AI-based journal screening for patient safe MR-examinations

Important results from the project

** Denna text är maskinöversatt ** Detecting implants prior to an MRI scan is important to prevent potential harm to the individual, but also to the device and its associated components. Today, this part of the process is ensured through very time-consuming work by healthcare professionals who review medical records for the patient. By automating parts of this work, it will be possible to save time but also increase the level of security as the human factor is not decisive for the discovery and identification of contexts in journal text that indicate the presence of medical implants.

Expected long term effects

** Denna text är maskinöversatt ** Overall, this type of digital decision support is something we see healthcare providers today very interested in investing in from a business perspective, it´s step one in taking the idea forward for development and commercialization. What has not yet been investigated is where in the process the decision support should be triggered and in what way the results should be presented to the users. In order to better visualize the solution itself, that design needs to be designed, evaluated and clarified.

Approach and implementation

** Denna text är maskinöversatt ** Future steps that remain to be investigated in order to further investigate the business potential are the technical aspects regarding where in the architecture of the journal systems the algorithms are to be handled. Today, for the current record system, the hypothesis has been the adapter for the clinical decision support system that is already in place today. It is desirable to set a more generic model around this, considering scalability and adaptation to different suppliers´ record systems.