PhD Scientific Days 2023

Clinical Medicine - Posters A

Prognostic power of deep-learning automated coronary artery calcium score and quantitative pneumonia burden in patients hospitalized with COVID-19

Előadó neve

Szabó, István Viktor

Neptun code

XGJRJE

Előadó munkahelye

Semmelweis University Medical Imaging Centre

Előadó telefonszáma

+36202839347

Előadó e-mail címe

istvan.szabo97@gmail.com

Az előadás címe

Prognostic power of deep-learning automated coronary artery calcium score and quantitative pneumonia burden in patients hospitalized with COVID-19

Szerző(k) neve és munkahelye

Chiara Nardocci1, Judit Simon1,2 Hugo Aerts3, Roman Zeleznik4, Michael Lu5,
Julia Karady2,6, Marton Kolossvary2, Bernard Cosyns7, Bettina Budai1, Viktor Gál, Mihály
Radványi, Dávid Prait, Damini Dey8, Piotr Slomka, Veronika Müller9, Béla Merkely2, Pál
Maurovich-Horvat1,2

1 Semmelweis University Medical Imaging Centre, Budapest
2 MTA-SE Cardiovascular Imaging Research Group, Heart and Vascular Center, Semmelweis University, Budapest
3 Departments of Radiation Oncology and Radiology, Brigham and Women’s Hospital, Dana-Farber Cancer Institute, Harvard Medical School, Boston, United States of America
4 Artificial Intelligence in Medicine (AIM) Program, Mass General Brigham, Harvard Medical School, Boston
5 School of Public Health at the University of California Berkeley, in Berkeley, California
6 Cardiovascular Imaging Research Center, Harvard Medical School - Massachusetts General Hospital
7 Department of Cardiology, University of Brussels, Brussels
8 Cedars-Sinai Medical Center, Los Angeles
9 Department of Pulmonology, Semmelweis University, Budapest

Bemutatás módja

Poszter

Szekció

Clinical Medicine - Posters A

Language of the presentation

Hungarian

Preferred session

Clinical Medicine

Összefoglaló szövege

Introduction: Patient outcomes after SARS-CoV2 infection are associated with cardiovascular morbidity and pulmonary involvement both of which can be estimated using deep learning (DL) analysis of computed tomography.
Aims: To determine the predictive power of DL-estimated coronary artery calcium score (CACS) and quantitative pneumonia burden to predict SARS-CoV2 associated in-hospital mortality.
Method: Single-center retrospective analysis of 1,050 patients admitted to the emergency department with polymerase chain reaction (PCR)-confirmed wild type SARS-CoV2 infection between September 1st and December 31st in 2020. All patients underwent chest CT for the assessment of SARS-CoV2 infection related pneumonia. CACS and pneumonia burden were quantified using DL algorithms. We divided the patients into six CACS categories. Primary outcome was defined as in-hospital mortality.
Results: Chest CTs of 300 patients were used for the training and tuning of the DL-based solution of CACS measurement and data of 388 patients belonged to the test cohort. In total, 74 patients died in the hospital. There were significant differences in in-hospital mortalities among the CACS strata (mortality rate was 8.2% in patients with zero CACS vs 27.3% in patients with CACS >1,000). The CT-based model (pneumonia burden + CACS) had an excellent predictive power, which was further improved by adding the clinical parameters to the model (AUC: 0.77 [95%CI: 0.71-0.83] vs 0.85 [95%CI: 0.81-0.90], p<0.001). In the multivariate analysis, adjusting for age only pneumonia burden remained an independent predictor of in-hospital death (OR: 1.05 [95%CI: 1.04-1.07], p<0.001).
Conclusion: CACS has no additional predictive role over age in patients hospitalized with COVID-19, however pneumonia is an independent predictor of in-hospital mortality.
Funding: No funding.

University and Doctoral School

Semmelweis University, Doctoral School of Theoretical and Translational Medicine

Supervisor

Dr. Maurovich-Horvat Pál

Publication of my abstract

I do not give consent to the publication of my abstract on the website of the congress.

Kind

Szabad

Status

elfogadva

Accepted presentation method

poszter

Előadás fájl jóváhagyás

nem rendelkezett róla

Előadó

5715

Start

12:12

End

12:17

Authors (legacy)

Chiara Nardocci1, Judit Simon1,2 Hugo Aerts3, Roman Zeleznik4, Michael Lu5,
Julia Karady2,6, Marton Kolossvary2, Bernard Cosyns7, Bettina Budai1, Viktor Gál, Mihály
Radványi, Dávid Prait, Damini Dey8, Piotr Slomka, Veronika Müller9, Béla Merkely2, Pál
Maurovich-Horvat1,2

1 Semmelweis University Medical Imaging Centre, Budapest
2 MTA-SE Cardiovascular Imaging Research Group, Heart and Vascular Center, Semmelweis University, Budapest
3 Departments of Radiation Oncology and Radiology, Brigham and Women’s Hospital, Dana-Farber Cancer Institute, Harvard Medical School, Boston, United States of America
4 Artificial Intelligence in Medicine (AIM) Program, Mass General Brigham, Harvard Medical School, Boston
5 School of Public Health at the University of California Berkeley, in Berkeley, California
6 Cardiovascular Imaging Research Center, Harvard Medical School - Massachusetts General Hospital
7 Department of Cardiology, University of Brussels, Brussels
8 Cedars-Sinai Medical Center, Los Angeles
9 Department of Pulmonology, Semmelweis University, Budapest