Current status and future developments in predicting outcomes in radiation oncology

Abstract

Advancements in data-driven technologies and the inclusion of information-rich multi omics features have significantly improved the performance of outcomes modeling in radiation oncology. For this current trend to be sustainable, challenges related to robust data modeling such as small sample size, low size to feature ratio, noisy data, as well as issues related to algorithmic modeling such as complexity, uncertainty, and interpretability, need to be mitigated if not resolved. Emerging computational technologies and new paradigms such as federated learning, human-in-the-loop, quantum computing, and novel interpretability methods show great potential in overcoming these challenges and bridging the gap towards precision outcome modeling in radiotherapy. Examples of these promising technologies will be presented and their potential role in improving outcome modeling will be discussed.

Artificial intelligence (AI) and machine learning (ML) in precision oncology: a review on enhancing discoverability through multiomics integration

“… 24 Multiomics data along with advanced prognostication, 25–29 and outcome prediction 30,31 can offer possibilities for treatment selection or planning. After treatment has commenced, omics data can be leveraged to monitor response and provide decision-support 32–34 to therapy and adapt if needed. 35 Some examples of modeling designs include: (i) radiomics as predictors for molecular data, and (ii) omics or multiomics data as predictors for a specific clinical end point.…”

“…Information-rich multiomics data have been shown to improve model performance and help lower decision-making uncertainty. 32 Luo et al 31,79 have employed an interpretable Bayesian Network approach on multiomics data for predicting local control and radiation-induced complications in NSCLC. They have used the Markov blanket approach and pre-existing expert knowledge in feature selection and building a relational graph between the selected features and the outcome as shown in Figure 3 .…”

Designing combination therapies for cancer treatment: application of a mathematical framework combining CAR T-cell immunotherapy and targeted radionuclide therapy

“…In particular, mathematical models of tumor-immune system dynamics have been proposed to optimize and personalize immunotherapies ( 21 – 23 ) either on their own or in combination with chemotherapy ( 24 ). Additionally, mathematical modeling of radiation therapy using the linear-quadratic model ( 25 ) to optimize patient-specific treatment regimens has been investigated for decades ( 26 , 27 ). The proposed radiobiological models have been used to study both tumor and normal tissue radiation dose-response effects ( 28 – 30 ).…”

A clinical decision support system for AI-assisted decision-making in response-adaptive radiotherapy (ARCliDS)

“…Optimal decision-making in Knowledge Based Response-Adaptive Radiotherapy (KBR-ART) is a difficult task 1 . The difficulties arise from a slew of factors, such as, involvement of many variables, uncertainty in treatment response, and inter-patient heterogeneity 2 . In the absence of a quantitative framework, clinical decisions are primarily influenced by physician’s professional experiences, which may result in inter-physician variability.…”