Accurate computation: COVID-19 rRT-PCR positive test dataset using stages classification through textual big data...

Area of Science:

• Medical Informatics
• Computational Biology
• Epidemiology

Background:

• The COVID-19 pandemic highlighted the need for rapid diagnostic tools.
• Current standard testing (rRT-PCR) has limitations including long turnaround times and specialized laboratory requirements.
• Early identification and risk stratification of COVID-19 cases are crucial for effective public health response.

Purpose of the Study:

• To develop and evaluate a machine learning model for classifying COVID-19 patients based on textual clinical reports.
• To explore the feasibility of using text data mining and machine learning as an alternative to rRT-PCR.
• To improve the speed and accessibility of COVID-19 diagnosis and risk prediction.

Main Methods:

• Utilized an ensemble machine learning classifier for text classification.
• Employed Term Frequency-Inverse Document Frequency (TF/IDF) for feature extraction from clinical text data.
• Applied data mining techniques to analyze COVID-19 patient records.

Main Results:

• The machine learning model successfully classified textual clinical reports into distinct categories related to COVID-19.
• TF/IDF effectively extracted relevant features for classification and prediction.
• Demonstrated the potential of analyzing blood test data with machine learning for COVID-19 detection.

Conclusions:

• Machine learning techniques, combined with text data mining, show promise for classifying COVID-19 patients.
• This approach offers a viable alternative to rRT-PCR, potentially enabling faster and more accessible diagnosis.
• The study supports the feasibility of using computational methods for COVID-19 risk prediction and patient categorization.