Symptom Analysis Using Fuzzy Logic for Detection and Monitoring of COVID-19 Patients

Abstract

Recent developments regarding the Internet of Things (IoT), Artificial Intelligence (AI), and Machine Learning (ML) opened new horizons of healthcare opportunities. Moreover, these technological advancements give strength to face upcoming healthcare challenges. One of such challenges is the advent of COVID-19, which has adverse effects beyond comprehension. Therefore, utilizing the basic functionalities of IoT, this work presents a real-time rule-based Fuzzy Logic classifier for COVID-19 Detection (FLCD). The proposed model deploys the IoT framework to collect real-time symptoms data from users to detect symptomatic and asymptomatic Covid-19 patients. Moreover, the proposed framework is also capable of monitoring the treatment response of infected people. FLCD constitutes three components: symptom data collection using wearable sensors, data fusion through Rule-Based Fuzzy Logic classifier, and cloud infrastructure to store data with a possible verdict (normal, mild, serious, or critical). After extracting the relevant features, experiments with a synthetic COVID-19 symptom dataset are conducted to ensure effective and accurate detection of COVID-19 cases. As a result, FLCD successfully acquired 95% accuracy, 94.73% precision, 93.35% recall, and showed a minimum error rate of 2.52%.

Classification of COVID-19 on Chest X-Ray Images Through the Fusion of HOG and LPQ Feature Sets

“…The fusion strategy was the most potent for estimation, with an accuracy of 95.56 ± 0.57%, recall of 95.61 ± 0.79%, precision of 95.5 ± 0.61%, F-score of 95.55 ± 0.57%, specificity of 95.61 ± 0.81%, and MCC of 91.12 ± 1.15% compared to other feature extraction procedures. (6)(7)(8) contain a list of all the SVM models' performance outcomes. Based on the investigational consequences for all three scenarios, it can be confirmed that the features combined from the HOG and LPQ techniques performed better than the other scenarios using the SVMP, SVML, and SVMRBF classifiers.…”

“…However, the fusion technique with the SVMP classifier have delivered an extremely maximum average accuracy of 97.15 ± 0.38% among the remaining classifiers for all scenarios. The misclassification error rate indicator was included in this work to evaluate the performance of the suggested scenarios, as revealed in Figure (8). The consequences presented that combining HOG and LPQ feature sets to the SVM classifier with a polynomial kernel produced a lowest misclassification error of 2.85%, demonstrating that the suggested scenario operates significantly better than earlier suggested scenarios.…”

“…The Support Vector Machine (SVM) technique was performed to classify the effective features selected by the Relief feature selection procedure at all levels of the architecture. Tayyaba Ilyas [8] suggested for COVID-19 detection, it is proposed to offer a real time rule based Fuzzy Logic classifier (FLCD). The suggested methodology collects real-time symptom data from users through the use of an Internet of Things (IoT) platform to identify symptomatic and asymptomatic COVID-19 patients.…”

An Interval Type-3 Fuzzy–Fractal Approach for Plant Monitoring

“…Recently, we have seen the utilization of neural networks, evolutionary computing, and fuzzy systems in this area. In the particular case of fuzzy logic, we can find that most of the works in the literature for monitoring are based on the simplest form of fuzzy logic [1][2][3], which is called type-1, like the works that can be reviewed in [4][5][6][7][8][9][10][11][12]. More recently, type-2 has also been considered in this area, as a way to model uncertainty in a better fashion and achieve better results, as can be verified in [13,14].…”