Clin Chem Lab Med October 2020 PMID 33079698

Development, evaluation, and validation of machine learning models for COVID-19 detection based on routine blood tests.

Federico Cabitza, Andrea Campagner, Davide Ferrari, Chiara Di Resta, Daniele Ceriotti, Eleonora Sabetta, Alessandra Colombini, Elena De Vecchi, Giuseppe Banfi, Massimo Locatelli, Anna Carobene

Mentions

17
Peak: 8 in Nov 2020
across 2 sources since Oct 2020

Extended reach

69,634
potential audience

Total likes

8

Video views

443

Mentions per month

Oct 2020 – May 2023

Where it's discussed

  • X / Twitter13 posts
  • YouTube4 videos

AI synopsis

What people are saying

The overall discussions reflect cautious optimism, with phrases like 'potential for quick COVID detection' and mentions of the models' accuracy scores indicating interest in practical applications. Words such as 'evaluation,' 'validation,' and references to the ongoing peer review process suggest a focus on preliminary but promising findings, shaping a moderate overall tone.

AgreementModerate agreement

Most discussions acknowledge the potential of the machine learning models for COVID-19 detection, though some mention that the research is still preliminary or unreviewed.

InterestNeutral interest

The posts show moderate interest, with some curiosity about the approach, but few engage deeply with technical details or broader implications.

EngagementModerate engagement

Several users refer to the methodology and practical applications, but interactions are mostly surface-level, lacking extensive debate or critique.

ImpactModerate impact

There is general recognition that this research could influence future testing strategies, especially in resource-limited settings, but it remains somewhat speculative.

Abstract in brief

  • The study explores using machine learning models trained on routine blood test data as a faster and cheaper alternative to the RT-PCR test for detecting COVID-19, addressing issues like long test turnaround times and reagent shortages.
  • Researchers used data from 1624 patients, including blood test results, symptoms, and demographic info, to develop and validate five ML models, which showed good accuracy in distinguishing COVID-19 positive cases with AUROC scores around 0.83-0.90.
  • The results suggest that machine learning applied to blood tests could be a practical tool for quick, cost-effective COVID-19 detection, especially in resource-limited settings such as developing countries or during outbreaks.]

Top posts

Most liked on X and Bluesky
Fabio Vanorio
@FabioVanorio · X

Development, evaluation, and validation of machine learning models for COVID-19 detection based on routine blood tests https://t.co/8h7fuCDHkU

Nov 16, 2020 · 1 like · 4 shares · 867 followers View post →
medRxiv
@medrxivpreprint · X

Development, evaluation, and validation of machine learning models for COVID-19 detection based on routine blood tests https://t.co/iqprA3LSam #medRxiv

Oct 4, 2020 · 1 share · 17,448 followers View post →
Machine Learning Bot
@ML_Tweet_Bot · X

RT @safalabolo: Development, evaluation, and validation of machine learning models for COVID-19 detection based on routine blood tests http…

Nov 16, 2020 · 2 shares · 10,165 followers View post →
Machine Learning Bot
@ML_Tweet_Bot · X

RT @FabioVanorio: Development, evaluation, and validation of machine learning models for COVID-19 detection based on routine blood tests ht…

Nov 16, 2020 · 4 shares · 10,165 followers View post →
questo non è un uovo 
@PMO_W · X

RT @FabioVanorio: Development, evaluation, and validation of machine learning models for COVID-19 detection based on routine blood tests ht…

Nov 16, 2020 · 4 shares · 9,476 followers View post →
Vitruviano
@Il_Vitruviano · X

RT @FabioVanorio: Development, evaluation, and validation of machine learning models for COVID-19 detection based on routine blood tests ht…

Nov 16, 2020 · 4 shares · 6,819 followers View post →

All mentions

むらま
@BushiMurasaki · X

ノシ Development, evaluation, and validation of machine learning models for COVID-19 detection based on routine blood tests https://t.co/1b5iDH43zx

Nov 21, 2020 · 64 followers View post →
questo non è un uovo 
@PMO_W · X

RT @FabioVanorio: Development, evaluation, and validation of machine learning models for COVID-19 detection based on routine blood tests ht…

Nov 16, 2020 · 4 shares · 9,476 followers View post →
S. S. Xman
@SSXman2 · X

RT @FabioVanorio: Development, evaluation, and validation of machine learning models for COVID-19 detection based on routine blood tests ht…

Nov 16, 2020 · 4 shares · 4,720 followers View post →
S. S. Xman
@SSXman2 · X

RT @safalabolo: Development, evaluation, and validation of machine learning models for COVID-19 detection based on routine blood tests http…

Nov 16, 2020 · 2 shares · 4,720 followers View post →
Vitruviano
@Il_Vitruviano · X

RT @FabioVanorio: Development, evaluation, and validation of machine learning models for COVID-19 detection based on routine blood tests ht…

Nov 16, 2020 · 4 shares · 6,819 followers View post →
Machine Learning Bot
@ML_Tweet_Bot · X

RT @safalabolo: Development, evaluation, and validation of machine learning models for COVID-19 detection based on routine blood tests http…

Nov 16, 2020 · 2 shares · 10,165 followers View post →
Machine Learning Bot
@ML_Tweet_Bot · X

RT @FabioVanorio: Development, evaluation, and validation of machine learning models for COVID-19 detection based on routine blood tests ht…

Nov 16, 2020 · 4 shares · 10,165 followers View post →
Fabio Vanorio
@FabioVanorio · X

Development, evaluation, and validation of machine learning models for COVID-19 detection based on routine blood tests https://t.co/8h7fuCDHkU

Nov 16, 2020 · 1 like · 4 shares · 867 followers View post →
Outbreak Science
@outbreaksci · X

Update: Development, evaluation, and validation of machine learning models for COVID-19 detection based on routine blood tests https://t.co/ZlIli1Jyz0 read/write/request review https://t.co/RnurTL3DyM #coronavirus #COVID19

Oct 21, 2020 · 1,372 followers View post →
あきた
@kachimakeakita · X

(This hasn’t been peer reviewed yet.) [1] Development, evaluation, and validation of machine learning models for COVID-19 detection based on routine blood tests. By Cabitza Federico et al. medRxiv preprint doi: https://t.co/lLXRr0LoCh. this version posted October 4, 2020.

Oct 5, 2020 · 4 followers View post →
MBE
@MBELTD · X

Development, evaluation, and validation of machine learning models for COVID-19 detection based on routine blood tests: Background: The rRT PCR test, the current gold standard for the detection of coronavirus disease (COVID19), presents with known… https://t.co/sXICdq6Fjq

Oct 4, 2020 · 2,995 followers View post →
Chomes
@CHOMES102 · X

RT @medrxivpreprint: Development, evaluation, and validation of machine learning models for COVID-19 detection based on routine blood tests…

Oct 4, 2020 · 1 share · 326 followers View post →
medRxiv
@medrxivpreprint · X

Development, evaluation, and validation of machine learning models for COVID-19 detection based on routine blood tests https://t.co/iqprA3LSam #medRxiv

Oct 4, 2020 · 1 share · 17,448 followers View post →