ESTRO 2023 - Abstract Book

S1871

Digital Posters

ESTRO 2023

PO-2089 Toxicity modelling in RT using Bayesian network topology optimization with simulated annealing

K. Stenhouse 1 , P. McGeachy 2 , S. Spampinato 3 , K. Tanderup 3 , K. Martell 4 , S. Quirk 5 , K. Kirchheiner 6 , M. Schmid 6 , M. Roumeliotis 7 1 University of Calgary, Department of Physics and Astronomy, Calgary, Canada; 2 Tom Baker Cancer Centre, Department of Medical Physics, Calgary, Canada; 3 Aarhus University Hospital, Danish Center for Particle Therapy, Aarhus, Denmark; 4 Tom Baker Cancer Centre, Department of Radiation Oncology, Calgary, Canada; 5 Brigham and Women's Hospital, Department of Radiation Oncology, Boston, USA; 6 Medical University of Vienna, Department of Radiation Oncology - Comprehensive Cancer Centre, Vienna, Austria; 7 Johns Hopkins University, Department of Radiation Oncology and Molecular Radiation Sciences, Baltimore, USA Purpose or Objective To develop a simulated annealing optimization algorithm for building Bayesian Network (BN) topologies for late toxicity prediction in the EMBRACE I trial on locally advanced cervical cancer after radiochemotherapy. Materials and Methods EMBRACE morbidity outcomes were used to develop an algorithm for building two separate BN topologies for grade ≥ 2 cystitis and proctitis, as assessed by CTCAEv3. Baseline and at least one follow-up assessment were required. 40 clinical and treatment features were included as potential network nodes in the optimization. All features were initially included to ensure a direct comparison between the different model-building techniques. Restrictions on node connectivity were specified by experts to ensure that connections followed a logical sequential order (i.e. dosimetric features from treatment cannot be parent nodes of diagnostic features). Topology optimization followed a simulated annealing framework, with the cost function containing terms prioritizing model performance, mutual information between nodes, and simplicity. Results of the simulated annealing optimizations were compared to out-of-box optimization algorithms from the PyAgrum package (Tree-Augmented Naïve Bayes (TAN), Greedy Hill Climbing (GHC), Chow-Liu Optimization (CLO)). Results The proctitis model was built using 1108 patients and the cystitis model was built with 1073, excluding patients with bladder wall infiltration. Simulated annealing optimization for both toxicities terminated with convergence following 1000 unchanged iterations. Results for model performance and topology characteristics are outlined in Table 1 and networks are shown in Figure 1.