To search, Click below search items. |
| All Published Papers Search Service |
|
|
|
Title |
Precision in Action: COVID-19 Detection with Generalized Linear Classifier and Two Step-AS Algorithm |
|
Author |
Ahmed Hamza Osman, Hani Moetque Aljahdali, Sultan Menwer Altarrazi, and Ahmad A Alzahrani |
| Citation |
Vol. 24 No. 12 pp. 111-120 |
|
Abstract |
This research introduces a computer-aided intelligence model designed to automatically identify positive instances of COVID-19 for routine medical applications. The model, built on the Generalized Linear architecture, employs the TwoStep-AS clustering method with diverse filter relatives, abstraction, and weight-sharing properties to automatically identify distinctive features in chest X-ray images. Unlike the conventional transformational learning approach, our model underwent training both before and after clustering. The dataset was subjected to a compilation process that involved subdividing samples and categories into multiple sub-samples and subgroups. New cluster labels were then assigned to each cluster, treating each subject cluster as a distinct category. Discriminant features extracted from this process were used to train the Generalized Linear model, which was subsequently applied to classify instances. The TwoStep-AS clustering method underwent modification by pre-aggregating the dataset before employing the Generalized Linear model to identify COVID-19 cases from chest X-ray findings. Tests were conducted using the COVID-19 public radiology database guaranteed the correctness of the results. The suggested model demonstrated an impressive accuracy of 90.6%, establishing it as a highly efficient, cost-effective, and rapid intelligence tool for the detection of Coronavirus infections. |
|
Keywords |
Generalized Liner model, Covid-19, Two Step-AS, Clustering, X-ray images |
|
URL |
