|
KMID : 1137820170380050227
|
|
의공학회지
2017 Volume.38 No. 5 p.227 ~ p.231
|
|
|
Intelligent Diagnosing Method Based on the Conditional Probability for the Pancreatic Cancer Early Detection
|
|
Jang Ik-Gye
Jung Joon Ho
Ko Jae-Ho
Moon Hyun-Seok
Jo Yung-Ho
|
|
|
Abstract
|
|
|
|
Early diagnosis of pancreatic cancer had been considered one of the important barrier for successful therapy since the five year survival rate after treatment of pancreatic cancer was critically low. Nonetheless, patients often miss the golden time of treatment because they rarely visit the hospital until their symptoms are severe. To overcome these problems, a lot of information about the patient's symptoms should be applied as biomarkers for early diagnosis. For this reason, a biomarker for early detection of pancreatic cancer (CA19-9) has been developed as a diagnostic kit. However, since the diagnosis is not accurate enough, pancreatic symptoms (abdominal pain, jaundice, anorexia, diabetes, etc.) and biomarkers (CA19-9) should be considered together. We develop an intelligent diagnostic system that considers CA19-9 and the incidence of pancreatic cancer for pancreatic symptoms that was determined by studying a large number of patient information. It shows a higher accuracy than one using CA19-9 alone. It may increase the survival rate of pancreatic cancer because it can diagnose pancreatic cancer early.wkddl
|
|
|
KEYWORD
|
|
|
Pancreatic cancer , Intelligent diagnosis , Bayesian network
|
|
|
FullTexts / Linksout information
|
|
|
|
|
|
Listed journal information
|
|
 
|