This week we were thrilled to get the opportunity to present our poster on the topic “Using AI Based Nuclear Segmentation to Infer Toxicologic Pathology Outcomes such as Liver Hypertrophy from Cell Density Maps”.

This project was performed in collaboration with Charles River Laboratories. Together we demonstrated that the application of AI-enabled nuclear segmentation in combination with cellular density maps, can help with the rapid and reliable identification of hepatocellular hypertrophy. Liver hypertrophy is a lesion informed by subtle differences in individual hepatocyte size and extent which can be challenging to find without in-depth investigation by the pathologist at high magnification.

This study has provided a tool to enable pathologists to identify hypertrophic regions at low magnification with speed and accuracy. This is performed in an unsupervised fashion using a pretrained model which requires no additional annotations for the studies under investigation.

To gain access to the full poster please fill out the form below: