bims-aukdir Biomed News
on Automated knowledge discovery in diabetes research
Issue of 2025–06–29
ten papers selected by
Mott Given



  1. PLoS One. 2025 ;20(6): e0325956
      Hypoglycemia is a major challenge for people with diabetes. Therefore, glycemic monitoring is an important aspect of diabetes management. However, current methods such as finger pricking and continuous glucose monitoring systems (CGMS) are invasive, and hypoglycemia has still been shown to occur despite advancements in CGMS. Consequently, a growing body of research has been directed toward noninvasive hypoglycemia detection, relying on data from medical devices and wearables that can record physiological changes elicited by hypoglycemia. Consumer-grade wearables such as smartwatches remain an attractive yet underexplored candidate for such applications. Therefore, we explored the potential of a consumer-grade wearable for hypoglycemia prediction and investigated differing feature importance during waking and sleeping times. Smartwatch data from 18 adults with type 1 diabetes was collected, preprocessed, and imputed. Machine learning (ML) models were built using a tree-based ensemble algorithm to detect hypoglycemic events registered by CGMS. Models were built in a personalized manner using the same participant's data for training and testing, with separate modeling for daytime and nighttime. The relative importance of input features on model decisions was analyzed using SHAP (SHapley Additive exPlanations). Seventeen personalized models were built with an average area under the receiver operating characteristic curve (AUROC) score of 0.74 ± 0.08. Average specificity and sensitivity were 0.76 ± 0.18 and 0.71 ± 0.15, respectively. Time-of-day, activity, and cardiac features showed comparable importance in daytime models (29.9%, 28.5%, and 24%, respectively), while in nighttime models, cardiac features demonstrated the highest importance (42.2%) followed by time-of-day features (37.5%) and respiratory features (15.2%). In summary, we demonstrate the potential of consumer-grade wearables in noninvasive hypoglycemia detection. By additionally considering different physiological states (waking and sleeping) during modeling, our results offer further insights into differences in relative feature importance influencing the model's decision, guiding future research in this area.
    DOI:  https://doi.org/10.1371/journal.pone.0325956
  2. J Family Med Prim Care. 2025 May;14(5): 1871-1877
       Introduction: Diabetes mellitus (DM) is a metabolic disorder characterized by chronic hyperglycaemia either due to insulin resistance or due to relative or absolute insulin deficiency. Poorly controlled DM may result in both macrovascular and/or microvascular complications like diabetic retinopathy [DR]. Dilated eye examination is the most commonly employed method to diagnose DR. Nonmydriatic artificial intelligence [AI]-based technologies are the now available to screen DR.
    Methods: A cross-sectional observational study was conducted in urban field practice area of our medical college for 2 months duration. A total of 95 patients with type 2 DM were interviewed using predesigned, pretested semistructured schedule to collect data. Medical records were reviewed to collect relevant information. DR was screened using AI-based DR screening instrument, and venous blood sample was collected for glycated hemoglobin (HbA1C) testing. Data were analyzed using IBM SPSS [version 16]. Univariate and multivariate logistic regression tests were used, and P value ≤ 0.05 was taken as statistically significant.
    Results: The prevalence of DR was 17.9% in our study. Around 76.9% respondents had high fasting blood glucose [FBG: ≥126 mg/dl], and majority of the respondents [73.7%] had HbA1C value >7%. DR was significantly associated with FBG level, longer duration of diabetes, presence of hypertension, dyslipidemia, and kidney disease in univariate logistic regression and, in multivariable logistic regression, FBG level, presence of dyslipidemia and kidney disease retained their significance.
    Conclusion: This study had used AI-based DR screening instrument, to screen DR among T2DM patients. AI-based DR screening system can be encouraged in mass screening camps, especially in areas with inadequate number of ophthalmologists. This study also evaluated some important modifiable predictors of DR. Appropriate and early identification of such predictors may prevent DR-related blindness.
