Premium
IC‐P‐042: Classification of Alzheimer's subject based on PiB‐MR manifold learning
Author(s) -
Bourgeat Pierrick,
Salvado Olivier,
Raniga Parnesh,
Dore Vincent,
Zhou Luping,
Martins Ralph,
Macaulay Lance,
Masters Colin,
Ames David,
Ellis Kathryn,
Villemagne Victor,
Rowe Christopher,
Fripp Jurgen
Publication year - 2012
Publication title -
alzheimer's and dementia
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 6.713
H-Index - 118
eISSN - 1552-5279
pISSN - 1552-5260
DOI - 10.1016/j.jalz.2012.05.074
Subject(s) - artificial intelligence , atlas (anatomy) , pattern recognition (psychology) , population , support vector machine , segmentation , nonlinear dimensionality reduction , mathematics , computer science , dimensionality reduction , anatomy , medicine , environmental health
dance between algorithms was 81%-96% for conversion, a nd kappa values were 0.61-0.79. Combination of MRI and CSF measures in the same subjects did not yield improved prediction performance compared with MRI alone. Concordance between HCV measures and CSF Ab was 79%-85% for conversion, but kappa values were 0.12-0.29.Conclusions:HCV prediction performance was algorithm insensitive, although the level of concordance between algorithms needs further investigation. CSF measures did not provide additional prediction power. Patients identified using CSF Ab were only partially concordant with those identified using HCV.