TY - JOUR
T1 - The significance of image analysis for cancer diagnosis
AU - Sela, Jona J.
AU - Bruckstien, Alfred
AU - Goshen, Gal
AU - Dubin, Uri
AU - Karasikov, Nir
AU - Kopolovic, Juri
PY - 2012/6
Y1 - 2012/6
N2 - Cancer diagnosis is established by pathologists by means of microscopical analysis of cells and tissues. The accuracy of this methodology is dependent upon human skills, training, and judgment. The conventional diagnostic methods are currently re-examined in view of innovative trends in image analysis. The progress in cancer diagnosis by integration of computer assisted analysis is forthcoming. The current study presents several methods of computer assisted histopathological analysis that were recently developed. Biopsy tissue-sections served for the acquisition of microscopical images. These were enhanced and analyzed by dedicated algorithmic functions to evaluate complex and customized features of cell and tissue textures. The methods were designated to differentiate normal from neoplastic features, assisting the pathologist with the diagnosis. Consequently, the assembly of the data by the system, identification of specific cellular and tissue patterns and properties, has been proven to be useful for assessing efficacy of cancer therapy. The technological development was based on models of carcinoma that served for computer analysis of typical neoplastic changes in cells and tissues. The developing of a computerized diagnostic system on one type of cancer was followed effortlessly by its application to other types of cancer. Decision making system, to replace the pathologist, is not proposed. However, a substantial support to the diagnostic process is pertinent. In time, with increase of practice, this methodology is expected to become more accurate than the human eye and mind, in detecting minute deviations in cellular and tissue structures.
AB - Cancer diagnosis is established by pathologists by means of microscopical analysis of cells and tissues. The accuracy of this methodology is dependent upon human skills, training, and judgment. The conventional diagnostic methods are currently re-examined in view of innovative trends in image analysis. The progress in cancer diagnosis by integration of computer assisted analysis is forthcoming. The current study presents several methods of computer assisted histopathological analysis that were recently developed. Biopsy tissue-sections served for the acquisition of microscopical images. These were enhanced and analyzed by dedicated algorithmic functions to evaluate complex and customized features of cell and tissue textures. The methods were designated to differentiate normal from neoplastic features, assisting the pathologist with the diagnosis. Consequently, the assembly of the data by the system, identification of specific cellular and tissue patterns and properties, has been proven to be useful for assessing efficacy of cancer therapy. The technological development was based on models of carcinoma that served for computer analysis of typical neoplastic changes in cells and tissues. The developing of a computerized diagnostic system on one type of cancer was followed effortlessly by its application to other types of cancer. Decision making system, to replace the pathologist, is not proposed. However, a substantial support to the diagnostic process is pertinent. In time, with increase of practice, this methodology is expected to become more accurate than the human eye and mind, in detecting minute deviations in cellular and tissue structures.
KW - Artificial Intelligence
KW - Cancer
KW - Computerized Image Analysis
KW - Fuzzy Logic
KW - Histopathological Diagnosis
KW - Microscopy
KW - Neoplasia
KW - Neural Networks
UR - http://www.scopus.com/inward/record.url?scp=84870855727&partnerID=8YFLogxK
U2 - 10.1166/jamr.2012.1102
DO - 10.1166/jamr.2012.1102
M3 - ???researchoutput.researchoutputtypes.contributiontojournal.article???
AN - SCOPUS:84870855727
SN - 2156-7573
VL - 7
SP - 91
EP - 97
JO - Journal of Advanced Microscopy Research
JF - Journal of Advanced Microscopy Research
IS - 2
ER -