    Keywords:  Artificial intelligence; HbA1C; diabetes mellitus; diabetes retinopathy; fasting blood glucose
    DOI:  https://doi.org/10.4103/jfmpc.jfmpc_1693_24
  3. Biomedicines. 2025 Jun 12. pii: 1446. [Epub ahead of print]13(6):
      Background: Despite advances in artificial intelligence (AI) for Diabetic Retinopathy (DR) classification, traditional severity-based approaches often lack interpretability and fail to capture specific lesion-centered characteristics. To address these limitations, we constructed the National Medical Center (NMC) dataset, independently annotated by medical professionals with detailed labels of major DR lesions, including retinal hemorrhages, microaneurysms, and exudates. Methods: This study explores four critical research questions. First, we assess the analytical advantages of lesion-centered labeling compared to traditional severity-based labeling. Second, we investigate the potential complementarity between these labeling approaches through integration experiments. Third, we analyze how various model architectures and classification strategies perform under different labeling schemes. Finally, we evaluate decision-making differences between labeling methods using visualization techniques. We benchmarked the lesion-centered NMC dataset against the severity-based public Asia Pacific Tele-Ophthalmology Society (APTOS) dataset, conducting experiments with EfficientNet-a convolutional neural network architecture-and diverse classification strategies. Results: Our results demonstrate that binary classification effectively identifies severe non-proliferative Diabetic Retinopathy (Severe NPDR) exhibiting complex lesion patterns, while relationship-based learning enhances performance for underrepresented classes. Transfer learning from NMC to APTOS notably improved severity classification, achieving performance gains of 15.2% in mild cases and 66.3% in severe cases through feature fusion using Bidirectional Feature Pyramid Network (BiFPN) and Feature Pyramid Network (FPN). Visualization results confirmed that lesion-centered models focus more precisely on pathological features. Conclusions: Our findings highlight the benefits of integrating lesion-centered and severity-based information to enhance both accuracy and interpretability in DR classification. Future research directions include spatial lesion mapping and the development of clinically grounded learning methodologies.
    Keywords:  APTOS dataset; dataset integration; diabetic retinopathy; fine-grained lesion detection; knowledge transfer; lesion-centered labeling; medical expert labeling
    DOI:  https://doi.org/10.3390/biomedicines13061446
  4. JMIR Med Inform. 2025 Jun 27. 13 e66200
       Background: Building machine learning models that are interpretable, explainable, and fair is critical for their trustworthiness in clinical practice. Interpretability, which refers to how easily a human can comprehend the mechanism by which a model makes predictions, is often seen as a primary consideration when adopting a machine learning model in health care. However, interpretability alone does not necessarily guarantee explainability, which offers stakeholders insights into a model's predicted outputs. Moreover, many existing frameworks for model evaluation focus primarily on maximizing predictive accuracy, overlooking the broader need for interpretability, fairness, and explainability.
    Objective: This study proposes a 3-stage machine learning framework for responsible model development through model assessment, selection, and explanation. We demonstrate the application of this framework for predicting cardiovascular disease (CVD) outcomes, specifically myocardial infarction (MI) and stroke, among people with type 2 diabetes (T2D).
    Methods: We extracted participant data comprised of people with T2D from the ACCORD (Action to Control Cardiovascular Risk in Diabetes) dataset (N=9635), including demographic, clinical, and biomarker records. Then, we applied hold-out cross-validation to develop several interpretable machine learning models (linear, tree-based, and ensemble) to predict the risks of MI and stroke among patients with diabetes. Our 3-stage framework first assesses these models via predictive accuracy and fairness metrics. Then, in the model selection stage, we quantify the trade-off between accuracy and fairness using area under the curve (AUC) and Relative Parity of Performance Scores (RPPS), wherein RPPS measures the greatest deviation of all subpopulations compared with the population-wide AUC. Finally, we quantify the explainability of the chosen models using methods such as SHAP (Shapley Additive Explanations) and partial dependence plots to investigate the relationship between features and model outputs.
