MAS-AI: Model for ASsessment of Artificial Intelligence

The use of AI could affect the organisation of the treatment, patient safety, the effect of the treatments, the patient's experience of the treatment, inequality in health, finances, staff consumption and ethics.

But at the start of this project, there had been no models developed to evaluate the effect and consequences of using artificial intelligence, and decision-makers and health professionals are therefore left to themselves in the assessment of a given solution.

AIM

The aim of the project MAS-AI (Model for ASessment of Artificial Intelligence) was to develop a model which can help decision-makers to assess and designate AI technologies in image diagnostics. With a thorough model for evaluation, you get an overview of which AI technologies create value, so that you can choose them, and technologies with no or inappropriate effect can be selected from.

The model was developed based on the following steps:

• A review of existing guides, evaluations and assessments of the value of AI in the field of medical imaging
• Interviews with leading researchers within AI in Denmark
• Workshops with decision makers, patient organisations and researchers

With input from the above, the project developed the MAS-AI model, which can give administrative and health professional managers an overview of the consequences of using AI technologies.

RESULTS

The project has developed the model MAS-AI, which is an MTV-based tool that can support the introduction of new technology with artificial intelligence in the healthcare system.

The model was developed by an interdisciplinary group of experts, researchers and patient representatives, and it consists of three parts:

An early MAS-AI with four areas of evaluation

1. The health problem and the current use of technology
2. Technology
3. Ethical aspects
4. Legal aspects

A full MAS-AI with five other evaluation areas

1. Security
2. Clinical aspects
3. Economy
4. Organisation
5. Patient perspectives

as well as five process factors that are recommended to be considered during the evaluation process.

An evaluation using MAS-AI can support the decision-making process in relation to the use of artificial intelligence in healthcare and increase transparency for all parties involved.

Read the article Model for Assessing the value of Artificial Intelligence in medical imaging (MAS-AI) in the International Journal of Technology Assessment in Health Care.

PARTNERS

The project was carried out in collaboration with the Department of Radiology and the Department of Nuclear Medicine at OUH. In addition, a Canadian validation of MAS-AI is being carried out in collaboration with Theta Collaborative and Transform HF.

This project involved artificial intelligence and was therefore also anchored at the Centre for Clinical Artificial Intelligence (CAI-X).

EXTERNAL FUNDING

The project was financed by the pool for OUH's Competitive Funds.