Maxima of discretely sampled random fields, with an application to 'bubbles'

A smooth Gaussian random field with zero mean and unit variance is sampled on a discrete lattice, and we are interested in the exceedance probability or P-value of the maximum in a finite region. If the random field is smooth relative to the mesh size, then the P-value can be well approximated by results for the continuously sampled smooth random field (Adler, 1981; Worsley, 1995a; Taylor & Adler, 2003; Adler & Taylor, 2007). If the random field is not smooth, so that adjacent lattice values are nearly independent, then the usual Bonferroni bound is very accurate. The purpose of this paper is to bridge the gap between the two, and derive a simple, accurate upper bound for intermediate mesh sizes. The result uses a new improved Bonferroni-type bound based on discrete local maxima. We give an application to the 'bubbles' technique for detecting areas of the face used to discriminate fear from happiness. Copyright 2007, Oxford University Press.

[1]  R. Adler,et al.  Random Fields and Geometry , 2007 .

[2]  Keith J. Worsley,et al.  An improved theoretical P value for SPMs based on discrete local maxima , 2005, NeuroImage.

[3]  Philippe G Schyns,et al.  Accurate statistical tests for smooth classification images. , 2005, Journal of vision.

[4]  Akimichi Takemura,et al.  MATHEMATICAL ENGINEERING TECHNICAL REPORTS Validity of the expected Euler characteristic heuristic , 2003 .

[5]  P. Schyns,et al.  A mechanism for impaired fear recognition after amygdala damage , 2005, Nature.

[6]  Peter Tittmann,et al.  Bonferroni-Galambos Inequalities for Partition Lattices , 2004, Electron. J. Comb..

[7]  Thomas Lewiner,et al.  Applications of Forman's discrete Morse theory to topology visualization and mesh compression , 2004, IEEE Transactions on Visualization and Computer Graphics.

[8]  K. Worsley.  Detecting activation in fMRI data , 2003, Statistical methods in medical research.

[9]  Robert J. Adler,et al.  Euler characteristics for Gaussian fields on manifolds , 2003 .

[10]  P. Schyns,et al.  Show Me the Features! Understanding Recognition From the Use of Visual Information , 2002, Psychological science.

[11]  Torsten Hothorn,et al.  On the Exact Distribution of Maximally Selected Rank Statistics , 2002, Comput. Stat. Data Anal..

[12]  Frédéric Gosselin,et al.  Bubbles: a technique to reveal the use of information in recognition tasks , 2001, Vision Research.

[13]  Klaus Dohmen,et al.  Improved Bonferroni Inequalities via Union-Closed Set Systems , 2000, J. Comb. Theory, Ser. A.

[14]  Alan C. Evans,et al.  A General Statistical Analysis for fMRI Data , 2000, NeuroImage.

[15]  M. Hallett.  Human Brain Function , 1998, Trends in Neurosciences.

[16]  S. Sarkar.  Some probability inequalities for ordered $\rm MTP\sb 2$ random variables: a proof of the Simes conjecture , 1998 .

[17]  R. Forman.  Morse Theory for Cell Complexes , 1998 .

[18]  Daniel Q. Naiman,et al.  Abstract tubes, improved inclusion-exclusion identities and inequalities and importance sampling , 1997 .

[19]  Karl J. Friston,et al.  Human Brain Function , 1997 .

[20]  Bradley Efron,et al.  The length heuristic for simultaneous hypothesis tests , 1997 .

[21]  K. Worsley,et al.  Boundary corrections for the expected Euler characteristic of excursion sets of random fields, with an application to astrophysics , 1995, Advances in Applied Probability.

[22]  D. Siegmund,et al.  Testing for a Signal with Unknown Location and Scale in a Stationary Gaussian Random Field , 1995 .

[23]  K. Worsley.  Estimating the number of peaks in a random field using the Hadwiger characteristic of excursion sets, with applications to medical images , 1995 .

[24]  Fred M. Hoppe,et al.  Beyond Inclusion-and-Exclusion: Natural Identities for P(exactly t events) and P(at least t events) and Resulting Inequalities , 1993 .

[25]  D. Naiman,et al.  INCLUSION-EXCLUSION-BONFERRONI IDENTITIES AND INEQUALITIES FOR DISCRETE TUBE-LIKE PROBLEMS VIA EULER CHARACTERISTICS , 1992 .

[26]  R. Simes,et al.  An improved Bonferroni procedure for multiple tests of significance , 1986 .

[27]  Ioan Tomescu,et al.  Hypertrees and Bonferroni inequalities , 1986, J. Comb. Theory, Ser. B.

[28]  R. Adler,et al.  The Geometry of Random Fields , 1982 .

[29]  K. Worsley.  An improved Bonferroni inequality and applications , 1982 .

[30]  David J. Hunter.  An upper bound for the probability of a union , 1976, Journal of Applied Probability.

[31]  Seymour M. Kwerel.  Most Stringent Bounds on Aggregated Probabilities of Partially Specified Dependent Probability Systems , 1975 .

[32]  E. Kounias.  Bounds for the Probability of a Union, with Applications , 1968 .

[33]  Klaus Dohmen,et al.  Improved Inclusion-Exclusion Identities and Bonferroni Inequalities with Reliability Applications , 2002, SIAM J. Discret. Math..

[34]  J. Galambos,et al.  Bonferroni-type inequalities with applications , 1996 .

[35]  David A. Grable,et al.  Sharpened Bonferroni Inequalities , 1993, J. Comb. Theory, Ser. B.