PATIENT DIGITAL TWINS AS THE FOUNDATION OF FUTURE MEDICINE: PERSONALIZED SIMULATION, PREDICTIVE MODELING, AND DATA-DRIVEN CLINICAL DECISION-MAKING

Background: The growing availability of high-resolution imaging, biosensors, molecular profiling, and artificial intelligence has enabled the development of digital patient twins—computational models that reproduce individual physiological and pathological processes in silico. While digital twins have been widely proposed as tools for personalised medicine, their clinical and translational value across major disease domains has not yet been systematically synthesised. Methods: A narrative review was conducted of full-text publications from 2020–2025 addressing digital patient twins in cardiology, oncology, chronic disease management, and rehabilitation. The analysed literature included translational and clinical studies, mechanistic modelling papers, and healthcare system implementations. Evidence was prioritised from studies reporting patient-specific simulations, comparisons with real clinical or imaging data, and therapy-support scenarios. Results: In cardiology, electrophysiological and haemodynamic digital twins demonstrated high concordance with invasive mapping and imaging data and were associated with improved ablation planning, device optimisation, and reduced arrhythmia recurrence. In oncology, tumour digital twins integrating imaging and molecular data predicted tumour growth and treatment response with clinically meaningful accuracy, supporting personalised and adaptive cancer therapy. In chronic diseases, sensor-driven digital twins enabled early detection of physiological deterioration and supported proactive intervention, reducing exacerbations and hospitalisations. In rehabilitation, biomechanical and neurophysiological digital twins improved functional recovery by guiding personalised and robot-assisted therapy. Conclusions: Digital patient twins are transitioning from experimental computational tools to clinically relevant systems capable of influencing diagnosis, therapy selection, monitoring, and patient outcomes. By enabling in silico testing of therapeutic strategies on a virtual representation of the patient, digital twins reduce uncertainty in clinical decision-making and support truly personalised care. Continued progress in data integration, model validation, and regulatory governance will be essential for their safe and widespread adoption in clinical practice.