Radiološki vjesnik

radiologija • radioterapija • nuklearna medicina

Challenges of introducing artificial intelligence into radiological clinical practice



We hope you have received a fair amount of all the positives when implementing Artificial Intelligence (AI) into the radiological service reading this Special Edition of our Journal. Now, in this paper, we will focus on the possible challenges and ethical concerns when implementing AI into radiology services.

Since this is not a strictly scientific Journal, and for practical reasons, I will list findings and discuss them at the same time.

We need to mention technical challenges when implementing AI into radiology practice for a start, but we will not spend much time on the discussion here, since we can overcome most of the challenges by increasing the funding to the project. Like discussed by Shaikh et al. (2020), one needs to focus on workflow integration, data transfer, management and deployment, user experience, and methodological issues. Of course, we do need to address the safety concerns and physical safety to the subjects. Since every radiology department is already heavily loaded with data, I will practically assume that most of the security and safety features are already implemented.


Hrvatsko društvo
radiološke tehnologije

Mlinarska cesta 38
10000 Zagreb