Taipei Medical University

A B C D E F G H I J K L M N O P Q R S T U V W X Y Z
Chang TC
------>authors3_c=
------>paper_class1=1
------>Impact_Factor=None
------>paper_class3=1
------>paper_class2=1
------>vol=4292
------>confirm_bywho=wingchan
------>insert_bywho=ed100975
------>Jurnal_Rank=None
------>authors4_c=
------>comm_author=
------>patent_EDate=None
------>authors5_c=
------>publish_day=1
------>paper_class2Letter=None
------>page2=38
------>medlineContent=
------>unit=E0119
------>insert_date=20081223
------>iam=5
------>update_date=None
------>author=???
------>change_event=4
------>ISSN=
------>authors_c=
------>score=145
------>journal_name=Lecture Notes in Computer Science
------>paper_name=The Diagnostic Application of Brain Image Processing and Analysis System for Ischemic Stroke
------>confirm_date=20081224
------>tch_id=097094
------>pmid=15780892
------>page1=31
------>fullAbstract=Immunochemical staining techniques are commonly used to assess neuronal, astrocytic and microglial alterations in experimental neuroscience research, and in particular, are applied to tissues from animals subjected to ischemic stroke. Immunoreactivity of brain sections can be measured from digitized immunohistology slides so that quantitative assessment can be carried out by computer-assisted analysis. Conventional methods of analyzing immunohistology are based on image classification techniques applied to a specific anatomic location at high magnification. Such micro-scale localized image analysis limits one for further correlative studies with other imaging modalities on whole brain sections, which are of particular interest in experimental stroke research. This report presents a semi-automated image analysis method that performs convolution-based image classification on micro-scale images, extracts numerical data representing positive immunoreactivity from the processed micro-scale images and creates a corresponding quantitative macro-scale image. The present method utilizes several image-processing techniques to cope with variances in intensity distribution, as well as artifacts caused by light scattering or heterogeneity of antigen expression, which are commonly encountered in immunohistology. Micro-scale images are composed by a tiling function in a mosaic manner. Image classification is accomplished by the K-means clustering method at the relatively low-magnification micro-scale level in order to increase computation efficiency. The quantitative macro-scale image is suitable for correlative analysis with other imaging modalities. This method was applied to different immunostaining antibodies, such as endothelial barrier antigen (EBA), lectin, and glial fibrillary acidic protein (GFAP), on histology slides from animals subjected to middle cerebral artery occlusion by the intraluminal suture method. Reliability tests show that the results obtained from immunostained images at high magnification and relatively low magnification are virtually the same.
------>tmu_sno=None
------>sno=20936
------>authors2=Lee JD
------>authors3=Huang CH
------>authors4=Wu T
------>authors5=Chen CJ
------>authors6=Wu SJ
------>authors6_c=
------>authors=Chang TC
------>delete_flag=0
------>SCI_JNo=None
------>authors2_c=
------>publish_area=0
------>updateTitle=Semi-automated image processing system for micro- to macro-scale analysis of immunohistopathology: application to ischemic brain tissue.
------>language=2
------>check_flag=None
------>submit_date=None
------>country=NULL
------>no=
------>patent_SDate=None
------>update_bywho=None
------>publish_year=2006
------>submit_flag=None
------>publish_month=11
A B C D E F G H I J K L M N O P Q R S T U V W X Y Z