A EU-based manufacturer of smart medical devices contracted Auriga’s data science team to R&D new medical image analysis algorithms to assist with automated detection of certain patterns in tissue images.
State-of-the-art neural network architecture. Initial data validation and modelling to make optimal dataset for training. AB testing to prove solution efficiency. Operational trials of the solution. HW platforms configuration and evaluation.
Data preprocessing algorithms optimization – abnormalities identification rate increased up to 20% compared to regular expert review. Solution can process each image in seconds dramatically accelerating diagnostics process. Significant cost reduction for HW platform configuration. Developed solution is planned to be adopted as a part of national screening examinations.
