MULTIFRACTAL DISCRIMINATION MODEL (MDM) OF HIGH-FREQUENCY PUPIL DIAMETER MEASUREMENTS FOR HUMAN-COMPUTER INTERACTION - Quantitative Medical Data Analysis Using Mathematical Tools and Statistical Techniques - page 333

CHAPTER 16
MULTIFRACTAL DISCRIMINATION MODEL (MDM) OF
HIGH-FREQUENCY PUPIL DIAMETER MEASUREMENTS
FOR HUMAN-COMPUTER INTERACTION
Bin Shi
The MathWorks, Inc.,
3 Apple Hill Dr., Natick, MA 01760, USA
E-mail: [e-mail]
Kevin P. Moloney, V. Kathlene Leonard, Julie Jacko,
Francois Sainfort and Brani Vidakovic
Health Systems Institute,
Wallace H. Coulter Department of Biomedical Engineering,
Georgia Institute of Technology,
313 Ferst Dr. NW, Atlanta, GA 30332-0535, USA
E-mails: f moloney, leonard, jacko, sainfort, brani g @hsi.gatech.edu
Multifractality present in high-frequency pupil diameter measurements,
usually connected with the irregular scaling behavior and self-similarity,
is modeled with statistical accuracy and discriminatory power. The Mul-
tifractal Discrimination Model (MDM) is proposed to determine ocular
pathology based on the pupillary response behavior (PRB) exhibited by
older adults with and without ocular disease during the performance
of a computer-based task. The MDM consists of two parts: (1) a dis-
criminatory summary of the multifractal spectrum and (2) a combined
k-nearest-neighbor classier. The multifractal spectrum is used to dis-
criminate the PRB from four groups of older adult users, diering in oc-
ular pathology. Spectral Mode, Broadness, and left Slope (the M.B.S.
summary), three measures characterizing the multifractal spectrum of
observations, are proposed as distinguishing features of PRB across the
groups. The combined k-nearest neighbor classier is shown to be a valid
classier for the accurate prediction of ocular pathology from the PRB
measurements.