The Annals of Statistics

Multiple testing of local maxima for detection of peaks in random fields

Dan Cheng and Armin Schwartzman

Full-text: Access denied (no subscription detected)

We're sorry, but we are unable to provide you with the full text of this article because we are not able to identify you as a subscriber. If you have a personal subscription to this journal, then please login. If you are already logged in, then you may need to update your profile to register your subscription. Read more about accessing full-text


A topological multiple testing scheme is presented for detecting peaks in images under stationary ergodic Gaussian noise, where tests are performed at local maxima of the smoothed observed signals. The procedure generalizes the one-dimensional scheme of Schwartzman, Gavrilov and Adler [Ann. Statist. 39 (2011) 3290–3319] to Euclidean domains of arbitrary dimension. Two methods are developed according to two different ways of computing p-values: (i) using the exact distribution of the height of local maxima, available explicitly when the noise field is isotropic [Extremes 18 (2015) 213–240; Expected number and height distribution of critical points of smooth isotropic Gaussian random fields (2015) Preprint]; (ii) using an approximation to the overshoot distribution of local maxima above a pre-threshold, applicable when the exact distribution is unknown, such as when the stationary noise field is nonisotropic [Extremes 18 (2015) 213–240]. The algorithms, combined with the Benjamini–Hochberg procedure for thresholding p-values, provide asymptotic strong control of the False Discovery Rate (FDR) and power consistency, with specific rates, as the search space and signal strength get large. The optimal smoothing bandwidth and optimal pre-threshold are obtained to achieve maximum power. Simulations show that FDR levels are maintained in nonasymptotic conditions. The methods are illustrated in the analysis of functional magnetic resonance images of the brain.

Article information

Ann. Statist., Volume 45, Number 2 (2017), 529-556.

Received: May 2014
Revised: February 2016
First available in Project Euclid: 16 May 2017

Permanent link to this document

Digital Object Identifier

Mathematical Reviews number (MathSciNet)

Zentralblatt MATH identifier

Primary: 62H35: Image analysis
Secondary: 62H15: Hypothesis testing

False discovery rate Gaussian random field kernel smoothing image analysis overshoot distribution selective inference topological inference


Cheng, Dan; Schwartzman, Armin. Multiple testing of local maxima for detection of peaks in random fields. Ann. Statist. 45 (2017), no. 2, 529--556. doi:10.1214/16-AOS1458.

