Reinforcement Learning for Precision Oncology

Abstract

Precision oncology is grounded in the increasing understanding of genetic and molecular mechanisms that underly malignant disease and offer different treatment pathways for the individual patient. The growing complexity of medical data has led to the implementation of machine learning techniques that are vastly applied for risk assessment and outcome prediction using either supervised or unsupervised learning. Still largely overlooked is reinforcement learning (RL) that addresses sequential tasks by exploring the underlying dynamics of an environment and shaping it by taking actions in order to maximize cumulative rewards over time, thereby achieving optimal long-term outcomes. Recent breakthroughs in RL demonstrated remarkable results in gameplay and autonomous driving, often achieving human-like or even superhuman performance. While this type of machine learning holds the potential to become a helpful decision support tool, it comes with a set of distinctive challenges that need to be addressed to ensure applicability, validity and safety. In this review, we highlight recent advances of RL focusing on studies in oncology and point out current challenges and pitfalls that need to be accounted for in future studies in order to successfully develop RL-based decision support systems for precision oncology.

Organized Breast and Cervical Cancer Screening: Attendance and Determinants in Rural China

“…Our research showed that women who participated in organized two-cancer screening reported a high level of overall perceived usefulness of the screening program, and its association with the willingness to participate in screening again and to recommend screening to others was statistically significant. As the final step in organized two-cancer screening, overall perceived usefulness can help participating women gain peace of mind by dispelling doubts, anxiety, and worry about two-cancer, which would be a reinforcer for screening participation [ 38 , 39 ].…”

Translating Data Science Results into Precision Oncology Decisions: A Mini Review

“…The goal of the algorithm is to maximize the rewards. RL applications in oncology have been proposed, for instance, in the domain of detection of lung nodules [ 22 ], and while potentially a tool for supporting clinical decisions, a few challenges have been discussed in precision oncology [ 23 ], treatment [ 24 ], personalized dose optimization [ 25 ] and model-informed precision dosing [ 26 ].…”

Unsupervised and semi‐supervised learning: the next frontier in machine learning for plant systems biology

“…The use of graph-linked unified embedding (GLUE) introduced the generative learning strategy of the variational graph autoencoder (VGAE) algorithm to realize the unsupervised integration and regulatory inference of millions of single-cell multi-omics data, with high efficiency and accuracy (Cao & Gao, 2022). Still, a number of promising ML approaches, such as PU-learning (Li et al, 2022), contrastive learning (Chen et al, 2020) and reinforcement learning (Eckardt et al, 2021), which have succeeded in the areas of human genomics, computer vison and games, respectively, need further exploration in plant research. There is no doubt that the increasing application of advanced algorithms will further promote plant systems biology research.…”