Finding the optimal cutoff value for amyloid β positivity using the iterative outlier method and concordance rate

Abstract

Purpose: The purpose of this study was to calculate the cutoff value for amyloid β (Aβ) positron emission tomography (PET) positivity using the iterative outlier method and to evaluate its validity based on the concordance rate. Methods: We performed the iterative outlier method on 373 cognitively unimpaired (CU) subjects and calculated the optimal cutoff value for Aβ positivity. The validation was performed using the independent dataset, comprising 83 subjects (27 CU, 27 amnestic mild cognitive impairment, and 29 Alzheimer’s dementia). We evaluated the validity of the Aβ cutoff value by calculating its concordance rate with the visual assessment and between two different Aβ tracers performed in the same subject. Results: The concordance rate of Aβ cutoff values with the visual assessment ranged from 84.3\% to 91.5\%, depending on the reference regions. The concordance rate of the cutoff values between the two Aβ tracers ranged from 90.3\% to 97.5\%, all of which were higher than that of the visual assessment (86.7\%). Conclusion: We demonstrated that the iterative outlier method could identify the cutoff value for Aβ PET positivity.