TY - JOUR
T1 - Accuracy evaluation for region centroid-based registration of fluorescent CLSM imagery
AU - Lee, Sang Chul
AU - Bajcsy, Peter
AU - Lin, Amy
AU - Folberg, Robert
PY - 2006
Y1 - 2006
N2 - We present an accuracy evaluation of a semiautomatic registrationtechnique for 3D volume reconstruction from fluorescent confocallaser scanning microscope (CLSM) imagery. The presentedsemiautomatic method is designed based on ourobservations that (a) an accurate point selection is much harderthan an accurate region (segment) selection for a human, (b) acentroid selection of any region is less accurate by a human thanby a computer, and (c) registration based on structural shape of aregion rather than based on intensity-defined point is more robustto noise and to morphological deformation of features acrossstacks. We applied the method to image mosaicking and imagealignment registration steps and evaluated its performance with 20human subjects on CLSM images with stained blood vessels. Ourexperimental evaluation showed significant benefits of automationfor 3D volume reconstruction in terms of achieved accuracy,consistency of results, and performance time. In addition, theresults indicate that the differences between registrationaccuracy obtained by experts and by novices disappear with theproposed semiautomatic registration technique while the absoluteregistration accuracy increases.
AB - We present an accuracy evaluation of a semiautomatic registrationtechnique for 3D volume reconstruction from fluorescent confocallaser scanning microscope (CLSM) imagery. The presentedsemiautomatic method is designed based on ourobservations that (a) an accurate point selection is much harderthan an accurate region (segment) selection for a human, (b) acentroid selection of any region is less accurate by a human thanby a computer, and (c) registration based on structural shape of aregion rather than based on intensity-defined point is more robustto noise and to morphological deformation of features acrossstacks. We applied the method to image mosaicking and imagealignment registration steps and evaluated its performance with 20human subjects on CLSM images with stained blood vessels. Ourexperimental evaluation showed significant benefits of automationfor 3D volume reconstruction in terms of achieved accuracy,consistency of results, and performance time. In addition, theresults indicate that the differences between registrationaccuracy obtained by experts and by novices disappear with theproposed semiautomatic registration technique while the absoluteregistration accuracy increases.
UR - http://www.scopus.com/inward/record.url?scp=33645699528&partnerID=8YFLogxK
U2 - 10.1155/ASP/2006/82480
DO - 10.1155/ASP/2006/82480
M3 - Article
AN - SCOPUS:33645699528
SN - 1110-8657
VL - 2006
SP - 1
EP - 11
JO - Eurasip Journal on Applied Signal Processing
JF - Eurasip Journal on Applied Signal Processing
ER -