PhD project with focus on digital microscopy and AI for diagnostics in resource-limited settings

Joar von Bahr holds a medical degree from Karolinska Institutet, where he graduated in January 2024. He recently had his ISP seminar at GPH, where he presented his research plans for the coming years. We got a few minutes of his time to dig deeper in to his PhD project.

Hello Joar, congratulations on a successful ISP seminar! Tell us a little about yourself and your background.

“I have long been interested in research focusing on the integration of new technologies into medicine and for a year during my medical studies I studied data science at KTH Royal Institute of Technology to achieve my aim of working with this.

For my medical thesis, I worked under the supervision of Professors Johan Lundin and Nina Linder within the MoMic-Project, which focuses on using artificial intelligence in screening for cancer and infectious diseases in resource-limited settings. After completing my thesis, I continued working on the project as a research assistant while finishing my medical degree. Since the MoMic-project is a collaboration between Karolinska Institutet, the University of Helsinki, Uppsala University and a hospital in Kenya, my doctoral project was initially registered at the University of Helsinki and this summer the project was transferred to Karolinska Institutet.”

What will your planned PhD be about?

“My PhD project focuses on the implementation of digital microscopy combined with artificial intelligence for diagnostics in resource-limited settings, specifically targeting cervical cancer and soil-transmitted helminths. We are collaborating with a primary health care hospital in Kinondo, Kenya where the samples are collected – fecal smears for the soil-transmitted helminths and cervical cytology for the cervical cancer.

Currently, manual microscopy is used for diagnostics of these diseases, but it has some drawbacks with it being time-consuming and requiring experts on site. The lack of experts in resource-limited regions amplifies these weaknesses. By scanning microscopy slides with small portable scanners and analyzing them with artificial intelligence, it has the potential to alleviate the work-burden of experts and help increase the access to high-quality diagnostics to people lacking it.

The doctoral project will evaluate the diagnostic accuracy, and the cost of artificial intelligence-supported digital microscopy compared to manual microscopy.”

Do you have any take-home messages from your ISP?

“I think the whole ISP process, both the preparation and seminar, was a good way to kick-start the doctoral project. Presenting the whole project to external researchers helped me understand what aspects of it I did not grasp completely and needed to delve deeper into.

I also think it was excellent to have the chance to receive feedback and discuss with senior researchers working with different projects. For example, I received questions about the broader implications of the research, such as how the artificial intelligence-supported microscopy might be incorporated into screening programs for cervical cancer in the future. This pointed out to me that while focusing on details is necessary when conducting research, it’s equally important to keep in mind that the goal of research is to provide knowledge that can improve our societies.”

/Public Release. View in full here.