S Rashmi, S Srinath, S Murthy Prashanth, Deshmukh Seema
Dept. of Computer Science and Engineering, Sri Jayachamarajendra College of Engineering, JSS Science and Technology University, Mysuru, 570006, India.
Dept. of Pediatric & Preventive Dentistry, JSS Dental College & Hospital, JSS Academy of Higher Education & Research, Mysuru, 570015, India.
Dentomaxillofac Radiol. 2024 Feb 8;53(2):115-126. doi: 10.1093/dmfr/twad011.
The objectives of this study are to explore and evaluate the automation of anatomical landmark localization in cephalometric images using machine learning techniques, with a focus on feature extraction and combinations, contextual analysis, and model interpretability through Shapley Additive exPlanations (SHAP) values.
We conducted extensive experimentation on a private dataset of 300 lateral cephalograms to thoroughly study the annotation results obtained using pixel feature descriptors including raw pixel, gradient magnitude, gradient direction, and histogram-oriented gradient (HOG) values. The study includes evaluation and comparison of these feature descriptions calculated at different contexts namely local, pyramid, and global. The feature descriptor obtained using individual combinations is used to discern between landmark and nonlandmark pixels using classification method. Additionally, this study addresses the opacity of LGBM ensemble tree models across landmarks, introducing SHAP values to enhance interpretability.
The performance of feature combinations was assessed using metrics like mean radial error, standard deviation, success detection rate (SDR) (2 mm), and test time. Remarkably, among all the combinations explored, both the HOG and gradient direction operations demonstrated significant performance across all context combinations. At the contextual level, the global texture outperformed the others, although it came with the trade-off of increased test time. The HOG in the local context emerged as the top performer with an SDR of 75.84% compared to others.
The presented analysis enhances the understanding of the significance of different features and their combinations in the realm of landmark annotation but also paves the way for further exploration of landmark-specific feature combination methods, facilitated by explainability.
本研究的目的是利用机器学习技术探索和评估头影测量图像中解剖标志点定位的自动化，重点是特征提取与组合、上下文分析以及通过夏普利值（SHAP）进行模型可解释性分析。
我们在一个包含300张侧位头影图的私有数据集上进行了广泛实验，以深入研究使用像素特征描述符（包括原始像素、梯度幅值、梯度方向和方向梯度直方图（HOG）值）获得的标注结果。该研究包括对在不同上下文（即局部、金字塔和全局）下计算的这些特征描述进行评估和比较。使用个体组合获得的特征描述符通过分类方法来区分标志点像素和非标志点像素。此外，本研究解决了地标间LightGBM集成树模型的不透明性问题，引入SHAP值以增强可解释性。
使用平均径向误差、标准差、成功检测率（SDR）（2毫米）和测试时间等指标评估特征组合的性能。值得注意的是，在所有探索的组合中，HOG和梯度方向操作在所有上下文组合中均表现出显著性能。在上下文层面，全局纹理表现优于其他，尽管其代价是测试时间增加。局部上下文中的HOG表现最佳，与其他相比，SDR为75.84%。
所呈现的分析不仅增强了对不同特征及其组合在标志点标注领域重要性的理解，还为在可解释性的推动下进一步探索特定于标志点的特征组合方法铺平了道路。
