Publications by authors named "Ahatov Ainur"
Article Synopsis
- A new fully automated deep learning framework has been developed to measure vertebral morphometry and Cobb angles from 3D CT scans of the spine and validated on an external dataset.
- The framework uses an ensemble of U-Nets for localizing and segmenting vertebrae, achieving accurate measurements of vertebral body heights, intervertebral disk heights, and spinal curvature with minimal errors compared to manual measurements.
- Results show a strong correlation between the automated and manual measurements, indicating that this framework can effectively handle external data and is efficient in terms of time and computational resources.
