Page 130 - AIH-1-3

Artificial Intelligence in Health Interpretability of deep models for COVID-19

(SPIRA-BM) and by Coordenação de Aperfeiçoamento References
de Pessoal de Nível Superior - Brasil (CAPES) - Finance 1. Who Director-General’s Opening Remarks at the Media
Code 001.
Briefing on Covid-19. World Health Organization; 2020.
Conflict of interest Available from: https://www.who.int/director-general/
speeches/detail/who-director-general-s-opening-remarks-
The authors declare that they have no competing interests. at-the-media-briefing-on-covid-19-11-march-2020 [Last
accessed on 2024 Jul 19].
Author contributions
2. Brinati D, Campagner A, Ferrari D, Locatelli M, Banfi G,
Conceptualization: Arnaldo Candido Junior, Marcelo Finger Cabitza F. Detection of COVID-19 infection from routine
Investigation: Daniel Peixoto Pinto da Silva, Edresson blood exams with machine learning: A feasibility study.
Casanova, Arnaldo Candido Junior J Med Syst. 2020;44(8):135.
Methodology: Daniel Peixoto Pinto da Silva, Lucas Rafael doi: 10.1007/s10916-020-01597-4
Stefanel Gris, Flaviane Romani Fernandes Svartman, 3. Zoabi Y, Deri-Rozov S, Shomron N. Machine learning-based
Beatriz Raposo de Medeiros, Marcus Vinícius Moreira prediction of COVID-19 diagnosis based on symptoms. NPJ
Martins, Larissa Cristina Berti Digit Med. 2021;4(1):3.
Writing – original draft: Daniel Peixoto Pinto da Silva,
Arnaldo Candido Junior, Flaviane Romani Fernandes doi: 10.1038/s41746-020-00372-6
Svartman, Beatriz Raposo de Medeiros, Marcus 4. Ozturk T, Talo M, Yildirim EA, Baloglu UB, Yildirim O,
Vinícius Moreira Martins, Larissa Cristina Berti Acharya UR. Automated detection of COVID-19 cases
Writing – review & editing: Marcelo Matheus Gauy, using deep neural networks with x-ray images. Comput Biol
Arnaldo Candido Junior, Sandra Maria Aluísio, João Med. 2020;121:103792.
Paulo Teixeira, Marcelo Finger doi: 10.1016/j.compbiomed.2020.103792
Ethics approval and consent to participate 5. Acar E, Şahin E, Yılmaz İ. Improving effectiveness of different
deep learning-based models for detecting COVID-19 from
The research described in the paper was developed within computed tomography (CT) images. Neural Comput Appl.
the scope of the SPIRA Project (System for the Early 2021;33:17589-17609.
Detection of Respiratory Insufficiency via Audio), which doi: 10.1007/s00521-021-06344-5
was approved by the Research Ethics Committee (IRB)
of the Hospital das Clínicas da Faculdade de Medicina 6. Han J, Brown C, Chauhan J, et al. Exploring Automatic
da Universidade de São Paulo (HCFM/USP), Report COVID-19 Diagnosis via Voice and Symptoms from
3.988.088, approved on April 24, 2020. The report states that Crowdsourced Data. In: IEEE International Conference on
Acoustics, Speech and Signal Processing. p. 8328-8332.
this research does not require signed informed consent, as
data collection involves voice assessment, and participants doi: 10.1109/ICASSP39728.2021.9414576
consent to participate by recording their acceptance on the 7. Brown C, Chauhan J, Grammenos A, et al. Exploring
equipment (cell phone) used in the study. Automatic Diagnosis of COVID-19 from Crowdsourced
th
Respiratory Sound Data. In: Proceedings of the 26 ACM
Consent for publication SIGKDD International Conference on Knowledge Discovery
Due to the pandemic, the IRB of the Hospital das Clínicas and Data Mining, KDD; 2020. p. 3474-3484.
authorized us to collect patients’ agreement to participate doi: 10.1145/3394486.3412865
in the form of a recorded acceptance only. All participants 8. Aluísio SM, Camargo Neto AC, Casanova E, et al. Detecting
expressed their agreement in a recorded audio. Respiratory Insufficiency via Voice Analysis: The SPIRA
Project. In: Practical Machine Learning for Developing
Availability of data Countries on the Tenth International Conference on Learning
The audio data can be found at https://github.com/SPIRA- Representations; 2022.
COVID19/SPIRA-ACL2021/tree/master 9. Casanova E, Gris L, Camargo A, et al. Deep learning against
COVID-19: Respiratory insufficiency detection in Brazilian
Further disclosure Portuguese speech. In: Findings of the Association for
This paper has been uploaded to Arxiv at: https:// Computational Linguistics: ACL-IJCNLP; 2021. p. 625-633.
arxiv.org/pdf/2211.14372.pdf. The code for the models doi: 10.18653/v1/2021.findings-acl.55
can be found at: https://github.com/danpeixoto/covid19- 10. Selvaraju RR, Cogswell M, Das A, Vedantam R, Parikh D,
interpretability-analysis. Batra D. Grad-cam: Visual Explanations from Deep

Volume 1 Issue 3 (2024) 124 doi: 10.36922/aih.2992