Abstract:Tooth segmentation is a critical technology in the field of medical image segmentation, with applications ranging from orthodontic treatment to human body identification and dental pathology assessment. Despite the development of numerous tooth image segmentation models by researchers, a common shortcoming is the failure to account for the challenges of blurred tooth boundaries. Dental diagnostics require precise delineation of tooth boundaries. This paper introduces an innovative tooth segmentation network that integrates boundary information to address the issue of indistinct boundaries between teeth and adjacent tissues. This network's core is its boundary feature extraction module, which is designed to extract detailed boundary information from high-level features. Concurrently, the feature cross-fusion module merges detailed boundary and global semantic information in a synergistic way, allowing for stepwise layer transfer of feature information. This method results in precise tooth segmentation. In the most recent STS Data Challenge, our methodology was rigorously tested and received a commendable overall score of 0.91. When compared to other existing approaches, this score demonstrates our method's significant superiority in segmenting tooth boundaries.

What problem does this paper attempt to address?

The problem that this paper attempts to solve is the challenge of blurred boundaries in tooth image segmentation. Specifically, tooth image segmentation has important applications in the field of medical image processing, such as orthodontic treatment, human body recognition, and dental pathology assessment. However, the common problem in existing tooth image segmentation models is that they fail to fully consider the issue of blurred tooth boundaries, which makes it difficult to accurately divide the boundaries between teeth and adjacent tissues. To address this challenge, the paper proposes an innovative tooth segmentation network - the Boundary Feature Fusion Network (BFFNet). The core of this network lies in its Boundary Feature Extraction Module (BFEM), which aims to extract detailed boundary information from high - level features. Meanwhile, the Feature Cross - Fusion Module (FCFM) synergistically fuses the detailed boundary information with global semantic information to achieve the layer - by - layer transfer of feature information, thereby achieving accurate tooth segmentation results. By introducing these modules, BFFNet can more effectively solve the problem of blurred boundaries in tooth images, providing a more effective and accurate tool for oral medical image analysis. In addition, this method has achieved remarkable results in the recent STS data challenge, with an overall score of 0.91, demonstrating its superiority over other existing methods in segmenting tooth boundaries. ### Formula Summary - **Boundary Feature Extraction Module Formula**: \[ Fe_i=(Fi\otimes Wi)\oplus Fi\quad\text{(Formula 1)} \] \[ Wi = E-(\sigma(\text{Up}(Si + 1)))\quad\text{(Formula 2)} \] where $\text{Up}(\cdot)$ represents the up - sampling operation, $\sigma(\cdot)$ represents the Sigmoid function, $E$ is a matrix of all ones, $\otimes$ represents element - wise multiplication, and $\oplus$ represents element - wise addition. - **Feature Cross - Fusion Module Formula**: \[ Flocal_{j,j + 1}=\text{Con}\nu3(Fj,j + 1)\otimes Wlocal\oplus\text{Con}\nu3(Fj,j + 1)\quad\text{(Formula 3)} \] \[ Wlocal=\sigma(P-\text{Con}\nu_2(\delta(P-\text{Con}\nu_1(\text{Con}\nu_3(Fj,j + 1))))) \quad\text{(Formula 4)} \] \[ Fu_{j,j + 1}=\text{Con}\nu_3(\text{Concat}(Flocal_{j,j + 1},\text{Con}\nu_3(Fj,j + 1)))\quad\text{(Formula 5)} \] - **Loss Function**: \[ \text{Loss}=Lw_{IOU}+Lw_{BCE}\quad\text{(Formula 6)} \] \[ Ltotal = L(G,Fup_G)+L(G,Fs - up_3)+L(G,Fs - up_4)+L(G,Fs - up_5)\quad\text{(Formula 7)} \] - **Evaluation Metric Formula**: \[ \text{Dice}=\frac{2*|A\cap B|}{|A|+|B|}\quad\text{(Formula 8)} \] \[ \text{IOU}=\frac{|A\cap B|}{|A\cup B|}\quad\text{(Formula 9)} \] \[ H(d)=\min(|x1 - x2|+|y1 - y2|)\quad\text{

Boundary feature fusion network for tooth image segmentation

FDNet: Feature Decoupled Segmentation Network for Tooth CBCT Image

[Research on intelligent tooth segmentation method combining multiple seed region growth and boundary extension]

A novel difficult-to-segment samples focusing network for oral CBCT image segmentation

Multi-feature fusion 3D-CNN for tooth segmentation

Center-Sensitive and Boundary-Aware Tooth Instance Segmentation and Classification from Cone-Beam CT.

THISNet: Tooth Instance Segmentation on 3D Dental Models Via Highlighting Tooth Regions

Refined tooth and pulp segmentation using U-Net in CBCT image

Teeth And Root Canals Segmentation Using ZXYFormer With Uncertainty Guidance And Weight Transfer

ToothSegNet: Image Degradation meets Tooth Segmentation in CBCT Images

TSASNet: Tooth segmentation on dental panoramic X-ray images by Two-Stage Attention Segmentation Network

STSN-Net: Simultaneous Tooth Segmentation and Numbering Method in Crowded Environments with Deep Learning

A progressive framework for tooth and substructure segmentation from cone-beam CT images

Accurate and Automatic Tooth Image Segmentation Model with Deep Convolutional Neural Networks and Level Set Method

DeMambaNet: Deformable Convolution and Mamba Integration Network for High-Precision Segmentation of Ambiguously Defined Dental Radicular Boundaries

Refined Individual Tooth Segmentation from Cone Beam CT Images

Automatic Classification and Segmentation of Teeth on 3D Dental Model Using Hierarchical Deep Learning Networks

Two-Stream Graph Convolutional Network for Intra-Oral Scanner Image Segmentation

Boundary Guidance Hierarchical Network for Real-Time Tongue Segmentation

Dental panoramic X-ray image segmentation for multi-feature coordinate position learning