Clustering 3D image voxels based on their location and value

My goal is to detect whether a MRI image contains an anomaly and the location of the anomaly. In my dataset I have MRI brain images which contain values of electrical conductivity of brain tissues.

Tissue segmentation image slice (left) and MRI conductivity image slice (right):

Anomaly has bigger conductivity as white matter, so I have to detect an anomaly based on that property. On tissue segmentation image, anomaly is presented with yellow color.

I have segmented white matter from an image:

All actions should be performed on this segmented white matter, which is a 3D array (74x84x77) of conductivity values.

I am a beginner in data science, so I am not familiar with many data science algorithms.

My question is: which algorithm would be best used to cluster together values (voxels) with higher conductivity. Clustering should take into account the location of regions and their size (because around the edges of white matter there are some measuring errors resulting in higher conductivity, which have to be ignored).

For reference, I have plotted all voxels that have conductivity larger than 110% of median white matter conductivity: