Design of an AI Support for Diagnosis of Dyspneic Adults at Time of Triage in the Emergency Department

Ellen Tolestam Heyman; Awais Ashfaq; Ulf Ekelund; Mattias Ohlsson; Jonas Björk; Ardavan M. Khoshnood; Markus Lingman

Design of an AI Support for Diagnosis of Dyspneic Adults at Time of Triage in the Emergency Department

Ellen Tolestam Heyman, Awais Ashfaq, Ulf Ekelund, Mattias Ohlsson, Jonas Björk, Ardavan M. Khoshnood, Markus Lingman

Research output: Contribution to conference › Poster

44 Downloads (Pure)

Abstract

We created an AI support for diagnosis in dyspneic adults at time of triage in the emergency department.

Complete data from an entire regional health care system was analyzed, to find AI-derived, unknown, important diagnostic predictors. Most important were prior diagnoses of heart failure or COPD, daily smoking, atrial fibrillation/flutter, life difficulties and maternal care.

Sensitivity for AHF, eCOPD and pneumonia was 75%, 93%, and 54%, respectively, with a specificity above 75%.

Each patient visit received an individual graph with the AI´s underlying decision basis.

Original language	English
Publication status	Published - 2023 Sept
Event	European Emergency Medicine Congress 2023 - Barcelona, Spain Duration: 2023 Sept 17 → 2023 Sept 20

Conference

Conference	European Emergency Medicine Congress 2023
Country/Territory	Spain
City	Barcelona
Period	2023/09/17 → 2023/09/20

UN SDGs

This output contributes to the following UN Sustainable Development Goals (SDGs)

SDG 3 Good Health and Well-being

Subject classification (UKÄ)

Cardiology and Cardiovascular Disease

Free keywords

artificial intelligence
AI
Dyspnea

Access to Document

Eposter_DyspneaFinal published version, 1.09 MB

1 Doctoral Thesis (compilation)

Exploring Artificial Intelligence in Emergency Medicine for Predicting Disposition, Death, and Diagnosis
Tolestam Heyman, E., 2024, Lund: Lund University, Faculty of Medicine. 73 p.
Research output: Thesis › Doctoral Thesis (compilation)

Open Access
File

2 Finished

Resource Management in the Emergency Department by using Machine Learning
Tolestam Heyman, E. (Research student), Khoshnood, A. (Supervisor), Ekelund, U. (Assistant supervisor), Björk, J. (Assistant supervisor), Ohlsson, M. (Assistant supervisor), Lingman, M. (Assistant supervisor) & Dahlén Holmqvist, L. (Assistant supervisor)
2021/02/01 → 2024/11/21
Project: Dissertation
AIR Lund: AIR Lund - Artificially Intelligent use of Registers
Björk, J. (Researcher), Ohlsson, M. (Researcher), Melander, O. (Researcher), Ekelund, U. (Researcher), Larsson, S. (Researcher), Mattsson, T. (Researcher), Mokhtari, A. (Researcher), de Vries, K. (Researcher), Nilsson, A. (Researcher), Lingman, M. (Researcher), Pashami, S. (Researcher), Ottosson, F. (Researcher), Larsson, A. E. (Administrator), Henriksen, A. (Researcher), Högberg, C. (Research student), Olsson, M. (Research student), Nyström, A. (Research student), Olsson de Capretz, P. (Research student), Björkelund, A. (Researcher), Nilsson, A. (Researcher), Åkesson, A. (Research student), Nordberg, A. (Researcher), Ekström, M. (Researcher), Tolestam Heyman, E. (Research student), Lundager Forberg, J. (Researcher), Berg, J. (Research student), Johansson, A. (Research student), Amirahmadi, A. (Research student), Eriksson, T. (Research student), Khoshnood, A. (Researcher), Mäkinen, M. (Research student), Holmberg, P. (Researcher) & Nilsson, T. N. (Research student)
2018/07/01 → 2025/12/31
Project: Research

