Limitations
Given the non-categorical and non-algorithmic nature of our probability
maps - direct comparison with other methods was not possible. Moreover,
it was not possible to quantify accuracy and traditional test
characteristics like specificity and sensitivity.