ESTRO 2022 - Abstract Book

S688

Abstract book

ESTRO 2022

Conclusion We propose a DL-based solution that provides accurate and fast segmentation of the urethra. Our results showed that DL takes advantage of DMs to predict invisible regions in the CT scans, such as the urethra. The inputs of the proposed framework are only segmentation masks, so the translation into the MR domain would be straightforward. This method can be potentially applied to the clinical workflow to devise personalized treatment with reduced urinary toxicity in prostate radiation therapy.

OC-0771 Uncertainty map for error prediction in deep learning-based head and neck tumor auto-segmentation

J. Ren 1 , J. Teuwen 2 , J. Nijkamp 3 , M. Rasmussen 1 , J. Eriksen 4 , J. Sonke 2 , S. Korreman 1

1 Aarhus University Hospital, Danish Center for Particle Therapy, Department of Oncology, Aarhus, Denmark; 2 Netherlands Cancer Institute, Department of Radiation Oncology, Amsterdam, The Netherlands; 3 Aarhus University Hospital, Danish Center for Particle Therapy, Aarhus, Denmark; 4 Aarhus University Hospital, Department of Experimental Clinical Oncology, Aarhus, Denmark Purpose or Objective Deep learning (DL) based auto-segmentation has shown to be performant in a variety of radiotherapy applications. Even though auto-segmentation of the gross tumor volume (GTV) is acceptable for a large group of patients, it still fails in a subgroup of patients. In this study, we investigate the use of uncertainty map to visualize potential uncertainties and to indicate patient-level segmentation failure. Materials and Methods We collected HNSCC patients (n=301) comprising Larynx, Pharynx, Oral, Sinonasal and Salivary gland carcinomas. Furthermore, treatment planning CT, PET, and MRI (T1w mDixon and T2w) images, as well as clinical delineations of the primary tumor (GTV-T) and nodal metastases (GTV-N) were also included. MRIs were deformable registered to PET/CT. The union of GTV-T and GTV-N were treated as ground truth (GTV-Clinic) for the DL prediction (GTV-DL). We trained a 3D UNet for 1000 epochs in a five-fold cross-validation fashion. At test time, for each patient, 50 stochastic samples were drawn from the UNet with Monte Carlo dropouts(p=0.1) from snapshot-saved models. The mean of all output softmax probability maps was used to aggregate GTV-DL and uncertainty map. The uncertainty map is a heatmap representing prediction uncertainties. We correlated the geometric location of the thresholded uncertainty map, the uncertainty regions (UR), with false predictions of the GTV-DL to locate potential predicted error regions (ER). We used the Dice similarity coefficient (Dice) to quantify the degree of overlap between UR and ER. In order to detect patient-level segmentation failure, we employed overlap metrics, False Omission Rate, False Negative Rate, and Surface Dice between the UR and GTV-DL to estimate GTV-DL performance in Dice. A Gradient Boosting Regressor was applied for the Dice estimation. We evaluated the regression result using Root Mean Squared Error (RMSE), Mean Absolute Error (MAE), and Coefficient of Determination (R2). Results The 3D UNet achieved reasonable and acceptable segmentation performance (Figure 1-A). Figure 1-B illustrates the functionality of uncertainty maps for voxel-wise uncertainty estimation with three examples. In the first two cases, UR can clearly indicate the location of errors; but, in the last case, UR does not correlate with ER. This specific example is restricted due to the ambiguity of GTV location by insufficient information from images. Figure 2 indicates that using an uncertainty threshold of �� =0.7, the regressor could estimate the actual segmentation Dice with an RMSE of 0.14, MAE of 0.09, and R2 of 0.4.