With the spread of the COVID-19 pandemic, the scientific community took prompt action, first of all to sequence the virus and, therefore, to seek adequate treatments and vaccines, now finally available, for the prevention of the disease. In fact, the sequence of the new virus was already published by the Chinese at the beginning of the pandemic and for this reason the diagnosis immediately appeared to be easy to pronounce with the technique of nasopharyngeal swabs, serological and antigenic tests, chest radiographs, all methods already well known. to the scientific community. Without sequencing, diagnosis would have been more difficult. In fact, in most of the pandemics of the past the pathogen was unknown while in the case of COVID- 19, the sequencing of the virus made it possible to start from a very specific genetic identity. This then allowed the rapid development of vaccines with new technologies such as mRNA, or messenger RNA. For this reason, vaccination diagnosis and prevention are closely l inked. However, what seems less developed at present are methods for early diagnosis of the disease which would be useful especially when it is becoming more complicated towards interstitial pneumonia which is the main cause of ICU admissions and deaths. The aim of this work is to show how there are technologies typically used in electronics for the acquisition and processing of data, especially images, with particular reference to artificial intelligence (AI) algorithms, also known as Deep Learning (DL) and Machine Learning. (ML), which could allow a very early diagnosis of the onset of COVID-19 interstitial pneumonia and more. For this purpose, at least for a first screening, the use of medium -capacity smartphones may also be sufficient, without the need to resort to expensive medical equipment and diagnostic tests that are generally prohibitive for waiting times and very onerous. Despite the awareness that more and more scientists and research groups are using AI tools for the medical diagnosis of COVID-19 and beyond, the aim of this work is not to present a critical and / or descriptive analysis of the activities in progress by the scientific community but to present the engineering tools underlying the activities of the various research groups with AI tools. Therefore, we will first see how it is possible from a medical point of view to diagnose the onset of interstitial pneumonia early; subsequently we will make an overview of the algorithmic me thods useful for this purpose; finally, we will see how a smartphone can be useful as a tool for early diagnosis

Con il dilagare della pandemia da COVID -19 la comunità scientifica si è attivata tempestivamente innanzitutto per sequenziare il virus e, quindi, per cercare cure adeguate e vaccini, ora finalmente disponibili, per la prevenzione della malattia. Infatti, la sequenza del nuovo virus fu pubblicata già dai cinesi ad inizio pandemia e per questo la diagnosi è apparsa subito di facile pronuncia con la tecnica dei tamponi naso- faringei, dei test sierologici e antigenici, delle radiografie toraciche, metodologie tutte già ben note alla comunità scientifica. In assenza di sequenziamento la diagnosi sarebbe stata più difficile. Infatti, nella gran parte delle pandemie del passato l’agente patogeno era sconosciuto mentre nel caso del COVID -19 il sequenziamento del virus ha consentito di partire da una identità genetica ben precisa. Ciò ha permesso poi il rapido sviluppo dei vaccini con nuove tecnologie come quella del mRNA, o RNA messaggero. P er questo motivo diagnosi e prevenzione vaccinale sono strettamente legate. Tuttavia, ciò che sembra meno sviluppato attualmente sono metodi per la diagnosi precoce della malattia che sarebbero utili soprattutto per prevenire o curare ai primi sintomi la polmonite interstiziale che è la causa principale dei ricoveri in terapia intensiva e dei decessi. Obiettivo di questo lavoro è mostrare come ci siano tecnologie tipicamente utilizzate in elettronica per l’acquisizione ed elaborazione di dati, specialmente di immagini, con particolare riferimento ad algoritmi di intelligenza artificiale (IA), nota anche come Deep Learning (DL) e Machine Learning (ML), che potrebbero permettere una diagnosi assai precoce dell’insorgere della polmonite interstiziale da COVID-19 e non solo. A tale scopo, almeno per un primo screening, potrebbe essere anche sufficiente l’utilizzo di smartphone di media capacità, senza necessità di ricorrere a costosa strumentazione medica ed esami diagnostici generalmente proibitivi per tempi di attesa e molto onerosi. Pur nella consapevolezza che sono sempre più gli scienziati ed i gruppi di ricerca che utilizzano strumenti di IA per la diagnosi medica del COVID-19 e non solo, l’obiettivo del presente lavoro non è quello di presentare un’analisi critica e/o descrittiva delle attività in corso di svolgimento da parte della comunità scientifica bensì quello di presentare gli strumenti ingegneristici alla base delle attività dei vari gruppi di ricerca con strumenti di IA. Pertanto, vedremo innanzitut to come sia possibile diagnosticare precocemente l’insorgere della polmonite interstiziale; successivamente faremo una panoramica comparativa, qualitativa e quantitativa, dei metodi algoritmici utili allo scopo; infine, vedremo come uno smartphone possa es sere utile come strumento per la diagnosi precoce

Intelligenza artificiale e diagnosi precoce del COVID-19 / Giorgio, Agostino. - In: LA COMUNICAZIONE. - ISSN 1590-864X. - ELETTRONICO. - 63:(2020).

Intelligenza artificiale e diagnosi precoce del COVID-19

Giorgio Agostino
2020-01-01

Abstract

With the spread of the COVID-19 pandemic, the scientific community took prompt action, first of all to sequence the virus and, therefore, to seek adequate treatments and vaccines, now finally available, for the prevention of the disease. In fact, the sequence of the new virus was already published by the Chinese at the beginning of the pandemic and for this reason the diagnosis immediately appeared to be easy to pronounce with the technique of nasopharyngeal swabs, serological and antigenic tests, chest radiographs, all methods already well known. to the scientific community. Without sequencing, diagnosis would have been more difficult. In fact, in most of the pandemics of the past the pathogen was unknown while in the case of COVID- 19, the sequencing of the virus made it possible to start from a very specific genetic identity. This then allowed the rapid development of vaccines with new technologies such as mRNA, or messenger RNA. For this reason, vaccination diagnosis and prevention are closely l inked. However, what seems less developed at present are methods for early diagnosis of the disease which would be useful especially when it is becoming more complicated towards interstitial pneumonia which is the main cause of ICU admissions and deaths. The aim of this work is to show how there are technologies typically used in electronics for the acquisition and processing of data, especially images, with particular reference to artificial intelligence (AI) algorithms, also known as Deep Learning (DL) and Machine Learning. (ML), which could allow a very early diagnosis of the onset of COVID-19 interstitial pneumonia and more. For this purpose, at least for a first screening, the use of medium -capacity smartphones may also be sufficient, without the need to resort to expensive medical equipment and diagnostic tests that are generally prohibitive for waiting times and very onerous. Despite the awareness that more and more scientists and research groups are using AI tools for the medical diagnosis of COVID-19 and beyond, the aim of this work is not to present a critical and / or descriptive analysis of the activities in progress by the scientific community but to present the engineering tools underlying the activities of the various research groups with AI tools. Therefore, we will first see how it is possible from a medical point of view to diagnose the onset of interstitial pneumonia early; subsequently we will make an overview of the algorithmic me thods useful for this purpose; finally, we will see how a smartphone can be useful as a tool for early diagnosis
2020
Intelligenza artificiale e diagnosi precoce del COVID-19 / Giorgio, Agostino. - In: LA COMUNICAZIONE. - ISSN 1590-864X. - ELETTRONICO. - 63:(2020).
File in questo prodotto:
Non ci sono file associati a questo prodotto.

I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.

Utilizza questo identificativo per citare o creare un link a questo documento: https://hdl.handle.net/11589/219298
Citazioni
  • Scopus ND
  • ???jsp.display-item.citation.isi??? ND
social impact