Deep Learning-based Non-rigid Image Registration for High-dose Rate Brachytherapy in Inter-fraction Cervical Cancer

Abstract

In this study, an inter-fraction organ deformation simulation framework for the locally advanced cervical cancer (LACC), which considers the anatomical flexibility, rigidity, and motion within an image deformation, was proposed. Data included 57 CT scans (7202 2D slices) of patients with LACC randomly divided into the train (n = 42) and test (n = 15) datasets. In addition to CT images and the corresponding RT structure (bladder, cervix, and rectum), the bone was segmented, and the coaches were eliminated. The correlated stochastic field was simulated using the same size as the target image (used for deformation) to produce the general random deformation. The deformation field was optimized to have a maximum amplitude in the rectum region, a moderate amplitude in the bladder region, and an amplitude as minimum as possible within bony structures. The DIRNet is a convolutional neural network that consists of convolutional regressors, spatial transformation, as well as resampling blocks. It was implemented by different parameters. Mean Dice indices of 0.89 ± 0.02, 0.96 ± 0.01, and 0.93 ± 0.02 were obtained for the cervix, bladder, and rectum (defined as at-risk organs), respectively. Furthermore, a mean average symmetric surface distance of 1.61 ± 0.46 mm for the cervix, 1.17 ± 0.15 mm for the bladder, and 1.06 ± 0.42 mm for the rectum were achieved. In addition, a mean Jaccard of 0.86 ± 0.04 for the cervix, 0.93 ± 0.01 for the bladder, and 0.88 ± 0.04 for the rectum were observed on the test dataset (15 subjects). Deep learning-based non-rigid image registration is, therefore, proposed for the high-dose-rate brachytherapy in inter-fraction cervical cancer since it outperformed conventional algorithms.

Deep Learning for Medical Image-Based Cancer Diagnosis

“…This registration pipeline starts with a classification network for coarse orientation estimation, followed by a segmentation network for predicting ground-truth planes in 3D volumes, enabling fully automated slice-to-volume registration in one shot. In [464], a method for the deformation simulation of inter-fraction in high-dose rate brachytherapy was proposed, which is applied to the deformable image registration (DIR) algorithm based on deep learning, which can directly realize the inter-fraction image alignment of HDR sessions for inter-fraction high dose rate brachytherapy in cervical cancer.…”

Automatic landmark correspondence detection in medical images with an application to deformable image registration

“…The source image is then warped by resampling from the mapped locations. Some of the potential applications of DIR in medical imaging are dose accumulation in radiation treatment, contour propagation, tumor growth tracking, and creating a digital atlas (Mohammadi et al, 2019;Rigaud et al, 2019;Zhao et al, 2022;Salehi et al, 2022).…”

“…With the advent of deep learning in the past few years, multiple deep learning based DIR approaches have been proposed (Balakrishnan et al, 2019;de Vos et al, 2017;Li and Fan, 2018;Li et al, 2022;Salehi et al, 2022;Rigaud et al, 2019), which provide the possibility to predict the DVF for an entire volumetric scan within seconds. However, to the best of our knowledge, there is no work done in the direction of MO DIR using deep learning.…”

Progressively refined deep joint registration segmentation (ProRSeg) of gastrointestinal organs at risk: Application to MRI and cone‐beam CT

“…One approach to robustly handle large deformations is to use regularization constraints such as rigidity penalty 27 and geometry matching constraints used for successfully aligning images exhibiting large anatomic deformations such as upper GI organs 27 and female reproductive organs such as the uterus and cervix 28 . A recent DLIR method by Han et al 29 .…”

“…[14][15][16]26 One approach to robustly handle large deformations is to use regularization constraints such as rigidity penalty 27 and geometry matching constraints used for successfully aligning images exhibiting large anatomic deformations such as upper GI organs 27 and female reproductive organs such as the uterus and cervix. 28 A recent DLIR method by Han et al 29 have also shown that using geometry constraints can benefit handling large anatomic differences inherent in organs like the small and large bowel when aligning CBCT images with CT images. Our proposed method also uses geometry matching losses during training to better regularize the registration and segmentation sub-networks with a key difference that such losses are used also as deep supervision losses to optimize the incremental deformations computed by the network.…”