About Project
Application to identify the leg type (Left/Right), as well as detect the calibration sphere from the raw AP view X-ray images.
Client
Johnson & Johnson
Category
ML Backend Development
Key Highlights
- Trained a Machine Learning Model (Support Vector Machine) with features extracted from DICOM x-ray images to identify the leg type for the AP view of leg x-rays.
- Incorporated a feature to detect the metal calibration sphere in the x-rays.
LIBRARIES AND TOOLS
- Pydicom
- Pandas
- Numpy
- Matplotlib
- Seaborn
- Scikit-learn
- Scikit-image
- OpenCV
- DICOM X-ray files