Detection of Antibiotic Constituent in Aspergillus flavus Using Quantum Convolutional Neural Network

Abstract

Treatment of influenza and its complications is a major challenge for healthcare systems. Pyrazine is one drug used in treating influenza. Aspergillic acid is major antibiotic constituent in pyrazine compounds mined from Aspergillus flavus' final stage. This stage of flavus is detected through color change forming a pale-yellow crystal structure. Detection of the same is complex and demands an experienced fraternity to continuously monitor the growth of fungus and identify its color change. However, researches proved that the task needs to be perfect and a tiny human error leads to a catastrophe in antibiotic creation. To avoid these flaws, druggists make a huge investment on costly equipment for accurate detection. To overcome these drawbacks, this article proposes a hybrid quantum convolutional neural network that predicts various stages of the fungus from the microscope's sample. To train the network, about 47,000 samples were poised under typical lab settings. The proposed system was tested in usual conditions and positively isolated the mature samples with 96% efficiency.