Pathologists expertise and domain knowledge can inspire us engineers to develop more sophisticated models able to solve the task at hand.
In order to implement artificial intelligence into laboratory practice we need cooperation. Competences provided by computers differ greatly from ours, and so, pathologists will benefit from these tools. But this also can apply in the opposite direction. Pathologists expertise and domain knowledge can inspire us engineers to develop more sophisticated models able to solve the task at hand. This intuition led me to implement AI algorithms inspired from real-world WSI processing and extracting regions of interest. The idea was submitted as an abstract to NOBIM (Norwegian Association for Image Processing and Machine Learning) conference in Oslo where I presented an ongoing research in weakly supervised learning and interpretability.
It’s exciting to see how a multidisciplinary environment can change your mindset and improve the way you face novel challenges.
NOBIM picture of Saul
Saul Fuster Navarro – ESR5.