Automatically computed rating scales from MRI for patients with cognitive disorders

Abstract

Objectives. To study whether visual MRI rating scales used in diagnostics of cognitive disorders can be estimated computationally and to compare the visual rating scales with their computed counterparts in differential diagnostics. Methods. A set of volumetry and voxel-based morphometry imaging biomarkers was extracted from T1-weighted and FLAIR images. A regression model was developed for estimating visual rating scale values from a combination of imaging biomarkers. We studied three visual rating scales: medial temporal lobe atrophy (MTA), global cortical atrophy (GCA) and white matter hyperintensities (WMHs) measured by the Fazekas scale. Images and visual ratings from the Amsterdam Dementia Cohort (ADC) (N=513) were used to develop the models and cross validate them. The PredictND (N=672) and ADNI (N=752) cohorts were used for independent validation to test generalizability. Results. The correlation coefficients between visual and computed rating scale values were 0.83/0.78 (MTA-Left), 0.83/0.79 (MTA-Right), 0.64/0.64 (GCA) and 0.76/0.75 (Fazekas) in ADC/PredictND cohorts. When performance in differential diagnostics was studied for the main types of dementia, the highest balanced accuracy, 0.75-0.86, was observed for separating different dementias from cognitively normal subjects using computed GCA. The lowest accuracy of about 0.5 for all the visual and computed scales was observed for the differentiation between Alzheimer's disease and frontotemporal lobar degeneration. Computed scales produced higher balanced accuracies than visual scales for MTA and GCA (statistically significant). 2 Conclusions. MTA, GCA and WMHs can be reliably estimated automatically helping to provide consistent imaging biomarkers for diagnosing cognitive disorders, even among less-experienced readers.

Medial temporal atrophy in preclinical dementia: visual and automated assessment during six year follow-up

“…All sets of ratings showed strong correlation to both HC and ILV volumes. This was reasonable, given that another recently proposed model estimating MTA was based on a linear combination of HC and ILV volumes (Koikkalainen et al, 2019). Our reported Spearman correlations between MTA and HC volume were stronger than previously reported, with r s ∈ [-0.26,-0.37] (Wahlund et al, 1999;Cavallin et al, 2012a).…”

Selection of memory clinic patients for CSF biomarker assessment can be restricted to a quarter of cases by using computerized decision support, without compromising diagnostic accuracy

“…We extracted five imaging markers using the cNeuro 1 cMRI quantification tool: 1) computed medial temporal lobe atrophy (cMTA), 2) computed global cortical atrophy (cGCA), 3) AD similarity scale, 4) Anterior Posterior index and 5) white matter hyperintensities (WMH). First, we automatically computed the MTA and GCA scores [24,25]. To do so, we defined volumes of brain structures from T1 image segmentations produced by a multiatlas segmentation algorithm [9,26].…”

Midlife Insulin Resistance as a Predictor for Late-Life Cognitive Function and Cerebrovascular Lesions

“…Three computer-based measures were used to evaluate cerebrovascular lesions: 1) total WMH volume normalized for age, gender, and total brain volume, 2) a computational counterpart for the Fazekas scale that was computed from the deep WMH volume using a regression-based model which has previously been defined elsewhere [41], and 3) total vascular burden measure that was calculated as the weighted combination of deep WMH volume, volume of cortical infarcts and volume of lacunar infarct volumes as described in detail in a previously published article [40].…”