Regression Fitting & Plotting
Enter independent variable x and dependent variable y, and it automatically fits linear, quadratic, and cubic polynomial regressions, compares each model's R² and adjusted R², selects the best model by adjusted R², and plots it. Useful for analysing trends such as how ADC changes with gestational age.
① Input data
One data point per row, format: x (e.g. gestational age) y (e.g. ADC value), separated by space, Tab, or comma. You can paste two columns from Excel.
How to use & methodology
What is this tool good for?
It suits analysing how one continuous measure changes with another continuous variable, e.g. how fetal brain-region ADC changes with gestational age. The tool fits linear, quadratic, and cubic models simultaneously to help judge whether the trend is a straight line or a curve.
What is the difference between R² and adjusted R²?
R² (coefficient of determination) reflects how well the model fits the data — the closer to 1, the better; but raising the model's degree always increases R², which can cause overfitting. Adjusted R² penalises model complexity and is better for comparing models of different degrees, so this tool selects the best model by adjusted R².
Is the model with the highest R² always best?
Not necessarily. A high-order model's high R² may just be overfitting noise and need not be more biologically reasonable. Prefer a model that fits biological laws, has a good adjusted R², and is simple in form, rather than blindly chasing the highest R².
How do I prepare the data?
One data point per row, format 'x value + y value' (e.g. GA + ADC value), separated by space, Tab, or comma; you can paste two columns from Excel. At least 3 points are required, and at least 4 are recommended for fitting a cubic model.