ROC Analysis

 
   Lead Investigator:    Piea Peng Lee
   Institution      :    Singapore Management University
   E-Mail           :    lee.piea.peng@outlook.com
   Proposal ID      :    999



   Proposal Description:
   Main variables of interest: MMSE scores, dMMSE, clinical evaluation score for dementia CDD, and patients' characteristics.  My methods: Multi-objective optimization multi-class ROC analysis  Main outcomes of interest: (i) Determining optimal threshold values of MMSE and CDD (ii) Evaluate the diagnostic accuracy of MMSE and CDD (iii) Generate multi-biomarker ROC analysis and (iv) Find the optimal linear combination of MMSE and CDD that maximizes the partial area under the ROC curves.  Type of data to be used: National Alzheimer's Coordinating Center's national database on Alzheimer's Disease (AD).