Monte Carlo Study of Confidence Region Accuracy for MEG Inverse Dipole Solutions
Localization and characterization of neuromagnetic sources is often attempted assuming the model of a dipole current source in a spherical homogeneous conductor volume and solving the inverse problem for a given measured external field (Tanday (1987)). Evaluation of experimental results and clinical procedures also requires accurate confidence regions. Statistical methods exist for defining these regions by analyzing the fit of the data to the best fit model prediction without independent estimates of the measurement noise. We have empirically evaluated three of these methods in a Monte Carlo study.
KeywordsConfidence Region Monte Carlo Study Conducting Sphere Maximum Likelihood Criterion Dipole Parameter
Unable to display preview. Download preview PDF.
- Dennis, Jr, J.E. and Schnabel, Robert B. (1983). Numerical Methods for Unconstrained Optimization and Nonlinear Equations. Prentice-Hall Inc., Englewood Cliffs N.J.Google Scholar
- Donaldson, Janet R., (1985), Computation Experience with Confidence Regions. Thesis, Univ. of Colorado at Boulder.Google Scholar
- Halperin, M. (1963) “Confidence Interval Estimation in Nonlinear Regression”, Journal of the Royal Statistical Society (B), 25(2), pp 330–333.Google Scholar
- Hartley, H.O. (1964). “Exact Confidence Regions for the parameters in Nonlinear Regression Laws”, Biometrika, 51(3,4) pp. 347–353.Google Scholar
- Maksoudian, Y. Leon (1969). Probability and Statistics, with Applications. International Textbook Co., Scranton, PA. Pg. 188.Google Scholar
- Williams, E.J. (1962). “Exact Fiducial Limits in Nonlinear Estimation”, Journal of the Royal Statistical Society (B), 24(1), pp.125–139.Google Scholar