Multi-source localization in MEG using simulated annealing: model order determination and parameter accuracy

Conference · Sat Jun 01 00:00:00 EDT 1996

OSTI ID:243440

Huang, M; Supek, S; Aine, C

Empirical neuromagnetic studies have reported that multiple brain regions are active at single instants in time as well as across time intervals of interest. Determining the number of active regions, however, required a systematic search across increasing model orders using reduced chi-square measure of goodness-of-fit and multiple starting points within each model order assumed. Simulated annealing was recently proposed for noiseless biomagnetic data as an effective global minimizer. A modified cost function was also proposed to effectively deal with an unknown number of dipoles for noiseless, multi-source biomagnetic data. Numerical simulation studies were conducted using simulated annealing to examine effects of a systematic increase in model order using both reduced chi-square as a cost function as well as a modified cost function, and effects of overmodeling on parameter estimation accuracy. Effects of different choices of weighting factors are also discussed. Simulated annealing was also applied to visually evoked neuromagnetic data and the effectiveness of both cost functions in determining the number of active regions was demonstrated.

• Research Organization: Los Alamos National Lab. (LANL), Los Alamos, NM (United States)
• Sponsoring Organization: USDOE, Washington, DC (United States)
• DOE Contract Number: W-7405-ENG-36
• OSTI ID: 243440
• Report Number(s): LA-UR-96-1719; CONF-9602101-2; ON: DE96011291
• Resource Relation: Conference: Biomagnetism conference, Santa Fe, NM (United States), Feb 1996; Other Information: PBD: [1996]
• Country of Publication: United States
• Language: English

Related Subjects

55 BIOLOGY AND MEDICINE
BASIC STUDIES
66 PHYSICS
NERVE CELLS
ELECTROPHYSIOLOGY
MAGNETIC DIPOLE MOMENTS
DISTRIBUTION FUNCTIONS
BIOLOGICAL LOCALIZATION
NUMERICAL ANALYSIS
BIOELECTRICITY
MAGNETIC FIELDS
BIOLOGICAL MODELS
MAGNETIC FIELD CONFIGURATIONS
WEIGHTING FUNCTIONS
STATISTICAL MECHANICS
NUCLEAR MODELS