    Results: Our proposed framework demonstrates that the GLMnet model offers the best balance between predictive performance and fairness for both MI and stroke. For MI, GLMnet achieves the highest RPPS (0.979 for gender and 0.967 for race), indicating minimal performance disparities, while maintaining a high AUC of 0.705. For stroke, GLMnet has a relatively high AUC of 0.705 and the second-highest RPPS (0.961 for gender and 0.979 for race), suggesting it is effective across both subgroups. Our model explanation method further highlights that the history of CVD and age are the key predictors of MI, while HbA1c and systolic blood pressure significantly influence stroke classification.
    Conclusions: This study establishes a responsible framework for assessing, selecting, and explaining machine learning models, emphasizing accuracy-fairness trade-offs in predictive modeling. Key insights include: (1) simple models perform comparably to complex ensembles; (2) models with strong accuracy may harbor substantial differences in accuracy across demographic groups; and (3) explanation methods reveal the relationships between features and risk for MI and stroke. Our results underscore the need for holistic approaches that consider accuracy, fairness, and explainability in interpretable model design and selection, potentially enhancing health care technology adoption.
    Keywords:  MI; T2D; cardiology; cardiovascular; cardiovascular disease; clinical practice; diabetes; explainability; fairness; interpretable machine learning; machine learning; myocardial infarction; prediction; responsible framework; stroke; type 2 diabetes
    DOI:  https://doi.org/10.2196/66200
  5. PLoS One. 2025 ;20(6): e0324759
      Within the healthcare sector, the application of machine learning is gaining prominence, notably enhancing the efficiency and precision of diagnostic procedures. This study focuses on this key area of diabetes prediction and aims to develop an innovative prediction method. Using the data set published by Kare, this paper constructs and compares various intelligent systems based on multilayer algorithms, and specifically introduces improved reptile search algorithm (IRSA) to optimize the weight and threshold initialization of traditional backpropagation (BP) neural networks. This improvement aims to improve the network performance and accuracy in diabetes detection. In the study, the IRSA-BP hybrid algorithm and many other machine learning algorithms were used for diabetes prediction, and the algorithm performance was comprehensively evaluated using multiple classification metrics. The experimental results showed that the IRSA-BP algorithm performed the best among all the evaluated algorithms, with an accuracy of up to 83.6%, showing its superior performance in diabetes prediction. Therefore, the IRSA-BP classifier has an important potential for application in the medical field. It can assist medical professionals to identify diabetes risk earlier and assess the condition more accurately, thus improving diagnostic efficiency and accuracy. This is important for early intervention and treatment of patients with diabetes and to improve their health status and quality of life.
    DOI:  https://doi.org/10.1371/journal.pone.0324759
  6. Comput Biol Med. 2025 Jun 20. pii: S0010-4825(25)00967-9. [Epub ahead of print]195 110616
      This study aims to develop a robust and accurate model to forecast 30-day readmissions for patients with diabetes by leveraging machine learning techniques. Diabetes, being a chronic condition with complex care needs, often leads to frequent hospital readmissions. By predicting the likelihood of readmission within this critical timeframe, the study aims to empower healthcare providers to identify high-risk patients and implement targeted interventions proactively. The study utilized a dataset of 352 records from a diabetes specialty clinic in Varanasi, India. The study constructed prediction models utilizing Logistic Regression, Decision Tree, Random Forest, and XGBoost. The models were assessed using precision, recall, F1-score, and AUC-ROC. The results demonstrate that XGBoost attained the highest precision (0.84), recall (0.87), and F1-score (0.85), establishing it as the most effective model for predicting 30-day readmissions. Nevertheless, the Random Forest model demonstrated a superior AUC-ROC value of 0.94, indicating its ability to detect readmission cases accurately. The study's findings indicate that XGBoost demonstrates superior prediction accuracy, although Random Forest exhibits greater suitability for smaller datasets because of its strength against overfitting. These findings emphasize the significance of carefully choosing and optimizing machine learning models for healthcare applications. The study enhances patient care by allowing healthcare practitioners to identify high-risk patients and conduct focused interventions, potentially lessening the burden of readmissions.
