Visual atrophy rating scales and amyloid PET status in an Alzheimer's disease clinical cohort

Abstract

Objectives Visual rating scales (VRS) are the quantification method closest to the approach used in routine clinical practice to assess brain atrophy. Previous studies have suggested that the medial temporal atrophy (MTA) rating scale is a reliable diagnostic marker for AD, equivalent to volumetric quantification, while others propose a higher diagnostic utility for the Posterior Atrophy (PA) scale in early‐onset AD. Methods Here, we reviewed 14 studies that assessed the diagnostic accuracy of PA and MTA, we explored the issue of cut‐off heterogeneity, and assessed 9 rating scales in a group of patients with biomarker‐confirmed diagnosis. A neuroradiologist blinded to all clinical information rated the MR images of 39 amyloid‐positive and 38 amyloid‐negative patients using 9 validated VRS assessing multiple brain regions. Automated volumetric analyses were performed on a subset of patients (n = 48) and on a group of cognitively normal individuals (n = 28). Results No single VRS could differentiate amyloid‐positive from amyloid‐negative patients with other neurodegenerative conditions. 44% of amyloid‐positive patients were deemed to have age‐appropriate levels of MTA. In the amyloid‐positive group, 18% had no abnormal MTA or PA scores. These findings were substantially affected by cut‐off selection. Amyloid‐positive and amyloid‐negative patients had comparable hippocampal and parietal volumes, and MTA but not PA scores correlated with the respective volumetric measures. Interpretation Consensus guidelines are needed before VRS can be recommended for use in the diagnostic workup of AD. Our data are suggestive of high intragroup variability and non‐superiority of volumetric quantification of atrophy over visual assessment.

Towards validation in clinical routine: a comparative analysis of visual MTA ratings versus the automated ratio between inferior lateral ventricle and hippocampal volumes in Alzheimer’s disease diagnosis

“…A continuous measurement would enable close monitoring of gradual decline, which prevents a situation where a patient aged 74, with an MTA score of 1.5 (pathological in this case) can have a normal MTA score in a follow-up examination the year after, solely due to the surpassing of the aforementioned (stringent) threshold, since subtle changes and trends are not depicted in a visual assessment. In addition, it needs to be noted that the existence and inconsistency in the use of different (age-specific and clinical population-based) cut-offs for defining visual MTA rating scale abnormalities is identified as a major issue and source of heterogeneity, with only a few studies addressing this concern so far [ 24 ]. Thus, preserving the advantage of automated volumetric imaging quantification, the ILV/Hip ratio score can be translated to a standardized value useful for interpretation purposes when compared to an age- and sex-matched normative population, while retaining a comparable performance to visual assessment.…”

“…Even though widely used in research settings, the integration of automated volumetry in routine clinical practice is still an ongoing evolving process [ 7 , 16 , 25 ]. As suggested by Vernooij et al [ 51 ], one of the main concerns hampering integration of automated software largely pertains to lack of standardization, validation, concerns about specificity, and the difficulty to transfer research findings into the clinical setting to help diagnosing an individual patient [ 24 ]. To overcome potential shortcomings of automated hippocampal volumetry alone for (early) AD diagnosis, we suggest the use of an automated MTA ratio, defined as the ratio between ILV and hippocampal volumes expressed as a percentage.…”

Radiomics Models Identifying Dementia in Patients with Normal Age-Related Hippocampal Atrophy: A study based on Machine Learning

“…In our study, we found 25.86% (30/116) of elderly CSVD patients with normal age-related HA had dementia, consistent with the results of previous studies that not everyone with a small hippocampus develops dementia [11,12], the result was also con rmed by a study from 10-year follow-up of hippocampal volume on MRI in 518 non-demented elderly in the Netherlands, they found that during the 6 years of follow-up, the absolute volume decline was 1.7% per year for the left hippocampus and 1.6% per year for the right hippocampus, but only 50 people developed incident dementia [12]. On the other hand, not everyone with dementia has a small hippocampus, because pathological studies suggested that part of AD patients have no atrophy over time [33,34], and 44% of AD patients with amyloid-positive were deemed to have age-appropriate levels of MTA [4]. Our research also con rms this point, because 3 dementia patients in our study were diagnosed AD clearly by the PETCT examinations.…”

“…No statistical differences were found among the three models. Given the highest AUC value in the T2-FLAIR model, converting the rad-score into binary T2-FLAIR (No, Yes) revealed a correlation between a 'Yes' response and dementia (OR = 13.21, 95% CI 3.487−59.922, p < 0.001) after adjusting for risk factors.Conclusion: Radiomic analysis of the hippocampus in T2-FLAIR images can effectively identify dementia in CSVD patients with normal age-related HA.Hippocampal atrophy (HA) has been associated with dementia in AD [4,5], PD [6,7], and post-stroke [8,9]. However, little is known about its association with CSVD, despite atrophy being recognized as a neuroimaging manifestation of CSVD [10].…”

“…Hippocampal atrophy (HA) has been associated with dementia in AD [4,5], PD [6,7], and post-stroke [8,9]. However, little is known about its association with CSVD, despite atrophy being recognized as a neuroimaging manifestation of CSVD [10].…”