THE COLORENHANCEIVIENT FOR DIAGNOSIS GF SMALL GASTRIC CANCERS

-USING NEW PROGRAMING SOFTWARE FOR ENDOSCOPIC IMAGE PRUCESSING-

Abstract

We have developed the image processing software for finding out small gastriccancers. This would call endoscopists' attention to small gastric cancers by color enhancement of the lesions and this is useful for even dark distant views and far small lesions withfine color difference. This software consists of two steps on the programs. The first stepis to convert the original pictures into R component pictures (Rn' =Rn/Rn+Gn+Bn × M: M=average of the brightness) . The second step is to enhance this R component picture bythe extend of dynamic range and the non-linear enhancement of brightness. We appliedthis image processing method far early gastric cancers and gastric adenoma. The finedifference among colors on gastric mucosal surface was enhanced and the difference of thelesions from surrounding mucosa on colors became conspicuous for attractingendoscopist's attention. The image processing of endoscopic pictures is expected todevelop the originality in ids algorithms for the diagnostic usefulness.