Variational autoencoder latent space as a robust and pragmatic clinical classification tool for dementia

William C Wakefield, Leland R Barnard, Hugo Botha, Jonathan Graff‐Radford, Kejal Kantarci, Brad F Boeve, Ronald C Petersen, Clifford R Jack, Val J Lowe, David T Jones · 2025

Read Variational autoencoder latent space as a robust and pragmatic clinical classification tool for dementia by William C Wakefield, Leland R Barnard, Hugo Botha, Jonathan Graff‐Radford, Kejal Kantarci, Brad F Boeve, Ronald C Petersen, Clifford R Jack, Val J Lowe, David T Jones (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/pmc-11716831.txt.

Description

Sorting dementia types from one brain scan would help clinics, and William Wakefield and co-authors tested a learning model for that task in work from twenty twenty-five. Their model squeezes each sugar-use brain scan into a short code, trained on two thousand scans with no labels. A simple sorting rule then reads the code.

A test set of more than a thousand labeled people covered healthy adults and three dementia types. Scores for telling groups apart were strong across the set. The team also mapped which scan traits drove each call, so the tool does not stay a black box. They pitch the method as a plain, sturdy aid that needs just one scan type.

Catalog record

CREATORWilliam C Wakefield, Leland R Barnard, Hugo Botha, Jonathan Graff‐Radford, Kejal Kantarci, Brad F Boeve, Ronald C Petersen, Clifford R Jack, Val J Lowe, David T Jones
YEAR2025
COLLECTIONPAPERS
FORMATTXT
FILE SIZE3 KB
VERSION1.0
LICENSECC BY 4.0
RIGHTSCC BY
SOURCEEurope PMC

Source: https://europepmc.org/articles/PMC11716831

Browse the full library · About the Tower of Babel