An AutomAtic cough clAssificAtion Ai tool for PneumoniA PAthology Detection bAseD mAchine leArning AnD DeeP leArning moDels

Abstract

The coronavirus appeared in late 2019, which is represented by a large family of viruses that affect the respiratory system, where its symptoms vary from one infection to another: difficulty breathing,fever and exhaustion, the majority of patients suffer from dry cough. These viruses are easily transmitted from one person to another unlike other viruses, which led to paralysis of the world inits various sectors. With the advancement of the world and the development of technology in recent years, the field ofautomated media and artificial intelligence has witnessed great progress, as it has contributed to solving many problems in various fields, the most important of which are health and generalmedicine. Etc., with the advent of this deadly epidemic, we have been using deep and machinelearning to reduce and prevent the spread of coronavirus. That is why we did this research, which classifies cough using deep learning into a positive cough or a negative cough for coronavirus through a cough recorded on a smartphone. This study helps us reduce the spread of this deadly epidemic. After several tests and attempts to use models ANN, we reached a result of 97 %Le coronavirus est apparu fin 2019, qui est repr ́esent ́e par une grande famille de virus qui infectent le syst`eme respiratoire, o`u ses symptˆomes varient d’une infection `a l’autre : difficult ́e `a respirer, fi`evre et ́epuisement, la majorit ́e des patients souffrent de toux s`eche, et ces virus se transmettent facilement d’une personne `a une autre contrairement `a d’autres virus, ce qui a conduit `a la paralysie du monde dans ses diff ́erents secteurs. Avec l’avancement du monde et le d ́eveloppement de la technologie ces derni`eres ann ́ees, le domaine des m ́edias automatis ́es et de l’intelligence artificielle a connu de grands progr`es, car il a contribu ́e `a r ́esoudre de nombreux probl`emes dans divers domaines, dont les plus importants sont la sant ́e et la m ́edecine g ́en ́erale ... Etc., avec l’av`enement de cette ́epid ́emie mortelle, nous avons utilis ́e l’apprentissage profond et automatique pour r ́eduire et pr ́evenir la propagation du coronavirus. C’est pourquoi nous avons fait cette recherche qui classe la toux `a l’aide de l’apprentissage profond en toux positive ou en toux n ́egative du coronavirus par le biais d’une toux enregistr ́ee sur un smartphone, car cette ́etude nous aide `a r ́eduire la