ReCoGNet: Recurrent Context-Guided Network for 3D MRI Prostate Segmentation

2025-06-24

AI Summary

To address the limitations of 2D models—neglecting inter-slice continuity—and 3D models—requiring extensive annotated data—in T2-weighted MRI prostate gland segmentation, this paper proposes a spatiotemporal hybrid network. The method leverages a pretrained DeepLabV3 to extract slice-wise semantic features and incorporates ConvLSTM to explicitly model anatomical temporal dependencies across adjacent slices, treating MRI volumes as spatiotemporal sequences for the first time. This design jointly preserves strong 2D representational capacity and enforces 3D spatial consistency, significantly enhancing robustness under limited annotations and low-contrast conditions. Evaluated on the PROMISE12 benchmark, the approach achieves superior IoU and Dice scores compared to state-of-the-art 2D and 3D methods, demonstrating its efficacy and clinical deployability.

Technology Category

• Computer Vision: Segmentation
• Machine Learning: Multi-instance/Multi-view Learning
• Knowledge Representation and Reasoning: Geometric, Spatial, and Temporal Reasoning

Application Category

• Graph Algorithms and Modeling for the Web: Algorithms and analysis for heterogeneous, signed, attributed, multi-relational, temporal, higher-order, and annotated Web-related graphs
• Search and Retrieval-Augmented AI: Web query analysis, representation and understanding
• Semantics and Knowledge: Methods to enhance, augment, integrate or synergize semantic models such as knowledge graphs and LLMs

Abstract

Prostate gland segmentation from T2-weighted MRI is a critical yet challenging task in clinical prostate cancer assessment. While deep learning-based methods have significantly advanced automated segmentation, most conventional approaches-particularly 2D convolutional neural networks (CNNs)-fail to leverage inter-slice anatomical continuity, limiting their accuracy and robustness. Fully 3D models offer improved spatial coherence but require large amounts of annotated data, which is often impractical in clinical settings. To address these limitations, we propose a hybrid architecture that models MRI sequences as spatiotemporal data. Our method uses a deep, pretrained DeepLabV3 backbone to extract high-level semantic features from each MRI slice and a recurrent convolutional head, built with ConvLSTM layers, to integrate information across slices while preserving spatial structure. This combination enables context-aware segmentation with improved consistency, particularly in data-limited and noisy imaging conditions. We evaluate our method on the PROMISE12 benchmark under both clean and contrast-degraded test settings. Compared to state-of-the-art 2D and 3D segmentation models, our approach demonstrates superior performance in terms of precision, recall, Intersection over Union (IoU), and Dice Similarity Coefficient (DSC), highlighting its potential for robust clinical deployment.

Problem

• Segments prostate gland from T2-weighted MRI accurately
• Overcomes limitations of 2D CNNs and 3D models
• Improves robustness in data-limited, noisy clinical settings

Innovation

• Hybrid architecture models MRI as spatiotemporal data
• DeepLabV3 backbone extracts high-level semantic features
• ConvLSTM layers integrate inter-slice spatial information