An explainable framework for convolutional neural networks detecting dementia in MRI scans

Devesh Singh, Alice Grazia, Martin Dyrba, Stefan Teipel · 2025

Read An explainable framework for convolutional neural networks detecting dementia in MRI scans by Devesh Singh, Alice Grazia, Martin Dyrba, Stefan Teipel (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-11712630.txt.

Description

Doctors hesitate to trust AI diagnoses they cannot inspect. Devesh Singh and co-authors built a framework that explains a neural network's dementia calls on MRI scans. Their network learned from more than three thousand scans across cohorts. An attribution method then highlights the pixels that swayed each decision.

The framework cross-checks those highlights against anatomy. A segmentation tool measures each brain region's shrinkage against healthy norms. A knowledge model of brain structure organizes the findings. Software merges both streams into plain textual explanations with pictures, at whatever detail a clinician wants. Each part can be swapped as tools improve. Clinicians will test whether the explanations actually help.

Catalog record

CREATORDevesh Singh, Alice Grazia, Martin Dyrba, Stefan Teipel
YEAR2025
COLLECTIONPAPERS
FORMATTXT
FILE SIZE3 KB
VERSION1.0
LICENSECC BY 4.0
RIGHTSCC BY
SOURCEEurope PMC

Source: https://europepmc.org/articles/PMC11712630

Browse the full library · About the Tower of Babel