METHODS: A dataset containing medical records of 809 patients suspected to suffer from ACS was used. For each subject, 266 clinical factors were collected. At first, a feature selection was performed based on interviews with 20 cardiologists; thereby 40 seminal features for classifying ACS were selected. Next, a feature selection algorithm was also applied to detect a subset of the features with the best classification accuracy. As a result, the feature numbers considerably reduced to only seven. Lastly, based on the seven selected features, eight various common pattern recognition tools for classification of ACS were used.
RESULTS: The performance of the aforementioned classifiers was compared based on their accuracy computed from their confusion matrices. Among these methods, the multi-layer perceptron showed the best performance with the 83.2% accuracy.
CONCLUSION: The results reveal that an integrated AI-based feature selection and classification approach is an effective method for the early and accurate classification of ACS and ultimately a timely diagnosis and treatment of this disease.
METHODS: Diabetes data were derived from the Malaysian National Health and Morbidity Surveys conducted in 2006, 2011 and 2015. The air pollution data (NOx, NO2, SO2, O3 and PM10) were obtained from the Department of Environment Malaysia. Using multiple logistic and linear regression models, the association between long-term exposure to these pollutants and prevalence of diabetes among Malaysian adults was evaluated.
RESULTS: The PM10 concentration decreased from 2006 to 2014, followed by an increase in 2015. Levels of NOx decreased while O3 increased annually. The air pollutant levels based on individual modelled air pollution exposure as measured by the nearest monitoring station were higher than the annual averages of the five pollutants present in the ambient air. The prevalence of overall diabetes increased from 11.4% in 2006 to 21.2% in 2015. The prevalence of known diabetes, underdiagnosed diabetes, overweight and obesity also increased over these years. There were significant positive effect estimates of known diabetes at 1.125 (95% CI, 1.042, 1.213) for PM10, 1.553 (95% CI, 1.328, 1.816) for O3, 1.271 (95% CI, 1.088, 1.486) for SO2, 1.124 (95% CI, 1.048, 1.207) for NO2, and 1.087 (95% CI, 1.024, 1.153) for NOx for NHMS 2006. The adjusted annual average levels of PM10 [1.187 (95% CI, 1.088, 1.294)], O3 [1.701 (95% CI, 1.387, 2.086)], NO2 [1.120 (95% CI, 1.026, 1.222)] and NOx [1.110 (95% CI, 1.028, 1.199)] increased significantly from NHMS 2006 to NHMS 2011 for overall diabetes. This was followed by a significant decreasing trend from NHMS 2011 to 2015 [0.911 for NO2, and 0.910 for NOx].
CONCLUSION: The findings of this study suggest that long-term exposure to O3 is an important associated factor of underdiagnosed DM risk in Malaysia. PM10, NO2 and NOx may have mixed effect estimates towards the risk of DM, and their roles should be further investigated with other interaction models. Policy and intervention measures should be taken to reduce air pollution in Malaysia.