Abstract:
Diabetes mellitus is a complex endocrine and metabolic disorder involving impaired glucose regulation, altered insulin response, pancreatic β-cell dysfunction, adiposity-related metabolic load, vascular stress, renal vulnerability, and long-term complication risk. Conventional diabetes screening and prediction systems often classify patients using isolated laboratory or demographic variables, but they may provide limited explanation of how endocrine instability develops over time or how glucose–insulin imbalance relates to systemic disease progression. This work presents an endocrine-specific diabetes review and simulation framework implemented within the fully free SaMD (Software as a Medical Device) BVidAL Clinical Software Suite for interpretable, review-only clinical decision-support research.
The proposed software workflow integrates laboratory, clinical, and time-series glucose evidence into a structured endocrine model. Core variables include fasting glucose, random or postprandial glucose, HbA1c, fasting insulin, C-peptide, body mass index, blood pressure, lipid markers, renal markers, medication exposure, and continuous glucose monitoring parameters. These measurements are organized into pancreatic, metabolic, vascular, renal, and therapeutic review layers. Continuous glucose monitoring data are further analyzed through time-in-range, glycemic variability, hypoglycemia frequency, hyperglycemia burden, and glucose excursion patterns.
A simulation component is included to model glucose–insulin state behavior under different physiological and therapeutic contexts. The simulation layer supports review of glucose response trends, insulin-response sensitivity, carbohydrate-load influence, exercise-context effects, medication-associated modulation, and progression from stable glycemic control toward dysglycemia, prediabetes, or diabetes-risk states. The goal is not to replace clinical judgment or provide standalone diagnosis, but to generate transparent, reproducible, and interpretable endocrine-state summaries from public retrospective datasets and software-based simulation experiments.
Potential dataset evidence includes diabetes-risk tabular data, NHANES endocrine laboratory data, continuous glucose monitoring datasets, and hospital diabetes medication records. The BVidAL Medical Analyzer workflow supports dataset normalization, endocrine analyte mapping, glucose–insulin simulation, glycemic variability review, risk stratification, and structured report generation. Complication-oriented review can also connect diabetes-state patterns with renal risk, retinopathy context, neuropathy indicators, and diabetic foot-risk evidence where relevant datasets are available.
The clinical relevance of this approach is its ability to combine measurable diabetes variables with simulation-based endocrine interpretation. By linking laboratory markers, CGM behavior, medication context, and complication risk into a unified review workflow, the system may assist researchers and clinicians in understanding dominant contributors to dysglycemia, identifying unstable glucose–insulin patterns, and exploring individualized metabolic-risk trajectories. This study is presented as a proof-of-concept public-dataset and simulation-based demonstration for diabetes and endocrine decision-support research, without claims of regulatory approval, standalone diagnosis, or prospective clinical validation.
