METHODS: A total of 255 dental students in Universiti Malaya completed the modified Index of Learning Styles (m-ILS) questionnaire containing 44 items which classified them into their respective LS. The collected data, referred to as dataset, was used in a decision tree supervised learning to automate the mapping of students' learning styles with the most suitable IS. The accuracy of the ML-empowered IS recommender tool was then evaluated.
RESULTS: The application of a decision tree model in the automation process of the mapping between LS (input) and IS (target output) was able to instantly generate the list of suitable instructional strategies for each dental student. The IS recommender tool demonstrated perfect precision and recall for overall model accuracy, suggesting a good sensitivity and specificity in mapping LS with IS.
CONCLUSION: The decision tree ML empowered IS recommender tool was proven to be accurate at matching dental students' learning styles with the relevant instructional strategies. This tool provides a workable path to planning student-centered lessons or modules that potentially will enhance the learning experience of the students.
METHODS: We use Non-linear Iterative Partial Least Squares to perform the data dimensionality reduction, Self-Organizing Map technique for clustering task and ensembles of Neuro-Fuzzy Inference System for predicting the hepatitis disease. We also use decision trees for the selection of most important features in the experimental dataset. We test our method on a real-world dataset and present our results in comparison with the latest results of previous studies.
RESULTS: The results of our analyses on the dataset demonstrated that our method performance is superior to the Neural Network, ANFIS, K-Nearest Neighbors and Support Vector Machine.
CONCLUSIONS: The method has potential to be used as an intelligent learning system for hepatitis disease diagnosis in the healthcare.
METHODS: A decision tree model was developed based on literature and expert inputs. An epidemiological projection model was then added to the decision tree to calculate the target population size. The budget impact of adapting the different enteral nutrition (EN) formulas was calculated by multiplying the population size with the costs of the formula and ICU length of stay (LOS). A one-way sensitivity analysis (OWSA) was conducted to examine the effect each input parameter has on the calculated output.
RESULTS: Replacing SPF with SEF would lower ICU cost by MYR 1059 (USD 216) per patient. The additional cost of increased LOS due to EFI was MYR 5460 (USD 1114) per patient. If the MOH replaces SPF with SEF for ICU patients with high EFI risk (estimated 7981 patients in 2022), an annual net cost reduction of MYR 8.4 million (USD 1.7 million) could potentially be realized in the MOH system. The cost-reduction finding of replacing SPF with SEF remained unchanged despite the input uncertainties assessed via OWSA.
CONCLUSION: Early use of SEF in ICU patients with high EFI risk could potentially lower the cost of ICU care for the MOH system in Malaysia.
PURPOSE: To examine the accuracy of AI-based radiomics in diagnosis, prognosis assessment and predicting the diagnostic value of radiomics for pelvic LN metastasis in cervical cancer patients.
MATERIAL AND METHODS: The study included 118 female patients with 660 LNs and 118 merged LNs. Four imaging histology models-decision tree, random forest, logistic regression, and support vector machine (SVM)-were created in this study. The imaging histology features were extracted from both the independent and merged LN groups. The AUC values for the test sets and the training sets of the four imaging histology models were compared for the independent LN group and the merged LN group. The DeLong test was used to compare the models.
RESULT: The imaging histology prediction model developed in the merged LN group outperformed the independent LN group in terms of test set AUC (0.668 vs. 0.535 for decision tree, 0.841 vs. 0.627 for logistic regression, 0.785 vs. 0.637 for random forest, 0.85 vs. 0.648 for SVM) and accuracy (0.754 vs. 0.676 for decision tree, 0.780 vs. 0.671 for random forest, 0.848 vs. 0.685 for logistic regression, 0.822 vs. 0.657 for SVM).
CONCLUSION: The constructed SVM imaging histology model for the merged LN group might be advantageous in predicting pelvic LN metastasis in cervical cancer.
