Abstract
The relationship between diabetes mellitus (DM) and microvascular brain pathology is complex and not very well debated, especially when referred to lacunar strokes, a major subset of cerebral small vessel disease (CSVD). Chronic hyperglycaemia, insulin resistance, and endothelial dysfunction drive microvascular injury, often compounded by coexisting hypertension and dyslipidaemia. Imaging and cohort studies confirm that DM accelerates CSVD progression and increases the risk of lacunar strokes, while antihypertensive and lipid-lowering therapies can modulate this risk. Nonetheless, a differential diagnosis of demyelinating lesions of the brain remains complicated, especially in young patients, as DM increases the likelihood of vascular or infectious mimics, but does not fundamentally influence diagnostic or pharmacological algorithms for acquired demyelinating syndromes. The differential diagnosis for Multiple Sclerosis is also essential in young patients with an autoimmune background. There is also growing evidence that artificial intelligence (AI) and computational modelling are increasingly useful when applied to cerebrovascular disease with DM as a key risk factor, by providing: risk stratification, automated neuroimaging analysis and decision-support systems, that integrate multimodal data to predict risks, classify aetiologies and forecast outcomes (along with retinal imaging and neuroimaging AI). AI-driven cerebrovascular morphology analysis, CSVD markers and perfusion imaging analyses are advancing the ability to predict short-term and longer-term outcomes, including cognitive impairment. For all that, challenges still arise throughout data standardisation, generalisability across diabetic subgroups, integration with electronic health records or regulatory considerations that need further prospective multicenter validation. Although pharmacological management in diabetic patients with lacunar stroke may not benefit from a diabetes-specific protocol beyond standard guidelines, physiotherapy stands as both a rehabilitative and preventive tool, contributing to microvascular risk reduction through exercise, improved metabolic control, and overall lifestyle modification. While calling for larger, diabetes-stratified studies to guide future interventions and refine personalised treatment strategies, the following narrative review, with emphasis on an AI and computational modelling in diabetes-related cerebrovascular disease, offers a brief insight into the current level of knowledge and recent updates in the correlations between DM and lacunar strokes.