Deep Phrenology

The AFNI team is pleased to announce the latest annual contribution to the AFDI[sup]*[/sup] project. This year, we are combining the old and the new - phrenology together with deep learning, in what we call “Deep Phrenology”. Here, we believe for the first time, the well established science of phrenology is merged with the emerging computing trend of deep learning. In our implementation, we literally bring the phrenology mapping from the skull to deep within the brain.

The exciting application for this new synergy is included in our brand new “cranium and skull investigation” program, 3dCSI. This software uses both new and old technologies to determine brain “fingerprinting” of suspects. Syntax is show below:

3dCSI -enhance -enhance -ENHANCE -input DSET

Also find below the input data and the perhaps surprising results with just our first test! An example script is also provided.

Try it out for yourself, if you’ve got the gall![sup]1[/sup]


Franz Joseph Gall, On the Functions of the Brain and of Each of Its Parts: Critical review of some anatomico-physiological works; with an explanation of a new philosophy of the moral qualities and intellectual faculties, 1835.

Oiwi Parker Jones, Fidel Alfaro-Almagro, Saad Jbabd, “An empirical, 21st century evaluation of phrenology” bioRxiv 243089; doi:



We take the skull image, and using the extremely stable processes of analytic continuation and sheaf cohomology on complex (holomorphic) manifolds, the skull bump map can be projected to the cortical surface, and then correlated with the intrinsic Riemann curvature of the gyri and sulci. These phreno-cortical mappings are very revealing of individual subject propensities. Coincidentally, this breakthrough technique has demonstrated the outstanding character and intelligence of all the members of the AFDI development team. One of the leaders in this cutting edge technology is pictured here in his formal laboratory setting.