Visual Analysis of Tumor Control Models for Prediction of Radiotherapy Response

Abstract

In radiotherapy, tumors are irradiated with a high dose, while surrounding healthy tissues are spared. To quantify the probability that a tumor is effectively treated with a given dose, statistical models were built and employed in clinical research. These are called tumor control probability (TCP) models. Recently, TCP models started incorporating additional information from imaging modalities. In this way, patient-specific properties of tumor tissues are included, improving the radiobiological accuracy of models. Yet, the employed imaging modalities are subject to uncertainties with significant impact on the modeling outcome, while the models are sensitive to a number of parameter assumptions. Currently, uncertainty and parameter sensitivity are not incorporated in the analysis, due to time and resource constraints. To this end, we propose a visual tool that enables clinical researchers working on TCP modeling, to explore the information provided by their models, to discover new knowledge and to confirm or generate hypotheses within their data. Our approach incorporates the following four main components: (1) It supports the exploration of uncertainty and its effect on TCP models; (2) It facilitates parameter sensitivity analysis to common assumptions; (3) It enables the identification of inter-patient response variability; (4) It allows starting the analysis from the desired treatment outcome, to identify treatment strategies that achieve it. We conducted an evaluation with nine clinical researchers. All participants agreed that the proposed visual tool provides better understanding and new opportunities for the exploration and analysis of TCP modeling.

The State‐of‐the‐Art in Predictive Visual Analytics

“…This enables the user to find unusual flight behaviors and remove outliers from the training data. Raidou et al [RCMM*16] propose a visual analytics system for understanding Tumor Control Probability (TCP) models. This work supports the exploration of uncertainty and its effects on the model while facilitating parameter sensitivity with respect to user assumptions.…”

State‐of‐the‐Art Report: Visual Computing in Radiation Therapy Planning

“…(a) The Bladder Runner [RCA*18], proposed for the analysis of potential RT‐induced genito‐urinary toxicity risk, with respect to day‐to‐day exploration of bladder shape variations in a cohort of prostate cancer patients. (b) The approach proposed by Raidou et al [RCM*16] for the exploration and analysis of Tumor Control Probability (TCP) modeling in prostate cancer patients. (b1) shows the incorporation of ADC‐induced uncertainty and parameter sensitivity into the analysis and (b2) shows an inverse approach used for retrospective evaluation.…”

“…However, showing the uncertainties propagated from the respective imaging modality, or illustrating the sensitivity of the models to a number of parameter assumptions, has also become important. Raidou et al [RCM*16] propose a visual tool that enables clinical researchers to explore their TCP models, by supporting uncertainty and parameter sensitivity exploration, while enabling inter‐patient response variability analysis. This visual analysis tool has been later used in a clinical study, to quantify dose and TCP uncertainty bands when initial cell density is estimated from MRI‐based apparent diffusion coefficient maps of the patients [CRR*18].…”

“…Also, each step of the RT workflow involves heterogeneous sources of information. These might relate, for example, to multi‐modal registration [SFJ*16] and segmentation data [RMB*16], to ensemble data from the optimization phase of the dose planning [SRV16], or to modeling data from tumor control probability (TCP) [RCM*16] and normal tissue complication probability (NTCP) [RBGR18, RCA*18]. Understanding, exploring and analyzing all these data channels can be a demanding and time‐consuming task.…”

“…4D imaging and deformable registration can be used to model breathing motion [WSVHD08], but these introduce uncertainty [SFJ*16]. Uncertainties are also present in segmentation [RBGR18, RMB*16] and (radiobiological) modeling [RCA*18, RCM*16], to name a few. In particular, within radiobiological modeling, the careful quantitative reasoning about the probability of tumor control is unique for medical treatment.…”