Calculation of Femur Caput Collum Diaphyseal angle for X-Rays images using Semantic Segmentation

TL;DR

The study addresses automatic estimation of the femur CCD angle from X-ray images to reduce manual measurement variability and accelerate hip-fracture assessment. It introduces a segmentation-based approach using a UNet that outputs line heatmaps for neck and shaft centerlines on both sides, from which the CCD angle is derived via robust line fitting (RANSAC). The method achieves mean absolute errors of $MAE\approx 4.3^{\circ}$ on the left and $MAE\approx 4.9^{\circ}$ on the right, demonstrating practical accuracy for clinical use, complemented by a voice-controlled, touchless user interface suitable for the operating room. A user study reports strong usability (SUS 80–90%), while noting speech recognition biases with accents as a limitation and emphasizing the need for larger, more diverse datasets. Overall, the work presents a feasible, workflow-friendly tool for rapid CCD angle assessment with potential to improve diagnostic and surgical planning efficiency in hip pathologies.

Abstract

This paper investigates the use of deep learning approaches to estimate the femur caput-collum-diaphyseal (CCD) angle from X-ray images. The CCD angle is an important measurement in the diagnosis of hip problems, and correct prediction can help in the planning of surgical procedures. Manual measurement of this angle, on the other hand, can be time-intensive and vulnerable to inter-observer variability. In this paper, we present a deep-learning algorithm that can reliably estimate the femur CCD angle from X-ray images. To train and test the performance of our model, we employed an X-ray image dataset with associated femur CCD angle measurements. Furthermore, we built a prototype to display the resulting predictions and to allow the user to interact with the predictions. As this is happening in a sterile setting during surgery, we expanded our interface to the possibility of being used only by voice commands. Our results show that our deep learning model predicts the femur CCD angle on X-ray images with great accuracy, with a mean absolute error of 4.3 degrees on the left femur and 4.9 degrees on the right femur on the test dataset. Our results suggest that deep learning has the potential to give a more efficient and accurate technique for predicting the femur CCD angle, which might have substantial therapeutic implications for the diagnosis and management of hip problems.

Figures (5)

• Figure 1: U-shaped UNet architecture
• Figure 2: Labeled femur
• Figure 3: 1st column contains input X-Rays images with masks overlayed. 2nd column contains input mask that is fed to the UNet. 3rd Column contains masks predicted after the sigmoid cutoff applied at the output. 4th colums contains predicted lines after fitting the linear fit.
• Figure 4: Screenshot of user interface with a diagnostic CT scan of the hips opened
• Figure 5: System state: The system only executes commands after activation. Once activated the system using the key word activate, the system state will change to be listening and waits until it recognizes any commands. After execution of a command or after waiting for more than 5 seconds it automatically goes back to the default, not listening state.