Back to all blog posts
Dr. İlknur Türkmen is the founder pathologist and director of Memorial Pathology. The pathology department serves 12 hospitals of Memorial Healthcare Group in six cities in Turkey. Memorial Healthcare Group also serves patients worldwide, in over 150 countries.
Memorial Healthcare Group has a grand vision of digitalization, showing it with a fully digital pathology infrastructure in a private hospital setting. They get around 70,000 cases per year, including oncology, pathology, and transplant cases, and operate with a team of over 20 pathologists, who work with different subspecialties.
“After being digitized, it was the right time for the routine use of the clinical AI. After evaluating the probabilities of which AI company to select, we decided on Aiforia for breast, prostate, and lung PD-L1 cases,” Dr. Türkmen comments. We interviewed Dr. Türkmen about their clinical implementation journey and experiences with Aiforia. Continue reading to learn more or watch the video at the end of this article.
Why implement AI in a clinical setting?
Dr. Türkmen explains that while they could continue working without AI, it would be more difficult, and the pathologists would have to work longer hours. AI will enable them to work smarter and assist with the shortage of pathologists. “We should use AI in our digital infrastructure for efficiency and also for accuracy, because there's a shortage of pathologists globally as well as in our country, and especially the specialized pathologists are so difficult to find in all cases,” Dr. Türkmen describes.
The team at Memorial Pathology selected Aiforia because of its cloud-based, all-in-one platform, the possibility of using Aiforia® Create to develop their own AI models, and because of great support and communication.
“We wanted to use several AI applications for our routine workflow, so it was a good advantage that Aiforia could serve us with algorithms for different organ types in one setting. In choosing a partner, it was important for us that Aiforia’s solutions also enable us to develop AI models for research purposes. As a pioneer in Turkey, we have had to overcome some obstacles in implementing new technology. Aiforia’s team has been helpful in supporting and helping us tackle these challenges.”– Dr. Türkmen, Memorial Pathology
Dr. Türkmen’s tip for successful clinical implementation is to consider everything about integrations already at the start and involve all stakeholders, including IT, directors, and the pathology department, in the project on day one.
Memorial Pathology pioneers AI also in research
Already during the demo period, Dr. Türkmen’s team made a small AI application for detecting Helicobacter pylori with Warthin-Starry, using Aiforia® Create, Aiforia’s AI development tool. “It was so good to see that with a limited number of cases and limited time, we could make our own algorithm,” Dr. Türkmen comments.
The research side is also important for Memorial Pathology, as they aim to be a reference center and a pioneer in the country, collaborating with universities on AI development. Simultaneously with the clinical deployment, they have already started several research projects with the Istanbul Technical University.
“Using AI in our research will be in parallel to the clinical use: it will make it easier and decrease the time of the study. Another specific thing for Aiforia is that we can build other algorithms with the existing ones, making it easier and shortening the time for AI development.”
Digital pathology is teamwork
“We had good cooperation with the entire team, and in the interoperability case, we also had a good collaboration with the IT and pathology departments. Digital pathology is teamwork, and we saw that Aiforia is a good team worker,” Dr. Türkmen describes the collaboration.
She explains that Memorial and Aiforia are similar in that both are pioneers and want to advance digitalization in pathology. “These two common sides, I hope, will bring great projects together to build digital pathology for patients, especially cancer patients,” she concludes.
“With AI use and AI development, we are trying to reach a point where, for each organ for several disease cases, we can use AI for our help, to help the pathologist make the diagnosis in a shorter time and with increased accuracy.” - Dr. Türkmen, Memorial Pathology
Watch Dr. Türkmen’s interview:
Explore our other clinical articles:
- Guest blog: evolution of artificial intelligence in diagnostic pathology
- How can AI assist in detecting HER2-low breast cancer?
- Digital pathology deployment in Veneto: 8 steps to success