When CNNs Outperform Transformers and Mambas: Revisiting Deep Architectures for Dental Caries Segmentation

2025-11-18

AI Summary

Automatic segmentation of dental caries in panoramic radiographs remains challenging due to low image contrast, high morphological variability, and severe scarcity of annotated data. Method: This study systematically evaluates 12 state-of-the-art deep learning architectures—including CNNs, Transformers, and Mamba-based models—on the DC1000 dataset under a unified training framework to ensure fair comparison. Contribution/Results: Lightweight CNNs (e.g., DoubleU-Net) achieve superior performance, attaining Dice=0.7345, mIoU=0.5978, and precision=0.8145; the top three models are all CNNs. Crucially, this work empirically demonstrates that task-specific architectural suitability outweighs model complexity—a finding that challenges the prevailing trend of indiscriminately adopting large-scale models in medical image segmentation. It establishes a reproducible benchmark and provides methodological guidance for small-sample dental imaging analysis.

Technology Category

• Machine Learning: Deep Neural Architectures and Foundation Models
• Computer Vision: Segmentation
• Constraint Satisfaction and Optimization: Satisfiability Modulo Theories

Application Category

• Web Mining and Content Analysis: Large pretrained models with web data
• Graph Algorithms and Modeling for the Web: Graph neural networks and deep learning approaches for Web-related graphs
• Search and Retrieval-Augmented AI: Web evaluation methodologies and metrics

Abstract

Accurate identification and segmentation of dental caries in panoramic radiographs are critical for early diagnosis and effective treatment planning. Automated segmentation remains challenging due to low lesion contrast, morphological variability, and limited annotated data. In this study, we present the first comprehensive benchmarking of convolutional neural networks, vision transformers and state-space mamba architectures for automated dental caries segmentation on panoramic radiographs through a DC1000 dataset. Twelve state-of-the-art architectures, including VMUnet, MambaUNet, VMUNetv2, RMAMamba-S, TransNetR, PVTFormer, DoubleU-Net, and ResUNet++, were trained under identical configurations. Results reveal that, contrary to the growing trend toward complex attention based architectures, the CNN-based DoubleU-Net achieved the highest dice coefficient of 0.7345, mIoU of 0.5978, and precision of 0.8145, outperforming all transformer and Mamba variants. In the study, the top 3 results across all performance metrics were achieved by CNN-based architectures. Here, Mamba and transformer-based methods, despite their theoretical advantage in global context modeling, underperformed due to limited data and weaker spatial priors. These findings underscore the importance of architecture-task alignment in domain-specific medical image segmentation more than model complexity. Our code is available at: https://github.com/JunZengz/dental-caries-segmentation.

Problem

• Automated dental caries segmentation in panoramic radiographs
• Comparing CNN, Transformer and Mamba architectures for medical imaging
• Addressing challenges of low contrast and limited annotated data

Innovation

• CNN-based DoubleU-Net outperformed transformers and Mambas
• Comprehensive benchmarking of twelve architectures on DC1000 dataset
• Architecture-task alignment matters more than model complexity