In this letter, we present a new 3D statistical method for surface detection which provides improvements over competitive methods both in terms of noise suppression and detection of complete surfaces. The methods are applied to both synthetically created image volumes, and MRI data. Accuracy against a ground truth is assessed using the quantitative figure of merit performance measure, with the statistical methods outperforming both a 3D implementation of the gradient Canny operator and a 3D optimal steerable filter method. The results also confirm how 3D surface detection methods avoid missing surface information by successfully locating complete boundaries irrespective of the object orientation and plane of image capture. We conclude that the statistical 3D methods are capable of producing more accurate surface maps in textured images, that reflect the 3D boundary information, improving on current 2D and 3D standards.