Statistical Analysis for Correlation and Agreement

Assessing correlation and agreement between two methods of measurement.

• 1. Using
  • 1.1. Python
  • 1.2. Matlab
  • 1.3. R
  • 1.4. Julia
• 2. Correlation
• 3. Agreement

1. Using

1.1. Python

python ./Python/Correlation_Agreement.py \
    --M_predict 0.125 0.95 0.55 0.60 0.78 0.46 0.88 0.50 0.93 0.35 0.975 0.725 0.285 0.166 0.666 0.888 0.233 \
    --M_GT 0.127 0.97 0.53 0.57 0.72 0.49 0.91 0.52 0.90 0.37 0.982 0.718 0.277 0.175 0.666 0.88 0.2333

1.2. Matlab

cd ./Matlab
plot_Pearson_Correlation_Bland_Altman_Agreement([0.125, 0.95, 0.55, 0.60, 0.78, 0.46, 0.88, 0.50, 0.93, 0.35, 0.975, 0.725, 0.285, 0.166, 0.666, 0.888, 0.233], [0.127, 0.97, 0.53, 0.57, 0.72, 0.49, 0.91, 0.52, 0.90, 0.37, 0.982, 0.718, 0.277, 0.175, 0.666, 0.88, 0.2333])

1.3. R

Rscript ./R/Correlation_Agreement.R

1.4. Julia

julia ./Julia/Correlation_Agreement.jl

2. Correlation

The correlation (Pearson Correlation) test can be used to statistically test the degree of correlation between the measured values of the same object through two measurement methods.

Measurement_predict and Measurement_GT have a high correlation (Pearson correlation coefficient of 0.9966).

3. Agreement

Agreement (Bland-Altman Agreement) test is a measure of the mean and variance of the measurement values of the same object by two measurement methods.

Assuming that the difference between Measurement_predict and Measurement_GT conforms to a normal distribution, the difference between the two measurement methods is distributed within a confidence interval of 0.00145 ± 0.04668 (mm) with 95% confidence. That is, it can be considered that the two measurement methods, Measurement_predict and Measurement_GT, have good Agreement.

Contributing

If you find any errors, or have any suggestions, please feel free to please feel free to pull requests.