ABSTRACT
Background: The importance of early diagnosis of a hazardous illness cannot be overstated. The transmission rate is extremely high, especially in the current pandemic condition. The ability to predict epidemics will aid public health in reducing mortality and morbidity. Machine Learning (ML) approaches are used in the construction of an effective disease prognosis model. Furthermore, only if the model learns good associated features from the data is it possible to generate a speedy outcome. As a result, selecting features is also necessary before beginning the forecasting process. Objective(s): However, because of the virus's dynamic structure, it's difficult to predict Nipah disease and/or zoonotic infection. Furthermore, there is no clinical treatment for Nipah. The major goal of this research is to develop a prognostic model for early diagnosis of Nipah disease using a combination of several clinical factors such as symptoms, disease incubation information, and routine blood test results confirmed by a lab technician.Proposed System: The healthcare application and data are more complex to handle than other ML applications since various clinical features are assessed throughout disease manifestation. As a result, selecting the most relevant variables is critical when designing a prognosis model for any viral disease. To deal with clinical features from a vast number of features, we proposed a Restricted Boltzmann Machine (RBM) method in this research. Additionally, we employed a hybrid ensemble learning method to predict if the patient was infected with NiV after choosing features using the RBM. Data Collection: The proposed system is being implemented using the NiV infection dataset that erupted in Kozhikode, Kerala in 2018 and 2019. Result(s): The developed stacking-based ensemble Meta classifier was successfully implemented using the python programming language, and its performance was evaluated using a variety of metrics includingaccuracy, precision, recall, f1-score, log loss, AUROC and MCC. Our proposed Stacking Ensemble Meta Classifier (SEMC) model achieved an accuracy rate of 88.3% with a log loss of 0.36. Model precision, recall, f1-score, AUROC, and MCC value were 92.5%, 89.2%, 90.9%, 92.1%, and 0.74 respectively. In addition, we calculated the gravitational pull of each feature using the SHAP approach and discovered that altered sensorium, fever, headache, and cough were the most critical clinical indicators that distinguished NiVD infection from our dataset. Therefore, this classification may assist the pathologist in diagnosing NiVD with symptoms before performing the RT-PCR medical test. Conclusion(s): Using our proposed SEMC technique, we developed a prognostic model for the diagnosis of Nipah in humans. The proposed technique's discriminatory efficiency exhibited good NiVD diagnosis efficacy. We anticipate that this model will aid medics in determining a prognosis more quickly during future epidemics. However, to achieve maximum accuracy, the model requires more unique samples.Copyright © 2022 Wolters Kluwer Medknow Publications. All rights reserved.
ABSTRACT
Literature survey shows that convolutional neural network (CNN)-based pretrained models have been largely used for CoronaVirus Disease 2019 (COVID-19) classification using chest X-ray (CXR) and computed tomography (CT) datasets. However, most of the methods have used a smaller number of data samples for both CT and CXR datasets for training, validation, and testing. As a result, the model might have shown good performance during testing, but this type of model will not be more effective on unseen COVID-19 data samples. Generalization is an important term to be considered while designing a classifier that can perform well on completely unseen datasets. Here, this work proposes a large-scale learning with stacked ensemble meta-classifier and deep learning-based feature fusion approach for COVID-19 classification. The features from the penultimate layer (global average pooling) of EfficientNet-based pretrained models were extracted and the dimensionality of the extracted features reduced using kernel principal component analysis (PCA). Next, a feature fusion approach was employed to merge the features of various extracted features. Finally, a stacked ensemble meta-classifier-based approach was used for classification. It is a two-stage approach. In the first stage, random forest and support vector machine (SVM) were applied for prediction, then aggregated and fed into the second stage. The second stage includes logistic regression classifier that classifies the data sample of CT and CXR into either COVID-19 or Non-COVID-19. The proposed model was tested using large CT and CXR datasets, which are publicly available. The performance of the proposed model was compared with various existing CNN-based pretrained models. The proposed model outperformed the existing methods and can be used as a tool for point-of-care diagnosis by healthcare professionals.