COVID-19 Semantic Pneumonia Segmentation and Classification Using Artificial Intelligence

<table class="algorithm-group"><tr><td><table class="algorithm" id="alg3"><tr><td>(i)</td><td><b>Input</b>: Net, Percentage of Pruning (PP), maximum pruning (MP)</td></tr><tr><td> </td><td>Start Procedure</td></tr><tr><td> </td><td>Train and assess Net on B</td></tr><tr><td> </td><td><b>While</b> PP ≤ MP</td></tr><tr><td>(1)</td><td> Determine the number of filters in each hidden layer.</td></tr><tr><td>(2)</td><td> Recognize and eliminate percentage of filters in each hidden layer with the largest APoZ.</td></tr><tr><td>(3)</td><td> Retrain and assess the docked model on Net and choose the best-pruning weights.</td></tr><tr><td>(4)</td><td> PP++.</td></tr><tr><td>(5)</td><td>End While</td></tr><tr><td>(6)</td><td><b>Return</b> MP + 1</td></tr></table></td></tr></table>

Contrast Media & Molecular Imaging

alg3

Algorithm 3

Algorithm 3: COVID-19 Semantic Pneumonia Segmentation and Classification Using Artificial Intelligence 