METHODS: A large hospital-based breast cancer dataset retrieved from the University Malaya Medical Centre, Kuala Lumpur, Malaysia (n = 8066) with diagnosis information between 1993 and 2016 was used in this study. The dataset contained 23 predictor variables and one dependent variable, which referred to the survival status of the patients (alive or dead). In determining the significant prognostic factors of breast cancer survival rate, prediction models were built using decision tree, random forest, neural networks, extreme boost, logistic regression, and support vector machine. Next, the dataset was clustered based on the receptor status of breast cancer patients identified via immunohistochemistry to perform advanced modelling using random forest. Subsequently, the important variables were ranked via variable selection methods in random forest. Finally, decision trees were built and validation was performed using survival analysis.
RESULTS: In terms of both model accuracy and calibration measure, all algorithms produced close outcomes, with the lowest obtained from decision tree (accuracy = 79.8%) and the highest from random forest (accuracy = 82.7%). The important variables identified in this study were cancer stage classification, tumour size, number of total axillary lymph nodes removed, number of positive lymph nodes, types of primary treatment, and methods of diagnosis.
CONCLUSION: Interestingly the various machine learning algorithms used in this study yielded close accuracy hence these methods could be used as alternative predictive tools in the breast cancer survival studies, particularly in the Asian region. The important prognostic factors influencing survival rate of breast cancer identified in this study, which were validated by survival curves, are useful and could be translated into decision support tools in the medical domain.
Objective: The aim of this study was to conduct a cost-effectiveness analysis of spending on healthcare R&D to address the needs of developing innovative therapeutic products in Indonesia.
Methods: A decision tree model was developed by taking into account four stages of R&D: stage 1 from raw concept to feasibility, stage 2 from feasibility to development, stage 3 from development to early commercialization, and stage 4 from early to full commercialization. Considering a 3-year time horizon, a stage-dependent success rate was applied and analyses were conducted from a business perspective. Two scenarios were compared by assuming the government of Indonesia would increase GERD in health and medical sciences up to 2- and 3-times higher than the baseline (current situation) for the first and second scenario, respectively. Cost per number of innovative products in health and medical sciences was considered as the incremental cost-effectiveness ratio (ICER). Univariate sensitivity analysis was conducted to investigate the effects of different input parameters on the ICER.
Results: There was a statistically significant association (P-value<0.05) between countries' GERD in medical and health sciences with the number of innovative products. We estimated the ICER would be $8.50 million and $2.04 million per innovative product for the first and second scenario, respectively. The sensitivity analysis showed that the success rates in all stages and total GERD were the most influential parameters impacting the ICER.
Conclusion: The result showed that there was an association between GERD in medical and health sciences with the number of innovative products. In addition, the second scenario would be more cost-effective than the first scenario.
METHOD: This study proposes a combination of decision tree and logistic regression techniques to model crash severity (injury vs. noninjury), because the combined approach allows the specification of nonlinearities and interactions in addition to main effects. Both a scobit model and a random parameters logit model, respectively accounting for an imbalance response variable and unobserved heterogeneities, are tested and compared. The study data set contains a total of 5 years of crash data (2008-2012) on selected mountainous highways in Malaysia. To enrich the data quality, an extensive field survey was conducted to collect detailed information on horizontal alignment, longitudinal grades, cross-section elements, and roadside features. In addition, weather condition data from the meteorology department were merged using the time stamp and proximity measures in AutoCAD-Geolocation.
RESULTS: The random parameters logit model is found to outperform both the standard logit and scobit models, suggesting the importance of accounting for unobserved heterogeneity in crash severity models. Results suggest that proportion of segment lengths with simple curves, presence of horizontal curves along steep gradients, highway segments with unsealed shoulders, and highway segments with cliffs along both sides are positively associated with injury-producing crashes along rural mountainous highways. Interestingly, crashes during rainy conditions are associated with crashes that are less likely to involve injury. It is also found that the likelihood of injury-producing crashes decreases for rear-end collisions but increases for head-on collisions and crashes involving heavy vehicles. A higher order interaction suggests that single-vehicle crashes involving light and medium-sized vehicles are less severe along straight sections compared to road sections with horizontal curves. One the other hand, crash severity is higher when heavy vehicles are involved in crashes as single vehicles traveling along straight segments of rural mountainous highways.
CONCLUSION: In addition to unobserved heterogeneity, it is important to account for higher order interactions to have a better understanding of factors that influence crash severity. A proper understanding of these factors will help develop targeted countermeasures to improve road safety along rural mountainous highways.