Dear Trillium readers,
welcome back to our new edition of Trillium Pathology, the journal for pathologists interested in new trends and technologies in pathology.
Artificial intelligence (AI) is becoming increasingly popular in pathology, especially in clinical research. The tracks for digital and computational pathology at the DGP, ECP and ECDP conferences become more comprehensive every year and the number of publications in that field is constantly growing. However, the question remains as to how we can translate the new technology into efficient clinical practice.
The technical gap certainly plays a significant role: A network that has grown over time and the IT infrastructure of a hospital, optimized for a discipline that has largely been analog, is suddenly confronted with gigapixel sized digital images that need to be stored and streamed efficiently to be available inside and outside the hospital. Digital pathology is a requirement for the use of AI, and the change management needed to digitize pathology labs is still challenging. A digital lab comes with a risk of more technical errors – at least in the beginning.
On the other hand, the benefits of using AI, although evidently reported and illustrated in multiple studies, need to be experienced first hand. In this sense, AI in pathology can be compared with something that you might curiously embrace in the first place. But once you use it, you wonder how you have even worked without it before.
A key point for use of AI in clinical practice certainly is its robustness: It has to work despite of small perturbations or changes in the data distribution as they occur across different labs, different scanners and other factors. We will shed more light on this in the current issue.
Another important requirement for AI is its spread in the domain. If not only histology, but also cytology and molecular pathology can benefit from AI, the digital transition is even more rewarding. In this issue, we show an example of AI for cytology in Leukaemia detection, and we highlight currently available “AIs” for a wide range of pathology applications.
In research spatial transcriptomics and proteomics have recently gained more attention, with the advent of new high-resolution technologies. We explore the opportunities for AI in these two fields. Digitization in pathology also impacts biobanking, and we are proud to showcase this process in our current issue.
Finally, we present a short overview of the newly established nonprofit organization EMPAIA International e. V., that pioneers in bringing together various stakeholders in digital pathology such as medical doctors, pathologists, researchers, industry, and patient organizations, in order to foster digital transition, standardization and AI in pathology.
We hope you enjoy this year’s edition of Trillium Pathology and encourage you to contribute to our next issue in 2025 with experience reports, own opinions, original research, mini-reviews or other formats – simply contact us!
Author
