Deep learning-based dose map prediction for high-dose-rate brachytherapy

Background. Creating a clinically acceptable plan in the time-sensitive clinic workflow of brachytherapy is challenging. Deep learning-based dose prediction techniques have been reported as promising solutions with high efficiency and accuracy. However, current dose prediction studies mainly target EBRT which are inappropriate for brachytherapy, the model designed specifically for brachytherapy has not yet well-established. Purpose. To predict dose distribution in brachytherapy using a novel Squeeze and Excitation Attention Net (SE_AN) model. Method. We hypothesized the tracks of Ir-192 inside applicators are essential for brachytherapy dose prediction. To emphasize the applicator contribution, a novel SE module was integrated into a Cascaded UNet to recalibrate informative features and suppress less useful ones. The Cascaded UNet consists of two stacked UNets, with the first designed to predict coarse dose distribution and the second added for fine-tuning 250 cases including all typical clinical applicators were studied, including vaginal, tandem and ovoid, multi-channel, and free needle applicators. The developed SE_AN was subsequently compared to the classic UNet and classic Cascaded UNet (without SE module) models. The model performance was evaluated by comparing the predicted dose against the clinically approved plans using mean absolute error (MAE) of DVH metrics, including D (2cc) and D (90%). Results. The MAEs of DVH metrics demonstrated that SE_AN accurately predicted the dose with 0.37 & PLUSMN; 0.25 difference for HRCTV D (90%), 0.23 & PLUSMN; 0.14 difference for bladder D (2cc), and 0.28 & PLUSMN; 0.20 difference for rectum D (2cc). In comparison studies, UNet achieved 0.34 & PLUSMN; 0.24 for HRCTV, 0.25 & PLUSMN; 0.20 for bladder, 0.25 & PLUSMN; 0.21 for rectum, and Cascaded UNet achieved 0.42 & PLUSMN; 0.31 for HRCTV, 0.24 & PLUSMN; 0.19 for bladder, 0.23 & PLUSMN; 0.19 for rectum. Conclusion. We successfully developed a method specifically for 3D brachytherapy dose prediction. Our model demonstrated comparable performance to clinical plans generated by experienced dosimetrists. The developed technique is expected to improve the standardization and quality control of brachytherapy treatment planning.