Bladder Runner: Visual Analytics for the Exploration of RT‐Induced Bladder Toxicity in a Cohort Study

Abstract

We present the Bladder Runner, a novel tool to enable detailed visual exploration and analysis of the impact of bladder shape variation on the accuracy of dose delivery, during the course of prostate cancer radiotherapy (RT). Our tool enables the investigation of individual patients and cohorts through the entire treatment process, and it can give indications of RT-induced complications for the patient. In prostate cancer RT treatment, despite the design of an initial plan prior to dose administration, bladder toxicity remains very common. The main reason is that the dose is delivered in multiple fractions over a period of weeks, during which, the anatomical variation of the bladder -due to differences in urinary filling -causes deviations between planned and delivered doses. Clinical researchers want to correlate bladder shape variations to dose deviations and toxicity risk through cohort studies, to understand which specific bladder shape characteristics are more prone to side effects. This is currently done with Dose-Volume Histograms (DVHs), which provide limited, qualitative insight. The effect of bladder variation on dose delivery and the resulting toxicity cannot be currently examined with the DVHs. To address this need, we designed and implemented the Bladder Runner, which incorporates visualization strategies in a highly interactive environment with multiple linked views. Individual patients can be explored and analyzed through the entire treatment period, while inter-patient and temporal exploration, analysis and comparison are also supported. We demonstrate the applicability of our presented tool with a usage scenario, employing a dataset of 29 patients followed through the course of the treatment, across 13 time points. We conducted an evaluation with three clinical researchers working on the investigation of RT-induced bladder toxicity. All participants agreed that Bladder Runner provides better understanding and new opportunities for the exploration and analysis of the involved cohort data.

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

“…Exploration of the anatomical variability of OARs, together with analysis of potential toxicity risk has been researched by Raidou et al [RCA*18], within the Bladder Runner framework. The work uses the example of bladder toxicity in prostate cancer analysis, to present a novel tool for the detailed visual exploration and analysis of the impact of bladder shape variation on the accuracy of dose delivery.…”

“…(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.…”

“…Still, to compare patient cohorts, e.g., in retrospective or prospective Radiomics studies, where a multitude of information from all modalities should be analyzed and learned from, current options are limited. In this case, tools that allow a comprehensive inspection and exploration of all available patient data—going beyond single‐patient statistical analysis are required [RvD*15, RKS*16, RCA*18, RBGR18]. Therefore, ways to reduce the complexity of the data, while also being able to have a “digestible” overview of all of it, are essential.…”

“…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.…”