The study , led by Rudolph, Jan, et al. was published in the Alzheimer’s & Dementia: Diagnosis, Assessment & Disease Monitoring Journal in December 2024. The study provides evidence for improved detection of AD and FTD when leveraging the module Brain Volumetry of the AI-supported tool mdbrain with automatic lobe segmentation and age- and sex-adjusted percentile comparisons that allow for faster and more precise assessments.
More accuracy, less errors
In a cohort of 55 patients (AD: 17, FTD: 18, healthy controls: 20), diagnostic accuracy increased notably with AI support of mdbrain . Two board-certified neuroradiologists (BCNR), two board-certified radiologists (BCR), and three radiology residents (RR) assessed the scans twice: first without AI support and then with AI assistance.
Diagnostic accuracy for AD improved significantly (area under the curve increased from 0.800 to 0.926, p < 0.05).
Correct diagnoses increased (p < 0.01) and errors were reduced (p < 0.03) for AD.
For FTD diagnosis, overall consensus (p < 0.01), BCNR (p < 0.02), and BCR (p < 0.05) recorded significantly more correct diagnoses.
Radiologists of all experience levels benefited, including BCNR.
Enhanced Efficiency with AI
Beyond accuracy gains, the AI tool mdbrain significantly reduced brain volumetry processing time to under 5 minutes. With full PACS integration, it facilitates real-time clinical decision-making, making AI-assisted volumetry a valuable addition to dementia diagnostics.