    Keywords:  30-Day readmission; Decision tree; Diabetes; Logistic regression; Machine learning; Random forest; XGBoost
    DOI:  https://doi.org/10.1016/j.compbiomed.2025.110616
  7. Pharmaceutics. 2025 Jun 13. pii: 777. [Epub ahead of print]17(6):
      Diabetes is a global health challenge, and while current treatments offer relief, they often fall short in achieving optimal control and long-term outcomes. Nanotechnology offers a groundbreaking approach to diabetes management by leveraging materials at the nanoscale to improve drug delivery, glucose monitoring, and therapeutic precision. Early advancements focused on enhancing insulin delivery through smart nanosystems such as tiny capsules that gradually release insulin, helping prevent dangerous drops in blood sugar. Simultaneously, the development of nanosensors has revolutionised glucose monitoring, offering real-time, continuous data that empowers individuals to manage their condition more effectively. Beyond insulin delivery and monitoring, nanotechnology enables targeted drug delivery systems that allow therapeutic agents to reach specific tissues, boosting efficacy while minimising side effects. Tools like microneedles, carbon nanomaterials, and quantum dots have made treatment less invasive and more patient-friendly. The integration of artificial intelligence (AI) with nanotechnology marks a new frontier in personalised care. AI algorithms can analyse individual patient data to adjust insulin doses and predict glucose fluctuations, paving the way for more responsive, customised treatment plans. As these technologies advance, safety remains a key concern. Rigorous research is underway to ensure the biocompatibility and long-term safety of these novel materials. The future of diabetes care lies in the convergence of nanotechnology and AI, offering personalised, data-driven strategies that address the limitations of conventional approaches. This review explores current progress, persistent challenges, and the transformative potential of nanotechnology in reshaping diabetes diagnosis and treatment and improving patient quality of life.
    Keywords:  AI in diabetes management; carbon nanomaterials; continuous glucose monitoring (CGM); diabetes mellitus; diabetic wound healing; drug delivery systems; electrochemical biosensors; glucose biosensors; glucose monitoring; insulin delivery; microneedles; nanofibers; nanomedicine; nanotechnology; non-invasive monitoring; polymeric nanoparticles; quantum dots (QDs); smart nanocarriers; wearable biosensors
    DOI:  https://doi.org/10.3390/pharmaceutics17060777
  8. Ecotoxicol Environ Saf. 2025 Jun 23. pii: S0147-6513(25)00914-5. [Epub ahead of print]302 118569
       BACKGROUND: Diabetes Mellitus (DM) is a global health concern with rising prevalence, and its link to PFAS exposure remains unclear. No machine learning (ML) models have yet been developed to predict DM based on PFAS exposure.
    METHODS: We analyzed data from 10471 participants in National Health and Nutrition Examination Survey (NHANES, 2003-2018). Twelve ML models were compared, with LightGBM showing the best performance (AUC = 0.84, sensitivity = 0.83, accuracy = 73 %). Variable importance, Partial Dependence Analysis (PDA), SHapley Additive exPlanations (SHAP), and LOWESS smoothing were applied to assess predictor contributions and nonlinear effects. We developed a web-based calculator using Gradio to translate our findings into a clinical risk assessment tool.
    RESULTS: PFOA was identified as the strongest predictor and was negatively associated with DM risk. PFOS, PFNA, and MPAH showed positive associations, while PFDE had a slightly negative association. A PFOA threshold of 2.48 ng/ML was identified, below which DM risk was markedly reduced. At low PFOA levels, PFOS and PFNA exhibited mild synergistic effects, but these diminished at higher concentrations. SHAP analyses confirmed PFAS dominant protective contribution, and nonlinear patterns were observed for multiple PFAS. The deployed calculator provides clinicians with an accessible tool to assess individual DM risk based on patient profiles including PFAS exposure.
