Submanifold Sparse Convolutional Networks for Automated 3D Segmentation of Kidneys and Kidney Tumours in Computed Tomography

2025-11-06

AI Summary

To address the computational overhead–accuracy trade-off in 3D kidney and tumor segmentation from CT volumes, this paper proposes a two-stage framework integrating adaptive voxel sparsification with submanifold sparse convolution. The method preserves native high-resolution 3D input while reducing redundant computation via learnable voxel sparsification, and employs a lightweight submanifold sparse convolutional network for efficient hierarchical feature learning. Evaluated on the KiTS23 dataset, it achieves a combined kidney-and-lesion Dice score of 95.8% and a tumor-only Dice of 80.3%, matching state-of-the-art accuracy. Moreover, inference speed improves by 60% and GPU memory consumption decreases by 75%. To our knowledge, this is the first work to synergistically combine voxel sparsification and submanifold sparse convolutions for joint 3D organ–lesion segmentation in medical imaging—achieving clinical-grade accuracy while substantially enhancing deployment feasibility.

Technology Category

• Computer Vision: Segmentation
• Machine Learning: Learning with Manifolds
• Constraint Satisfaction and Optimization: Satisfiability Modulo Theories

Application Category

• Semantics and Knowledge: Scalable techniques for the creation, curation, publication, maintenance, and consumption of large, Web-based, structured, reusable, knowledge graphs and ontologies
• Graph Algorithms and Modeling for the Web: Representation, reconstruction, and subgraph or motif discovery in Web-related graphs
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking

Abstract

The accurate delineation of tumours in radiological images like Computed Tomography is a very specialised and time-consuming task, and currently a bottleneck preventing quantitative analyses to be performed routinely in the clinical setting. For this reason, developing methods for the automated segmentation of tumours in medical imaging is of the utmost importance and has driven significant efforts in recent years. However, challenges regarding the impracticality of 3D scans, given the large amount of voxels to be analysed, usually requires the downsampling of such images or using patches thereof when applying traditional convolutional neural networks. To overcome this problem, in this paper we propose a new methodology that uses, divided into two stages, voxel sparsification and submanifold sparse convolutional networks. This method allows segmentations to be performed with high-resolution inputs and a native 3D model architecture, obtaining state-of-the-art accuracies while significantly reducing the computational resources needed in terms of GPU memory and time. We studied the deployment of this methodology in the context of Computed Tomography images of renal cancer patients from the KiTS23 challenge, and our method achieved results competitive with the challenge winners, with Dice similarity coefficients of 95.8% for kidneys + masses, 85.7% for tumours + cysts, and 80.3% for tumours alone. Crucially, our method also offers significant computational improvements, achieving up to a 60% reduction in inference time and up to a 75% reduction in VRAM usage compared to an equivalent dense architecture, across both CPU and various GPU cards tested.

Problem

• Automated 3D kidney and tumor segmentation in CT scans
• Overcoming computational impracticality of high-resolution 3D medical imaging
• Reducing GPU memory and time requirements for tumor delineation

Innovation

• Uses voxel sparsification to reduce data density
• Employs submanifold sparse convolutional network architecture
• Enables high-resolution 3D segmentation with reduced resources