Medical follow-up optimization: A Monte-Carlo planning strategy

• DOI: 10.48550/arxiv.2401.03972
• Publication Date: 2024-01-01
• Authors: Benoîte de Saporta; Aymar Thierry d’A...; Régis Sabbadin; Alice Cleynen
• Keywords: Leverage (statistics); Personalized Medicine

Designing patient-specific follow-up strategy is a crucial step towards personalized medicine in cancer. Tools to help doctors deciding on treatment allocation together with next visit date, based patient preferences and medical observations, would be particularly beneficial. Such tools should realistic models of disease progress under the impact treatments, involve design (multi-)objective functions that optimize along patient's journey, include efficient resolution algorithms by taking history into account. We propose model cancer evolution Piecewise Deterministic Markov Process where patients alternate between remission relapse phases disease-specific tumor evolution. This controlled via online optimization long-term cost function accounting for side-effects, hospital visits burden quality life. Optimization noisy measurements blood markers at dates. leverage Partially-Observed Monte-Carlo Planning algorithm solve this continuous-time, continuous-state problem, advantage nearly-deterministic nature show approximate solution approach exact performs better than counterpart discrete model, while allowing more versatility model.