China
Africa
Explore chapters and articles related to this topic
Swarm Intelligence and Evolutionary Algorithms for Heart Disease Diagnosis
Published in Sandeep Kumar, Anand Nayyar, Anand Paul, Swarm Intelligence and Evolutionary Algorithms in Healthcare and Drug Development, 2019
In case of CHD, the input data consists of three significant attributes inorder to facilitate the prediction process. These attributes are age, sex and symptom intensity. The output attribute for prediction was a single value that represents the presence of disease in the particular patient. In case of coronary artery diseases (CAD), the transformation parameters correspond to heart blood vessels. The level of narrowing in the blood vessel indicates the health condition of artery. The blood vessels presenting the artery disease are left anterior descending (LAD) artery, right coronary artery (RCA), left circumflex (LCX) artery and left main (LM) coronary artery. The evaluation points for measuring the CAD based on LCX, RCA and LAD artery laid between 50% and 70%. In cardiovascular clinical observation, an artery narrowing evaluation value below 50% represents the health artery. The clinical value above 50% represents the patient with borderline artery disease. Finally, the clinical evaluation value 70% and above represents patient suffering from significant heart disease. The fourth critical artery is the LM blood vessel. The clinical threshold point for LM artery is between 30% and 50%. The reason is that LM contributes toward heart disease. The more the blockage of LM artery, the probability of heart disease is higher level than other three arteries. Dangare and Apte. [33] provided an overview of different data mining techniques for heart disease prediction. The decision tree algorithms have further variants namely C4.5, Iterative Dichotomized 3 (ID3) and Classification and Regression Tree (CART). In this chapter, the different heart disease diagnosis through various evaluation techniques were analyzed and compared. The main aim was to develop a systematic prototype for knowledge generation based on the patient medical records to diagnosis heart disease. For this purpose, hybridization of GA was incorporated. The authors purport that this prototype is significant to the medical practitioners to make correct decision on heart disease and hence facilitate toprovide timely recovery treatments for the patients. Bouktif [21] and Jabbar et al.[38] performed evaluative study on state-of-the-art work in research community towards diagnosis and prediction methods for heart diseases. The data mining techniques like Naive Bayes, K-nearest Neighbour Classification, Support Vector Machine, and Artificial Neural Network were considered for the purpose. Nguyen et al. [41], SVM exhibited overall performance and highly promising in the prediction of CHD. In case of CVD diagnosis, the decision tree-based feature reduction exhibited significant classification.
Ontology-Based decision tree model for prediction of fatty liver diseases
Published in Computer Methods in Biomechanics and Biomedical Engineering, 2023
Seyed Yashar Banihashem, Saman Shishehchi
RapidMiner Studio version 9.9 has various tools for data mining. Different decision trees are available in the operator section. Three decision tree types, i.e. ‘ID3’ (Peng et al., 2009), ‘CHAID’ (Milanović and Stamenković 2016), and ‘Decision Tree’ (Dryer et al. 2018), were selected to design the model. As aforementioned, all collected data were numerical. The first dataset loaded to the Rapidminer, then using the ‘Split Data’ tool, the dataset divided into learning (70%) and testing (30%) parts. All three decision tree types connected to the training dataset. After running, the error message appeared on the screen and informed that ID3 and CHAID could not be employed. Because they don’t support numerical attributes, it needs to mention that our dataset data type is numerical. Finally, the Decision tree model was chosen to use for modelling. In the rest of this paper the decision tree will refer as DT.
Tanshinone IIA alleviates ovalbumin-induced allergic rhinitis symptoms by inhibiting Th2 cytokine production and mast cell histamine release in mice
Published in Pharmaceutical Biology, 2022
Qing Chen, Liping Shao, Yong Li, Mian Dai, He Liu, Nan Xiang, Hui Chen
Mice serum was collected, and the concentrations of OVA-IgE and OVA-immunoglobulin G1 (IgG1) in mice serum were measured according to the instructions of mice OVA-IgE enzyme-linked immunosorbent assay (ELISA) kit (F10731, Westang, Shanghai, China) and mice OVA-IgG1 ELISA kit (3013, Chondrex, Woodinville, WA), respectively. Anti-mouse IgE monoclonal antibody (100 mL) was added to each well of 96-well ELISA plates, followed by incubation at 4 °C overnight. The plate was washed with diluted wash buffer after the coating solution was discarded. With the supplement of 100 μL blocking buffer to each well, the plate was maintained at room temperature for 1 h. After the solution was discarded, each well of 96-well ELISA plates was added with the prepared standard sample for 90 min of incubation at room temperature. Following the addition of 100 μL dissolved biotinylated OVA to each well, the plate was cultured at room temperature for another 90 min. Subsequently, the 96-well ELISA plates were washed, each well of which was supplemented with 100 μL TMB solution, followed by the culture in the dark at room temperature for 30 min. Finally, 50 μL stop solution was put to each well of 96-well ELISA plates. The optical density (OD) value was measured at a wavelength of 450 nm with a microplate reader (SpectraMax iD3, Molecular Devices, Silicon Valley, CA).
A hybrid clustering and classification approach for predicting crash injury severity on rural roads
Published in International Journal of Injury Control and Safety Promotion, 2018
Seyed Hessam-Allah Hasheminejad, Mohsen Zahedi, Seyed Mohammad Hossein Hasheminejad
To evaluate the quality of the identified rules with respect to their Support and Confidence measures, we compare the identified rules with those obtained by three famous rule-based classifications, including ID3, CART and C4.5 algorithms. In Table 2, the rules obtained by the last generation of the proposed GA method are compared to those obtained by ID3, CART and C4.5 algorithms according to the best rules and the average and standard deviation. As shown in Table 2, the proposed GA method outperforms other rule-based classification algorithms according to all the Support and Confidence measures.