groups, post-hoc matching was used. HRV features were extracted from five-minute ECG recordings using the PowerLab data acquisition system and LabChart HRV module (ADInstruments, Sydney, Australia), following the European Society of Cardiology Task Force guidelines. An 80:20 train-test split was used to train and assess ML models, such as Logistic Regression, K-Nearest Neighbors (KNNs), Random Forest, Gradient Boosting, XGBoost, LightGBM, CatBoost, and AdaBoost. Accuracy, precision, recall, F1-score, area under the curve (AUC) for the receiver operating characteristic (ROC), sensitivity, and specificity were among the performance indicators. GridSearchCV was used for hyperparameter adjustment to maximize model performance.
    RESULTS: The baseline characteristics of the non-diabetic and T2DM groups were similar (p>0.05). HRV analysis showed substantial decreases in the diabetic group's time-domain (SDNN - SD of Normal-to-Normal Intervals/RMSSD - RMS of Successive Differences), frequency-domain (Low/High Frequency - LF/HF), and nonlinear (SD2 - SD of Poincaré Plot/CVRR - Coefficient of Variation of R-R Intervals) parameters (p<0.001). With a 91.2% accuracy rate and an AUC of 0.91, CatBoost outperformed other ML models in terms of prediction. LightGBM and Random Forest, which demonstrated high sensitivity and specificity, trailed closely behind. KNN achieved the highest accuracy (98.2%) and AUC (0.99), followed by Random Forest (96.4%) and CatBoost (94.5%), while hyperparameter modification further enhanced performance. CatBoost demonstrated the highest predictive performance, with an accuracy of 91.2% and an AUC of 0.91. According to correlation analysis, the most important HRV characteristics for diabetes prediction were SD2, SDRR (SD of R-R Intervals), and CVRR.
    CONCLUSION: This study validates the utility of HRV-based ML models for non-invasive T2DM prediction, with ensemble models like CatBoost and LightGBM demonstrating superior performance when compared to the results of prior ML models. The optimized ML model, integrated with wearable medical technology for real-time monitoring, offers a scalable, affordable, and non-invasive alternative for diabetes screening. To improve generalizability and clinical use, future studies should investigate wearable-based HRV monitoring, multimodal AI models, and longitudinal validation.
    Keywords:  cardiac autonomic neuropathy; diabetes mellitus; heart rate variability; machine learning; non-invasive screening
    DOI:  https://doi.org/10.7759/cureus.80933
  13. Clin Neurol Neurosurg. 2025 Apr 17. pii: S0303-8467(25)00182-9. [Epub ahead of print]253 108899
       OBJECTIVE: Diabetes Insipidus (DI) is a common complication that occurs following transsphenoidal surgery for sellar lesions. DI is usually transient but can be permanent in select patients. Prior studies have described preoperative risk factors for developing postoperative DI. However, no predictive risk score has been created to risk stratify these patients.
    METHODS: A single-center retrospective review from 2017 - 2022 was performed, reviewing all patients who underwent transsphenoidal surgery for resection of a sellar lesion. Longterm DI was defined as a patient who met DI criteria for at least six months and required desmopressin therapy. Baseline patient, operative, and radiographic characteristics were obtained. A machine learning method (Risk-SLIM) was utilized to create a risk stratification score to identify patients at high risk for DI.
    RESULTS: In total, 252 patients were identified to have sellar lesions treated with transsphenoidal surgery. Of these, 27 (10.7 %) patients developed long-term DI and required desmopressin therapy. The DI after Transsphenoidal Surgery score (DITSS) was created with an area under the curve of 0.81 and a calibration error (CAL) error of 7.3 %. Predicative factors were tumor pathology, Tumor size, patient age, and endoscopic approach. The probability of developing DI requiring long-term desmopressin therapy ranged from < 1 % for a score of 0 and > 95 % for a score of 10 CONCLUSIONS: The DITSS model is a concise and accurate tool to assist in clinical decision-making for risk stratifying which patients undergoing transsphenoidal surgery for sellar lesions may go on to develop DI.
    Keywords:  Diabetes Insipidus; Predictive Model; Sellar Lesions; Transsphenoidal Surgery
    DOI:  https://doi.org/10.1016/j.clineuro.2025.108899