Tumor aware recurrent inter-patient deformable image registration of computed tomography scans with lung cancer

TL;DR

A tumor-aware recurrent registration (TRACER) deep learning method that is suitable for inter-patient registration involving LC occurring in both fixed and moving images and applicable to voxel-based analysis methods is evaluated.

Abstract

Background: Voxel-based analysis (VBA) for population level radiotherapy (RT) outcomes modeling requires topology preserving inter-patient deformable image registration (DIR) that preserves tumors on moving images while avoiding unrealistic deformations due to tumors occurring on fixed images. Purpose: We developed a tumor-aware recurrent registration (TRACER) deep learning (DL) method and evaluated its suitability for VBA. Methods: TRACER consists of encoder layers implemented with stacked 3D convolutional long short term memory network (3D-CLSTM) followed by decoder and spatial transform layers to compute dense deformation vector field (DVF). Multiple CLSTM steps are used to compute a progressive sequence of deformations. Input conditioning was applied by including tumor segmentations with 3D image pairs as input channels. Bidirectional tumor rigidity, image similarity, and deformation smoothness losses were used to optimize the network in an unsupervised manner. TRACER and multiple DL methods were trained with 204 3D CT image pairs from patients with lung cancers (LC) and evaluated using (a) Dataset I (N = 308 pairs) with DL segmented LCs, (b) Dataset II (N = 765 pairs) with manually delineated LCs, and (c) Dataset III with 42 LC patients treated with RT. Results: TRACER accurately aligned normal tissues. It best preserved tumors, blackindicated by the smallest tumor volume difference of 0.24\%, 0.40\%, and 0.13 \% and mean square error in CT intensities of 0.005, 0.005, 0.004, computed between original and resampled moving image tumors, for Datasets I, II, and III, respectively. It resulted in the smallest planned RT tumor dose difference computed between original and resampled moving images of 0.01 Gy and 0.013 Gy when using a female and a male reference.

Figures (7)

• Figure 1: Rationale for tumor conditioned inter-patient registration. (a) Lack of tumor conditioning in PACS, ill-preserves tumors and produces physically unrealistic stretching of tissues. (b) Tumor conditioning in TRACER preserves tumor and normal tissue topology.
• Figure 2: (a) Schematic of tumor aware recurrent inter-patient DIR (TRACER) showing the fixed image and its mask ($\{x_{f}, y_{f}\}$), moving image and its mask ($\{x_{m}, y_{m}\}$ and the hidden layer $h$ input to the TRACER. Tumor conditioning is automated using published method. (b) shows TRACER architecture. (c) shows the losses used to optimize the network.
• Figure 3: DSC and HD95 accuracies of various methods on the testing datasets.
• Figure 4: Medial axis skeletons produced from PACS, TRACER, and manual segmentation of aorta, PA, IVC, and trachea on a sample patient.
• Figure 5: Registration for 2 different patient pairs containing tumor on completely different locations as well as different gender (Example $\#$2). Red and yellow contours indicate the tumor region on moving and fixed images.