<ul><li>Hematotoxicity, drug-induced damage to the blood-forming system, poses a challenge in clinical practice due to its variability. Mechanistic models struggle with irregular patient outcomes.</li><li>Hybrid models combining mechanistic and data-driven approaches were developed for personalized platelet count prediction during chemotherapy.</li><li>Data-driven methods show improved prediction accuracy with sufficient data, especially for high-risk patients with irregular platelet dynamics.</li><li>Hybrid and mechanistic models excel in scenarios with limited or sparse data, showcasing potential in enhancing clinical decision-making in personalized medicine.</li></ul>

Developing hybrid mechanistic and data-driven personalized prediction models for platelet dynamics

Discover more