대한신장학회

Skip Navigation: Skip to contents

KOR
ENG
전체메뉴보기

간행물 검색

현재 페이지 경로

HOME
간행물
간행물 검색

논문분류	춘계학술대회 초록집
제목	Predicting kidney disease using artificial intelligence
저자	Jakir Hossain Bhuiyan Masud
출판정보	2023; 2023(1):
키워드
초록	Objectives: Artificial Intelligence (AI) solves many problems in data science. AI is the prediction of an outcome based upon existing data such as electronic health record. Text classification is an AI technology that uses natural language processing (NLP) to transform the free text in documents and databases into normalized, structured data suitable for analysis or to drive AI algorithms. AI can play an essential role in predicting presence or absence of kidney disease. This can provide important insights to doctors who can then adapt their diagnosis and treatment. This paper investigates classification method, which is used for predicting kidney disease from clinical note by multilabel classifiers. Methods: We collected clinical notes from the hospital electronic health record data of 2016. We used keras python tool for analyzing the data. Our method is multilabel text classification using convolutional neural network (CNN) to predict kidney disease from clinical notes. Experiments with this tool were performed using a nephrology department dataset. The concentration of this paper is to achieve state of art accuracy of classification algorithms, and to show its utility to predict disease. Results: The results of this study show that ensemble technique, such as CNN, is effective in improving the prediction accuracy of classifiers, and exhibit satisfactory performance in identifying risk of kidney disease. Our study model achieves better precision 57%, recall 93 % and F-measure 71 % for the kidney disease classification. This study also shows the higher-ranking prediction through probability. Conclusions: Our study achieves the state of art accuracy through prediction model. Thus, physician can find some alerts to identify most probable disease with different ranking of disease. In future we may include more important variables that can help to identify more convenient prediction.
원문(PDF)	PDF 원문보기

목록

위로가기