This work is published in Skeletal Radiology (https://doi.org/10.1007/s00256-024-04627-1).

The tool was developed based on the following dependencies:

1. NumPy
2. Scipy
3. Pandas
4. Lifelines
5. Statistics
6. Matplotlib
7. Sksurv
8. Sklearn
9. Jupyter Notebooks

All data files for this project are stored within the folder 'OAI_SBL_Analysis_Data'. Within this folder all information for patients regarding demographics, TKR, BML size, and SBL can be found.

Notice: You must update the path to this folder within each script before conducting Cox Proportional Hazards or Kaplan-Meier Analysis.

loc_data = "~/OAI_SBL_Analysis_Data/" --> path to data
loc_module = "~/OAI_Github_scripts/" --> path to analysis scripts

These scripts can be found under the 'Kaplan_Meier_Analysis' folder. By running these you will be able to generate Figures 2A and 2B from the paper.

Notice: When saving figures from Kaplan-Meier analysis, be sure to change the path to your suited location.

plt.savefig('~/KMF_Curves/kmf_BML.pdf', format='pdf', bbox_inches = 'tight')

These scripts can be found under the 'CPH_Analysis' folder. By running these you will be able to generate the other figures from the paper.

Notice: When saving figures from CPH analysis, be sure to change the path to your suited location.

plt.savefig('~AUC_Curves/AUC_ALL_Knees_SBL_BML_figure_all_SBL.pdf', format='pdf', bbox_inches = 'tight')

Each CPH Analysis script will produce a time-dependent AUC, Concordance Index, and Brier Score. Kaplan-Meier Analysis scripts will generate Kaplan Meier curves and the Log-Rank test associated with each curve. Generated figures are located in ~/AUC_Curves/ and ~/KMF_Curves/.