Computer Vision News - November 2022

61 Qritive whole system gets handled automatically. It even takes care of all the complex MLOps required to deploy this for clinical validation or clinical use. ” Another issue is that histopathology has many proprietary file formats, scanners, and image file types , so it can be challenging for people to work with them all. However, AI Engine supports almost all the major slide scanner manufacturers. Qritive has put a great deal of engineering into its platform over the past five years to build a solid AI infrastructure for histopathology. in a scalable Kubernetes cluster. They efficiently build these algorithms, test them, and take them through internal validation. The same system allows Qritive to automatically deploy the modules for clinical validation to see if they work in a real-world environment . “ When building algorithms, training AI models and running experiments, companies and academia face issues with scaling up , ” Bruno points out. “ With AI engine, we can run hundreds or even thousands of experiments with GPUs attached to the Kubernetes cluster and the Enlarged (Inset area) + AI Summary ancer Gleason Grading AI Module AI classification of Benign + Gleason Pattern 3,4 & 5 Non-Tumor Area Pattern 3 Pattern 4 Benign 1