Digital pathology, deep-learning and pancreatic cancer
When a tumor is resected from a patient, it can be studied by microscopy. Today, the microscopy slides are scanned, which enables more advanced computer-assisted analyses. Histology is a great tool, where the previously living tissue is captured as a snapshot in time. Thus, a tumor can be visualized in the context of all the surrounding, non-malignant cell types, which can help us to better understand how the tumor can spread through the pancreas. This might also help us to deduce what tumor characteristics that are informative in the clinical setting.
In the current project, we currently develop a deep-learning platform to segment (identify) various tissue components in pancreatic cancer, supported by Wallenberg Centre for Molecular Medicine Postdoc-program.
Background
I started my education at Linköping University, where I graduated from the Bachelor programme in Medicinsk biologi (Biomedicine) followed by the Master programme in Experimental and Medical Biosciences. In parallel, I was employed at Clinical Pathology at Region Östergötland, where I gained valuable experience in the diagnostic workflow.
During my time as a PhD student at Karolinska Institutet, supervised by Dr. Med. Marco Gerling and with outstanding training and support from my co-supervisor Dr. Med Carlos Fernández Moro, I developed an increasing interest for the pancreas, both regarding its functionality during normal, healthy conditions, and regarding the events of tumor initiation and progression. Here, I studied how tumors, primarily pancreatic ductal adenocarcinoma and neuroendocrine tumors of the pancreas spread in the pancreatic parenchyma. For example, I noticed how differentially the surrounding parenchyma of the pancreas was affected in different tumor types.
Now, I continue my research track in digital pathology and pancreatic cancer as a Postdoctoral researcher supervised by Dr. Docent Linda Bojmar and Dr. Med. Hakon Blomstrand.
