4,068 posters, 219 videos, 101 topics, 69 sessions, 14,365 authors, 6,128 institutions

Alzheimer's Association International Conference 2026

July 12-15, 2026 | London, United Kingdom, and Online

6500

Clinical Impact Of Computer Aided MRI Diagnosis For Patients Suspected Of Dementia

View Poster Contact Presenter

Poster Presenter

Stefan de Vries

Authors

Stefan de Vries, Hanneke Rhodius-Meester, Alle-Meije Wink, Lyduine Collij, Jyrki Lötjönen, Majon Muller, Wiesje van der Flier, Frederik Barkhof, Bas Jasperse, Mara ten Kate

Affiliations

Department of Radiology and Nuclear Medicine, Vrije Universiteit Amsterdam, Amsterdam UMC, location VUmc, Netherlands; Amsterdam Neuroscience, Brain Imaging, Netherlands, 3Geriatric Medicine, The Memory Clinic, Oslo University Hospital, Norway; Alzheimer Center Amsterdam, Neurology, Vrije Universiteit Amsterdam, Amsterdam UMC location VUmc, Netherlands; Amsterdam Neuroscience, Neurodegeneration, Netherlands,, Combinostics Ltd, Tampere, Finland, Department of Internal Medicine, Geriatric Medicine Section, Amsterdam UMC location Vrije Universiteit Amsterdam, The Netherlands, Amsterdam Neuroscience, Neurodegeneration, Netherlands; Epidemiology and Data Science, Vrije Universiteit Amsterdam, Amsterdam UMC location VUmc, Amsterdam, The Netherlands; Alzheimer Nederland, Amersfoort, the Netherlands; Amsterdam Public Health, Amsterdam, the Netherlands, Department of Radiology and Nuclear Medicine, Vrije Universiteit Amsterdam, Amsterdam UMC, location VUmc, Netherlands; Amsterdam Neuroscience, Brain Imaging, Netherlands; University College London (UCL), London, United Kingdom

Part of Topic

Background:

MRI supports dementia differential diagnosis, but characteristic patterns are often subtle and difficult to visually detect.

To assist radiologists, computer-aided diagnostic (CAD) systems are increasingly available.

Clinical impact of these systems remains underinvestigated.

Aim:

To assess the effect of CAD assistance on radiologist performance regarding:

1)Diagnostic accuracy in identifying dementia

2)Differential diagnosis

3)Diagnostic confidence

Methods:

Cases:

•Included from the Amsterdam Dementia Cohort (ADC) and Amsterdam Aging Cohort (AAC).

•80 cases spanning 6 diagnostic categories (SCD: 19, AD: 26, FTD: 10, VaD: 10, DLB: 10, and PSP: 5).

Readers:

•16 readers from 11 centers participated.

•12 neuroradiologists, 2 general radiologists, 2 final-year radiology residents.

•Median of 17 years of experience (range: 6-37 years).

Analysis:

Diagnostic Accuracy

•Outcome: reader performance, quantified with AUC per reader and condition.

•Differences in AUC tested using the DBM method for MRMC designs.

Differential Diagnosis

•Outcome: binary agreement with the multidisciplinary consensus diagnosis.

•Binomial linear mixed model (LMM) with random effects for reader and case and a fixed effect for CAD.

Diagnostic Confidence

•Outcome: Self-reported confidence on 5-point scale.

•Cumulative-linked LMM with random effects for reader and case and a fixed effect for CAD.

Results:

Diagnostic Accuracy

•AUC increased from 0.80 to 0.84 (∆ = 0.040; 95% CI 0.01-0.07; p = 0.01).

Differential Diagnosis

•CAD increased the odds of identifying the underlying etiology.

•Odds increased by 1.54 (95% CI 1.11-2.13; p = 0.01).

•Accuracy increase from 60.2% to 69.9% for the average case/reader pair.

Diagnostic Confidence

Confidence reporting

•Radiologists were more likely to report a higher confidence level when assisted.

•The odds increased by 1.68 (95% CI 1.43-1.97).

Calibration

•The relation between self-reported and objective performance was not altered by CAD assistance.

Take home message:

Computer-aided diagnostic assistance improved radiologists' identification of dementia in memory clinic patients. It helped identify the underlying etiology and increased diagnostic confidence.

Comments

Email is required but is not shown publicly.

Sort comments

Loading comments...