A region-based segmentation method for multichannel image data

被引:15
作者
Bose, S [1 ]
OSullivan, F [1 ]
机构
[1] UNIV WASHINGTON, DEPT STAT, SEATTLE, WA 98195 USA
关键词
cross-validation; empirical rates of convergence; image analysis; positron emission tomography; recursive merging; POSITRON-EMISSION TOMOGRAPHY; MODEL; APPROXIMATION; ERRORS; PET;
D O I
10.2307/2291453
中图分类号
O21 [概率论与数理统计]; C8 [统计学];
学科分类号
020208 ; 070103 ; 0714 ;
摘要
Segmentation has become a widely used tool in modern image analysis. Typical methods are highly nonlinear and often ad hoc, and none so far have proved accessible to detailed statistical analysis even by asymptotic approximations. Can such methods be any good? This article describes a segmentation procedure for multichannel image data and attempts to develop an understanding of its statistical performance characteristics in terms of nonparametric regression. The procedure is based on a recursive merging algorithm defined via a nested sequence of discretizations of the image domain. A cross-validation rule with a near-neighbor block replacement strategy is proposed for selecting the final segmentation model. Idealized numerical simulation experiments are used to evaluate the rare of convergence of the mean square error of estimation and also to study the efficiency of the cross-validation technique. Interestingly, the results show that the rate of convergence tends to decrease as the degree of smoothness of the underlying image increases. This complements the more familiar estimation characteristic associated with conventional nonparametric regression algorithms. The cross-validation is found to be effective at choosing a segmentation that minimizes the mean squared deviation between the segmented image and the underlying truth. Furthermore, by manipulating the block size in the replacement scheme, it is possible to maintain some robustness to artifacts caused by blurring. Physical phantom datasets taken from positron emission tomography (PET) are used to evaluate the segmentation procedure in a setting of practical interest. Some data from a human brain imaging study with PET are used for further illustration.
引用
收藏
页码:92 / 106
页数:15
相关论文
共 35 条
[1]  
ADAMS JB, 1986, J GEOPHYS RES-SOLID, V91, P8098, DOI 10.1029/JB091iB08p08098
[2]   SEEDED REGION GROWING [J].
ADAMS, R ;
BISCHOF, L .
IEEE TRANSACTIONS ON PATTERN ANALYSIS AND MACHINE INTELLIGENCE, 1994, 16 (06) :641-647
[3]   KERNEL SMOOTHING OF DATA WITH CORRELATED ERRORS [J].
ALTMAN, NS .
JOURNAL OF THE AMERICAN STATISTICAL ASSOCIATION, 1990, 85 (411) :749-759
[4]   POSITRON EMISSION TOMOGRAPHIC MEASUREMENTS OF CEREBRAL GLUCOSE-UTILIZATION USING [1-C-11]D-GLUCOSE [J].
BLOMQVIST, G ;
STONEELANDER, S ;
HALLDIN, C ;
ROLAND, PE ;
WIDEN, L ;
LINDQVIST, M ;
SWAHN, CG ;
LANGSTROM, B ;
WIESEL, FA .
JOURNAL OF CEREBRAL BLOOD FLOW AND METABOLISM, 1990, 10 (04) :467-483
[5]  
Breiman L., 1984, CLASSIFICATION REGRE
[6]   AN APPROXIMATION FORMULA FOR THE VARIANCE OF PET REGION-OF-INTEREST VALUES [J].
CARSON, RE ;
YAN, Y ;
DAUBEWITHERSPOON, ME ;
FREEDMAN, N ;
BACHARACH, SL ;
HERSCOVITCH, P .
IEEE TRANSACTIONS ON MEDICAL IMAGING, 1993, 12 (02) :240-250
[7]  
CHAMBERS JM, 1992, STAT MODELS S, DOI DOI 10.1201/9780203738535
[8]  
CHAUDHURI P, 1994, STAT SINICA, V4, P143
[9]   SPLIT-AND-MERGE IMAGE SEGMENTATION BASED ON LOCALIZED FEATURE ANALYSIS AND STATISTICAL TESTS [J].
CHEN, SY ;
LIN, WC ;
CHEN, CT .
CVGIP-GRAPHICAL MODELS AND IMAGE PROCESSING, 1991, 53 (05) :457-475
[10]  
CHOI HS, 1989, SPIE, V1092, P183