|
USING GREY-SCALE HIT-OR-MISS TRANSFORM FOR DETECTION OF ISOLATED FOREGROUND PIXELS IN CEREBRAL RMN DATASETS
|
|
|
L. Pana;S. Moldovanu;L. Moraru
|
|
|
||
|
|
|
|
1314-2704
|
|
|
||
|
English
|
|
|
19
|
|
|
6.3
|
|
|
|
|
|
||
|
MRI images contain a lot of subtle information related to various lesions which are difficult to be picked up by radiologists. Computer aided diagnosis CAD is a valuable tool to improve the ability of an average radiologist to diagnose the subtle lesions. This study uses the isolated foreground pixels in MRI images as a feature able to discern a stroke patient by a healthy one. Brain MRI image was divided into eight equal sectors. The isolated foreground pixels (i.e. pixels satisfying a neighborhood configuration that corresponds to an isolated foreground pixel) were extracted using hit-or-miss and skeleton transformations. We have tested the proposed algorithms using two cerebral image datasets (healthy and acute stroke patients). The mean ? SD values of isolated foreground pixels for stroke patients systematically exceed the corresponding values for healthy patients. The higher numbers of isolated foreground pixel determined using hit-or-miss transform indicate this method as a promising approach for a simple and quick evaluation of stroke.
|
|
|
conference
|
|
|
||
|
||
|
19th International Multidisciplinary Scientific GeoConference SGEM 2019
|
|
|
19th International Multidisciplinary Scientific GeoConference SGEM 2019, 9 - 11 December, 2019
|
|
|
Proceedings Paper
|
|
|
STEF92 Technology
|
|
|
International Multidisciplinary Scientific GeoConference-SGEM
|
|
|
Bulgarian Acad Sci; Acad Sci Czech Republ; Latvian Acad Sci; Polish Acad Sci; Russian Acad Sci; Serbian Acad Sci & Arts; Slovak Acad Sci; Natl Acad Sci Ukraine; Natl Acad Sci Armenia; Sci Council Japan; World Acad Sci; European Acad Sci, Arts & Letters; Ac
|
|
|
229-236
|
|
|
9 - 11 December, 2019
|
|
|
website
|
|
|
cdrom
|
|
|
6685
|
|
|
cerebral RMN; hit-or-miss; skeletonization; isolated foreground pixel; stroke
|
|