Auto-segmentation of important centers of growth in the pediatric skeleton to consider during radiation therapy based on deep learning.

BACKGROUND

Routinely delineating of important skeletal growth centers is imperative to mitigate radiation-induced growth abnormalities for pediatric cancer patients treated with radiotherapy. However, it is hindered by several practical problems, including difficult identification, time consumption, and inter-practitioner variability.

PURPOSE

The goal of this study was to construct and evaluate a novel Triplet-Attention U-Net (TAU-Net)-based auto-segmentation model for important skeletal growth centers in childhood cancer radiotherapy, concentrating on the accuracy and time efficiency.

METHODS

A total of 107 childhood cancer patients fulfilled the eligibility criteria were enrolled in the training cohort (N = 80) and test cohort (N = 27). The craniofacial growth plates, shoulder growth centers, and pelvic ossification centers, with a total of 19 structures in the three groups, were manually delineated by two experienced radiation oncologists on axial, coronal, and sagittal computed tomography images. Modified from U-Net, the proposed TAU-Net has one main branch and two bypass branches, receiving semantic information of three adjacent slices to predict the target structure. With supervised deep learning, the skeletal growth centers contouring of each group was generated by three different auto-segmentation models: U-Net, V-Net, and the proposed TAU-Net. Dice similarity coefficient (DSC) and Hausdorff distance 95% (HD95) were used to evaluate the accuracy of three auto-segmentation models. The time spent on performing manual tasks and manually correcting auto-contouring generated by TAU-Net was recorded. The paired t-test was used to compare the statistical differences in delineation quality and time efficiency.

RESULTS

Among the three groups, including craniofacial growth plates, shoulder growth centers, and pelvic ossification centers groups, TAU-Net had demonstrated highly acceptable performance (the average DSC = 0.77, 0.87, and 0.83 for each group; the average HD95 = 2.28, 2.07, and 2.86 mm for each group). In the overall evaluation of 19 regions of interest (ROIs) in the test cohort, TAU-Net had an overwhelming advantage over U-Net (63.2% ROIs in DSC and 31.6% ROIs in HD95, p = 0.001-0.042) and V-Net (94.7% ROIs in DSC and 36.8% ROIs in HD95, p = 0.001-0.040). With an average time of 52.2 min for manual delineation, the average time saved to adjust TAU-Net-generated contours was 37.6 min (p < 0.001), a 72% reduction.

CONCLUSIONS

Deep learning-based models have presented enormous potential for the auto-segmentation of important growth centers in pediatric skeleton, where the proposed TAU-Net outperformed the U-Net and V-Net in geometrical precision for the majority status.