METHODS: The algorithm was developed using data from 345 TDT patients. Spearman's rank correlation was used to evaluate the conceptual overlap between the instruments. Model specifications were chosen using a stepwise regression. Both direct and response mapping methods were attempted. Six mapping estimation methods ordinary least squares (OLS), a log-transformed response using OLS, generalized linear model (GLM), two-part model (TPM), Tobit and multinomial logistic regression (MLOGIT) were tested to determine the root mean squared error (RMSE) and mean absolute error (MAE). Other criterion used were accuracy of the predicted utility score, proportions of absolute differences that was less than 0.03 and intraclass correlation coefficient. An in-sample, leave-one-out cross validation was conducted to test the generalizability of each model.
RESULTS: The best performing model was specified with three out of the four PedsQL GCS scales-the physical, emotional and social functioning score. The best performing estimation method for direct mapping was a GLM with a RMSE of 0.1273 and MAE of 0.1016, while the best estimation method for response mapping was the MLOGIT with a RMSE of 0.1597 and MAE of 0.0826.
CONCLUSION: The mapping algorithm developed using the GLM would facilitate the calculation of utility scores to inform economic evaluations for TDT patients when EQ-5D data is not available. However, caution should be exercised when using this algorithm in patients who have poor quality of life.
METHODS: This cross-sectional study was conducted from March-November 2014 in the form of a telephone survey. Participants aged 40 years and above were randomly selected across Malaysia and interviewed using the validated Awareness Beliefs about Cancer (ABC) measurement tool. Linear regression was conducted to test the association between symptom and risk factor recognition and socio-demographic variables.
RESULTS: A sample of 1895 participants completed the survey. On average, participants recognised 5.8 (SD 3.2) out of 11 symptoms and 7.5 (SD 2.7) out of 12 risk factors. The most commonly recognised symptom was 'lump or swelling' (74.5%) and the most commonly recognised risk factor was 'smoking' (88.7%). Factors associated with prompted awareness were age, ethnicity, education and smoking status.
CONCLUSION: Recognition of symptom and risk factors for most cancers was relatively low across Malaysia compared to previous studies in high-income countries and to studies conducted in Malaysia. There is a need to conduct regular public health campaigns and interventions designed to improve cancer awareness and knowledge as a first step towards increasing the early detection of cancer.
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.