Machine learning (ML) is revolutionizing health sciences, specifically in medical imaging, drug discovery, and personalized care. This shift toward data-driven medicine utilizes computational models to derive clinical insights from complex health data. However, the domain faces significant hurdles, including strict privacy regulations, data heterogeneity, and extreme variance in dataset scales - from small clinical trials to massive omics repositories.To address these challenges, we propose a methodology centered on specialized data processing and encoding to ensure security and optimize data for varying sizes. By integrating quantum computing, our approach tackles complex medical datasets that remain intractable for classical ML. A case study on a heart disease dataset validates the proposed framework, demonstrating the efficacy of hybrid classical-quantum techniques in overcoming the limitations of modern data-driven medicine.
A Recipe for Data-driven Medicine: Data Encoding, Machine Learning, and Quantum Computing
Distefano, Salvatore
;Ziiatdinov, Mansur
2026-01-01
Abstract
Machine learning (ML) is revolutionizing health sciences, specifically in medical imaging, drug discovery, and personalized care. This shift toward data-driven medicine utilizes computational models to derive clinical insights from complex health data. However, the domain faces significant hurdles, including strict privacy regulations, data heterogeneity, and extreme variance in dataset scales - from small clinical trials to massive omics repositories.To address these challenges, we propose a methodology centered on specialized data processing and encoding to ensure security and optimize data for varying sizes. By integrating quantum computing, our approach tackles complex medical datasets that remain intractable for classical ML. A case study on a heart disease dataset validates the proposed framework, demonstrating the efficacy of hybrid classical-quantum techniques in overcoming the limitations of modern data-driven medicine.Pubblicazioni consigliate
I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.


