Comparison of prediction fuzzy modeling towards high-risk symptoms of lung cancer

• Published in: Journal of Intelligent and Fuzzy Systems
• Main Author: Zakaria A.S.; Shafi M.A.; Mohd Zim M.A.; Musa A.N.
• Format: Article
• Language: English
• Published: IOS Press BV 2024
• Online Access: https://www.scopus.com/inward/record.uri?eid=2-s2.0-85193707835&doi=10.3233%2fJIFS-233714&partnerID=40&md5=e8c82cebbfd6943f7e8bcca9bad354f8

Lung cancer constituted 12.2% of newly diagnosed cancer cases globally in 2020. The high fatality rate of the condition is attributed to delayed diagnosis and inadequate symptom recognition. In Malaysia, the incidence of lung cancer is estimated to be 1 in 60 males and 1 in 138 females, with a median age of 70 years or above. Most lung cancer cases were detected during advanced stages, specifically stages III and IV, with a prevalence exceeding 90% for both genders. In Malaysia, most patients are diagnosed in stages III and IV, which are associated with a lower likelihood of long-Term survival. Many cases are identified at a late stage, characterized by significant tumor expansion or the spread of cancer cells to areas that cannot be treated surgically. Malaysians are unaware of cancer symptoms; hence the situation is common. To improve survival and reduce mortality, Malaysians must recognize the symptoms of lung cancer. Fuzzy linear regression and multiple linear regression models have been compared to predict high-risk lung cancer symptoms in Malaysia. The fuzzy linear regression model analyses secondary data, eliminates irrelevant information and enhances precision in the results. Lung cancer patients at Al-Sultan Abdullah Hospital (UiTM Hospital) in Selangor provided data for this study. Data from 124 lung cancer patients were analyzed using Microsoft Excel, SPSS, and MATLAB. To improve data accuracy, the study used cross-validation measurement error (MSE and RMSE). According to data analysis, hemoptysis and chest pain are high-risk symptoms with MSE and RMSE values of 1.549 and 1.245, respectively. © 2024-IOS Press. All rights reserved.