1 Supervision of PhD students

Resource management in the emergency department by using machine learning
Khoshnood, A. (First/primary/lead supervisor), Ekelund, U. (Second supervisor), Ohlsson, M. (Joint second supervisor), Björk, J. (Joint second supervisor), Lingman, M. (Joint second supervisor) & Dahlén Holmqvist, L. (Joint second supervisor)
2021 Feb 1 → 2024 Nov 21
Activity: Examination and supervision › Supervision of PhD students

Cite this

@conference{9ba953a5a33147039a33bcd45d86516e,

title = "Design of an AI Support for Diagnosis of Dyspneic Adults at Time of Triage in the Emergency Department",

abstract = "We created an AI support for diagnosis in dyspneic adults at time of triage in the emergency department. Complete data from an entire regional health care system was analyzed, to find AI-derived, unknown, important diagnostic predictors. Most important were prior diagnoses of heart failure or COPD, daily smoking, atrial fibrillation/flutter, life difficulties and maternal care. Sensitivity for AHF, eCOPD and pneumonia was 75\%, 93\%, and 54\%, respectively, with a specificity above 75\%. Each patient visit received an individual graph with the AI´s underlying decision basis.",

keywords = "artificial intelligence, AI, Dyspnea, Artificiell intelligens, AI, Dyspne",

author = "\{Tolestam Heyman\}, Ellen and Awais Ashfaq and Ulf Ekelund and Mattias Ohlsson and Jonas Bj{\"o}rk and Khoshnood, \{Ardavan M.\} and Markus Lingman",

year = "2023",

month = sep,

language = "English",

note = "European Emergency Medicine Congress 2023 ; Conference date: 17-09-2023 Through 20-09-2023",

}

TY - CONF

T1 - Design of an AI Support for Diagnosis of Dyspneic Adults at Time of Triage in the Emergency Department

AU - Tolestam Heyman, Ellen

AU - Ashfaq, Awais

AU - Ekelund, Ulf

AU - Ohlsson, Mattias

AU - Björk, Jonas

AU - Khoshnood, Ardavan M.

AU - Lingman, Markus

PY - 2023/9

Y1 - 2023/9

N2 - We created an AI support for diagnosis in dyspneic adults at time of triage in the emergency department. Complete data from an entire regional health care system was analyzed, to find AI-derived, unknown, important diagnostic predictors. Most important were prior diagnoses of heart failure or COPD, daily smoking, atrial fibrillation/flutter, life difficulties and maternal care. Sensitivity for AHF, eCOPD and pneumonia was 75%, 93%, and 54%, respectively, with a specificity above 75%. Each patient visit received an individual graph with the AI´s underlying decision basis.

AB - We created an AI support for diagnosis in dyspneic adults at time of triage in the emergency department. Complete data from an entire regional health care system was analyzed, to find AI-derived, unknown, important diagnostic predictors. Most important were prior diagnoses of heart failure or COPD, daily smoking, atrial fibrillation/flutter, life difficulties and maternal care. Sensitivity for AHF, eCOPD and pneumonia was 75%, 93%, and 54%, respectively, with a specificity above 75%. Each patient visit received an individual graph with the AI´s underlying decision basis.

KW - artificial intelligence

KW - AI

KW - Dyspnea

KW - Artificiell intelligens

KW - AI

KW - Dyspne

M3 - Poster

T2 - European Emergency Medicine Congress 2023

Y2 - 17 September 2023 through 20 September 2023

ER -

Design of an AI Support for Diagnosis of Dyspneic Adults at Time of Triage in the Emergency Department

Abstract

Conference

UN SDGs

Subject classification (UKÄ)

Free keywords

Access to Document

Fingerprint

Research output

Exploring Artificial Intelligence in Emergency Medicine for Predicting Disposition, Death, and Diagnosis

Projects

Resource Management in the Emergency Department by using Machine Learning

AIR Lund: AIR Lund - Artificially Intelligent use of Registers

Activities

Resource management in the emergency department by using machine learning

Cite this