Deep Implicit Statistical Shape Models for 3D Medical Image Delineation
This addresses the problem of accurate and robust anatomical structure segmentation in medical imaging for clinical deployment, representing an incremental advancement by integrating existing methods.
The paper tackles 3D medical image delineation by introducing deep implicit statistical shape models (DISSMs), which combine convolutional neural networks with statistical shape models to impose anatomical constraints, resulting in improved robustness and performance over leading FCN models, such as reducing mean Hausdorff distance by 7.7-14.3mm and improving worst-case Dice-Sorensen coefficient by 1.2-2.3% in intra-dataset experiments, and by 3.5-5.9% and 12.3-24.5mm in cross-dataset experiments.
3D delineation of anatomical structures is a cardinal goal in medical imaging analysis. Prior to deep learning, statistical shape models that imposed anatomical constraints and produced high quality surfaces were a core technology. Prior to deep learning, statistical shape models that imposed anatomical constraints and produced high quality surfaces were a core technology. Today fully-convolutional networks (FCNs), while dominant, do not offer these capabilities. We present deep implicit statistical shape models (DISSMs), a new approach to delineation that marries the representation power of convolutional neural networks (CNNs) with the robustness of SSMs. DISSMs use a deep implicit surface representation to produce a compact and descriptive shape latent space that permits statistical models of anatomical variance. To reliably fit anatomically plausible shapes to an image, we introduce a novel rigid and non-rigid pose estimation pipeline that is modelled as a Markov decision process(MDP). We outline a training regime that includes inverted episodic training and a deep realization of marginal space learning (MSL). Intra-dataset experiments on the task of pathological liver segmentation demonstrate that DISSMs can perform more robustly than three leading FCN models, including nnU-Net: reducing the mean Hausdorff distance (HD) by 7.7-14.3mm and improving the worst case Dice-Sorensen coefficient (DSC) by 1.2-2.3%. More critically, cross-dataset experiments on a dataset directly reflecting clinical deployment scenarios demonstrate that DISSMs improve the mean DSC and HD by 3.5-5.9% and 12.3-24.5mm, respectively, and the worst-case DSC by 5.4-7.3%. These improvements are over and above any benefits from representing delineations with high-quality surface.