Quebec City, Quebec, Canada – NLP Logix competed in an international data science contest to train computers to be able to accurately detect white matter hyperintensities (WMH). The final tally was announced at the 20th International Conference on Medical Image Computing and Computer Assisted Intervention (MICCAI) 2017 on September 14th and NLP Logix placed 3rd out of dozens of competitors from across the world.
“This has been a very challenging competition,” said Matt Berseth, Lead Scientist, NLP Logix who was in Quebec City to present the NLP Logix methodology and accept the award. “We spent a number of months training the models and really pushing the limits on our team and tools.”
The value of a model to accurately read white matter hyperintensities is very important because now NLP Logix can process very large volumes of images, identify WMH and calculate its volume. This process gives physicians and patients with diseases such as multiple sclerosis insights into how the disease is progressing.
Over the next month, NLP Logix will be deploying the model to the Microsoft Azure cloud and will be testing the interface which will allow organizations to score T2 weighted MRI images.