Classification of the Cervical Vertebrae Maturation (CVM) stages Using the Tripod Network

TL;DR

This paper introduces TripodNet, a novel deep learning model with three parallel CNNs and directional filters, achieving state-of-the-art accuracy in classifying Cervical Vertebrae Maturation stages from X-ray images.

Contribution

The paper proposes TripodNet, a new deep learning architecture with directional filters and parallel networks, improving accuracy over existing models for CVM stage classification.

Findings

TripodNet achieves 81.18% accuracy in females.

TripodNet outperforms Swin Transformers.

Directional filters enhance edge detection in X-ray images.

Abstract

We present a novel deep learning method for fully automated detection and classification of the Cervical Vertebrae Maturation (CVM) stages. The deep convolutional neural network consists of three parallel networks (TriPodNet) independently trained with different initialization parameters. They also have a built-in set of novel directional filters that highlight the Cervical Verte edges in X-ray images. Outputs of the three parallel networks are combined using a fully connected layer. 1018 cephalometric radiographs were labeled, divided by gender, and classified according to the CVM stages. Resulting images, using different training techniques and patches, were used to train TripodNet together with a set of tunable directional edge enhancers. Data augmentation is implemented to avoid overfitting. TripodNet achieves the state-of-the-art accuracy of 81.18\% in female patients and 75.32\% in male patients. The proposed TripodNet achieves a higher accuracy in our dataset than the Swin Transformers and the previous network models that we investigated for CVM stage estimation.