Export citation


  • [1] Adler, R. J., Subag, E. and Taylor, J. E. (2012). Rotation and scale space random fields and the Gaussian kinematic formula. Ann. Statist. 40 2910–2942.
  • [2] Adler, R. J. and Taylor, J. E. (2007). Random Fields and Geometry. Springer, New York.
  • [3] Adler, R. J., Taylor, J. E. and Worsley, K. J. (2010). Applications of random fields and geometry: Foundations and case studies. Preprint. Available at
  • [4] Azaïs, J.-M. and Wschebor, M. (2009). Level Sets and Extrema of Random Processes and Fields. Wiley, Hoboken, NJ.
  • [5] Benjamini, Y. and Heller, R. (2007). False discovery rates for spatial signals. J. Amer. Statist. Assoc. 102 1272–1281.
  • [6] Benjamini, Y. and Hochberg, Y. (1995). Controlling the false discovery rate: A practical and powerful approach to multiple testing. J. Roy. Statist. Soc. Ser. B 57 289–300.
  • [7] Brutti, P., Genovese, C. R., Miller, C. J., Nichol, R. C. and Wasserman, L. (2005). Spike hunting in galaxy spectra. Technical report, Libera Università Internazionale degli Studi Sociali Guido Carli di Roma. Available at
  • [8] Candes, E. and Tao, T. (2007). The Dantzig selector: Statistical estimation when $p$ is much larger than $n$. Ann. Statist. 35 2313–2351.
  • [9] Cheng, D. and Schwartzman, A. (2016). Supplement to “Multiple testing of local maxima for detection of peaks in random fields.” DOI:10.1214/16-AOS1458SUPP.
  • [10] Cheng, D. and Schwartzman, A. (2015). Distribution of the height of local maxima of Gaussian random fields. Extremes 18 213–240.
  • [11] Cheng, D. and Schwartzman, A. (2015). Expected number and height distribution of critical points of smooth isotropic Gaussian random fields. Preprint. Available at arXiv:1511.06835.
  • [12] Cheng, D. and Schwartzman, A. (2015). On the explicit height distribution and expected number of local maxima of isotropic Gaussian random fields. Preprint. Available at arXiv:1503.01328.
  • [13] Chumbley, J. R. and Friston, K. J. (2009). False discovery rate revisited: FDR and topological inference using Gaussian random fields. Neuroimage 44 62–70.
  • [14] Chumbley, J. R., Worsley, K., Flandin, G. and Friston, K. J. (2010). Topological FDR for neuroimaging. Neuroimage 49 3057–3064.
  • [15] Cramér, H. and Leadbetter, M. R. (1967). Stationary and Related Stochastic Processes. Sample Function Properties and Their Applications. Wiley, New York.
  • [16] Egner, A., Geisler, C., von Middendorff, C., Bock, H., Wenzel, D., Medda, R., Andresen, M., Stiel, A. C., Jakobs, S., Eggeling, C., Schönle, A. and Hell, S. W. (2007). Fluorescence nanoscopy in whole cells by asynchronous localization of photoswitching emitters. Biophys. J. 93 3285–3290.
  • [17] Fithian, W., Sun, D. and Taylor, J. (2015). Optimal inference after model selection. Preprint. Available at arXiv:1410.2597.
  • [18] Fyodorov, Y. V. (2004). Complexity of random energy landscapes, glass transition, and absolute value of the spectral determinant of random matrices. Phys. Rev. Lett. 92 240601, 4.
  • [19] Geisler, C., Schönle, A., von Middendorff, C., Bock, H., Eggeling, C., Egner, A. and Hell, S. W. (2007). Resolution of $\lambda/10$ in fluorescence microscopy using fast single molecule photo-switching. Applied Physics A 88 223–226.
  • [20] Genovese, C. R., Lazar, N. A. and Nichols, T. E. (2002). Thresholding of statistical maps in functional neuroimaging using the false discovery rate. Neuroimage 15 870–878.
  • [21] Heller, R., Stanley, D., Yekutieli, D., Rubin, N. and Benjamini, Y. (2006). Cluster-based analysis of FMRI data. Neuroimage 33 599–608.
  • [22] Lee, J. D., Sun, D. L., Sun, Y. and Taylor, J. E. (2015). Exact post-selection inference, with application to the lasso. Preprint. Available at arXiv:1311.6238.
  • [23] Moran, J. M., Jolly, E. and Mitchell, J. P. (2012). Social-cognitive deficits in normal aging. J. Neurosci. 32 5553–5561.
  • [24] Nichols, T. and Hayasaka, S. (2003). Controlling the familywise error rate in functional neuroimaging: A comparative review. Stat. Methods Med. Res. 12 419–446.
  • [25] Perone Pacifico, M., Genovese, C., Verdinelli, I. and Wasserman, L. (2004). False discovery control for random fields. J. Amer. Statist. Assoc. 99 1002–1014.
  • [26] Perone Pacifico, M., Genovese, C., Verdinelli, I. and Wasserman, L. (2007). Scan clustering: A false discovery approach. J. Multivariate Anal. 98 1441–1469.
  • [27] Piterbarg, V. I. (1996). Asymptotic Methods in the Theory of Gaussian Processes and Fields. Translations of Mathematical Monographs 148. Amer. Math. Soc., Providence, RI.
  • [28] Poline, J.-B., Worsley, K. J., Evans, A. C. and Friston, K. J. (1997). Combining spatial extent and peak intensity to test for activations in functional imaging. Neuroimage 5 83–96.
  • [29] Pratt, W. K. (1991). Digital Image Processing. Wiley, New York.
  • [30] Schwartzman, A., Gavrilov, Y. and Adler, R. J. (2011). Multiple testing of local maxima for detection of peaks in 1D. Ann. Statist. 39 3290–3319.
  • [31] Siegmund, D. O., Zhang, N. R. and Yakir, B. (2011). False discovery rate for scanning statistics. Biometrika 98 979–985.
  • [32] Simon, M. (1995). Digital Communication Techniques: Signal Design and Detection. Prentice Hall, Englewood Cliffs, NJ.
  • [33] Smith, S. M. and Nichols, T. E. (2009). Threshold-free cluster enhancement: Addressing problems of smoothing, threshold dependence and localisation in cluster inference. Neuroimage 44 83–98.
  • [34] Subag, E. (2015). The complexity of spherical p-spin models—a second moment approach. Preprint. Available at arXiv:1504.02251.
  • [35] Sun, W., Reich, B. J., Cai, T. T., Guindani, M. and Schwartzman, A. (2015). False discovery control in large-scale spatial multiple testing. J. R. Stat. Soc. Ser. B. Stat. Methodol. 77 59–83.
  • [36] Taylor, J. E., Loftus, J. R. and Tibshirani, R. J. (2016). Inference in adaptive regression via the Kac–Rice formula. Ann. Statist. 44 743–770.
  • [37] Taylor, J. E. and Worsley, K. J. (2007). Detecting sparse signals in random fields, with an application to brain mapping. J. Amer. Statist. Assoc. 102 913–928.
  • [38] Worsley, K. J., Marrett, S., Neelin, P. and Evans, A. C. (1996). Searching scale space for activation in PET images. Hum. Brain Mapp. 4 74–90.
  • [39] Worsley, K. J., Marrett, S., Neelin, P., Vandal, A. C., Friston, K. J. and Evans, A. C. (1996). A unified statistical approach for determining significant signals in images of cerebral activation. Hum. Brain Mapp. 4 58–73.
  • [40] Worsley, K. J., Taylor, J. E., Tomaiuolo, F. and Lerch, J. (2004). Unified univariate and multivariate random field theory. Neuroimage 23 S189–195.
  • [41] Zhang, C.-H. (2010). Nearly unbiased variable selection under minimax concave penalty. Ann. Statist. 38 894–942.
  • [42] Zhang, H., Nichols, T. E. and Johnson, T. D. (2009). Cluster mass inference via random field theory. Neuroimage 44 51–61.

Supplemental materials

  • Online Supplement to “Multiple testing of local maxima for detection of peaks in random fields”. In this supplement, we provide proofs for Lemma 4 and Theorems 3, 5, 8, 10 and 11.