Automatic detection of preclinical neurodegeneration

Klöppel, Stefan; Chu, Carlton; Tan, GC; Draganski, Bogdan; Johnson, Hans J.; Paulsen, Jane S.; Kienzle, Wolf; Tabrizi, Sarah J.; Ashburner, John; Frackowiak, R. S. J.

doi:10.1212/01.wnl.0000341768.28646.b6

Cited by 90 publications

(78 citation statements)

References 30 publications

(32 reference statements)

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…Crucially, when caudate volume was used to retrospectively predict those patients who were within 5 years of clinical diagnosis (Fig 5), this measure was seen to be a robust and individualized identifier of real‐life clinical diagnosis. These findings accord and extend existing work,9, 19, 20, 21 providing additional support for the use of caudate volume as a reliable estimate of disease proximity and making it a potentially useful biomarker.…”

Section: Discussionsupporting

confidence: 88%

See 1 more Smart Citation

Predicting clinical diagnosis in Huntington's disease: An imaging polymarker

Mason

Daws

Soreq

et al. 2018

Annals of Neurology

View full text Add to dashboard Cite

ObjectiveHuntington's disease (HD) gene carriers can be identified before clinical diagnosis; however, statistical models for predicting when overt motor symptoms will manifest are too imprecise to be useful at the level of the individual. Perfecting this prediction is integral to the search for disease modifying therapies. This study aimed to identify an imaging marker capable of reliably predicting real‐life clinical diagnosis in HD.MethodA multivariate machine learning approach was applied to resting‐state and structural magnetic resonance imaging scans from 19 premanifest HD gene carriers (preHD, 8 of whom developed clinical disease in the 5 years postscanning) and 21 healthy controls. A classification model was developed using cross‐group comparisons between preHD and controls, and within the preHD group in relation to “estimated” and “actual” proximity to disease onset. Imaging measures were modeled individually, and combined, and permutation modeling robustly tested classification accuracy.ResultsClassification performance for preHDs versus controls was greatest when all measures were combined. The resulting polymarker predicted converters with high accuracy, including those who were not expected to manifest in that time scale based on the currently adopted statistical models.InterpretationWe propose that a holistic multivariate machine learning treatment of brain abnormalities in the premanifest phase can be used to accurately identify those patients within 5 years of developing motor features of HD, with implications for prognostication and preclinical trials. Ann Neurol 2018;83:532–543

show abstract

Section: Discussionsupporting

confidence: 88%

“…Researchers have postulated that neuroimaging markers of structural, functional, and connectivity changes in the premanifest brain have a more predictable relationship with the onset of clinically diagnosable HD 9, 10, 11. Over the past 5 years, the number of different imaging techniques has rapidly increased.…”

mentioning

confidence: 99%

Predicting clinical diagnosis in Huntington's disease: An imaging polymarker

Mason

Daws

Soreq

et al. 2018

Annals of Neurology

View full text Add to dashboard Cite

show abstract

“…Subclinical changes precede the onset of overt clinical manifestations of HD [21]; PMCs perform significantly worse on a variety of cognitive measures: psychomotor speed, timing and movement sequencing, learning and memory, working memory, face and emotion recognition and executive functions [11,22]. Moreover, PMCs also have worse recognition memory and score more highly for irritability than controls do [23].…”

Section: Discussionmentioning

confidence: 98%

Are gait initiation parameters early markers of Huntington's disease in pre-manifest mutation carriers?

et al. 2011

View full text Add to dashboard Cite

“…These methods make use of the entire multivariate pattern present in a dataset in order to create models that allow predictions to be made about new data. This approach was recently demonstrated by Klöppel et al using segmented grey matter data (Klöppel et al, 2009), as well as white matter data (Klöppel et al, 2008) to discriminate pre-HD from controls.…”

Section: Introductionmentioning

confidence: 99%