    CONCLUSION: This study provides novel ML-based insights into the associations between PFAS and DM. These findings warrant prospective validation and may inform environmental health strategies for diabetes prevention.
    Keywords:  Diabetes Mellitus; Environmental pollution; Interpretable machine learning; Partial Dependence Analysis; Per- and polyfluoroalkyl substances; SHapley Additive exPlanations
    DOI:  https://doi.org/10.1016/j.ecoenv.2025.118569
  9. Biomedicines. 2025 Jun 18. pii: 1494. [Epub ahead of print]13(6):
      Histopathological images represent a valuable data source for pathologists, who can provide clinicians with essential landmarks for complex pathologies. The development of sophisticated computational models for histopathological images has received significant attention in recent years, but most of them rely on free datasets. Materials and Methods: Motivated by this drawback, the authors created an original histopathological image dataset that resulted from an animal experimental model, acquiring images from normal female rats/rats with experimentally induced diabetes mellitus (DM)/rats who received an antidiabetic therapy with a synthetic compound (AD_SC). Images were acquired from vaginal, uterine, and ovarian samples from both MD and AD_DC specimens. The experiment received the approval of the Medical Ethics Committee of the "Gr. T. Popa" University of Medicine and Pharmacy, Iași, Romania (Approval No. 169/22.03.2022). The novelty of the study consists of the following aspects. The first is the use of a diabetes-induced animal model to evaluate the impact of an antidiabetic therapy with a synthetic compound in female rats, focusing on three distinct organs of the reproductive system (vagina, ovary, and uterus), to provide a more comprehensive understanding of how diabetes affects female reproductive health as a whole. The second comprises image classification with a custom-built convolutional neural network (CB-CNN), the extraction of textural features (contrast, entropy, energy, and homogeneity), and their classification with PyCaret Auto Machine Learning (AutoML). Results: Experimental findings indicate that uterine tissue, both for MD and AD_DC, can be diagnosed with an accuracy of 94.5% and 85.8%, respectively. The Linear Discriminant Analysis (LDA) classifier features indicate a high accuracy of 86.3% when supplied with features extracted from vaginal tissue. Conclusions: Our research underscores the efficacy of classifying with two AI algorithms, CNN and machine learning.
    Keywords:  PyCaret Auto Machine Learning; custom-built convolutional neural network; diabetes mellitus; histopathological image
    DOI:  https://doi.org/10.3390/biomedicines13061494
  10. Diabetol Metab Syndr. 2025 Jun 23. 17(1): 238
       BACKGROUND: Diabetes is a metabolic disease that can lead to severe cardiovascular diseases and neuropathy. The associated medical costs and complications make timely and effective management particularly important. Traditional diagnostic and management methods, like frequent glucose sampling and insulin injections, impose physical injuries on subjects. The development of artificial intelligence (AI) has opened new opportunities for diabetes management.
    METHODS: We conducted a meta-analysis integrating existing research, identifying a total of 1156 subjects to assess the effectiveness and safety of AI-based wearable devices, specifically closed-loop insulin delivery systems, in diabetes treatment.
    RESULTS: Compared to standard controls, AI-based closed-loop systems can analyze glucose data in real-time and automatically adjust insulin delivery, resulting in reduced time outside target glucose ranges (SMD = 0.90, 95% CI = 0.69 to 1.10, I2 = 58%, P < 0.001).
    CONCLUSION: AI-based closed-loop systems enhance the precision and convenience of diabetes treatment. This meta-analysis providing essential references for clinical treatment and policymaking in diabetes care.
    Keywords:  AI; Closed-loop systems; Diabetes; Insulin delivery; Wearable devices
    DOI:  https://doi.org/10.1186/s13098-025-01819-0