Skip to primary content

Implementation of clinical artificial intelligence in healthcare in Denmark

In 2023, the Danish Robusthedskommission recommended the use of e.g. artificial intelligence (AI) to help free up time for hospital staff to perform essential duties. Successful use of AI algorithms could provide patients with faster answers when it comes to medical imaging, in turn leading to faster relevant treatment or discharge. The time saved by staff in reviewing images could then be redirected to patient care.

Project period

Start: 1 March 2024

Aim

Based on the recommendations made by Robusthedskommissionen and the available research within the field of artificial intelligence and diagnostic imaging, this project seeks to identify where and how AI technology can lead to improvements for both staff and patients in the Emergency Department at Odense University Hospital (OUH).

The project will do this through three sub-projects that will:

  1. Investigate facilitating and hindering factors in the implementation process of artificial intelligence for medical imaging through a literature review.
  2. Evaluate an artificial intelligence algorithm approved according to Regulatory EU standards (CE marking), which can identify bone fractures in X-rays and test it on already available X-ray images.
  3. Assess the potential of using such an algorithm in the Emergency Department at OUH. This assessment is carried out via the AI-CARE (Artificial Intelligence for Coordinated Allocation and Rapid Evaluation) project, which consists of two phases, where each phase represents an existing workflow/guideline. A quality assurance of both workflows is carried out, after which the results are compared to shed light on any differences in fracture detection and the patients' length of stay. Radiologists and musculoskeletal specialists will continue to review all X-rays the next business day in both phases and contact patients if there are any changes.
    1. Phase 1: Evaluation of the current standard workflow in the Emergency Department for patients with suspected bone fractures.
    2. Phase 2: Evaluation of a workflow where an already approved AI algorithm is included as a support tool for the emergency department staff. This workflow may allow patients to be sent home immediately after the X-ray examination. The goal is to investigate whether this can shorten the patients' length of stay while maintaining diagnostic quality.

The project can thus help to illuminate solutions to key challenges in the healthcare system of the future.

Participants

UNIFY - Radiology Research and Innovation Unit

Department of Radiology, OUH

Sabine Morris Delhez

Sabine Morris Delhez

PhD Student

Odense University Hospital, Department of Radiology


APPFWU02V