Machine Learning has become a key component in most image processing workflows, and therefore in our framework. By learning from existing data, it allows to create and use complex algorithms without needing to implement hand-crafted rules. Our framework integrates natively the main deep learning frameworks like PyTorch and ONNXRuntime so that neural networks can be used in a convenient way. This allows the creation of AI-powered automated workflows.
ImFusion has a dedicated team of research scientists and engineers with extensive experience in machine learning applied to medical imaging to solve all kinds of medical imaging problems from the data annotation all the way to the integration in clinical products. They developped and use ImFusion Labels, a software to make our data management and annotation very efficient.
At ImFusion, we make extensive use of machine learning and have successfully applied it to a number of challenging problems, for instance:
Pixelwise classification of ultrasound images for real-time bone detection [1]
Multi-label bone classification in CT volumes
Trajectory estimation of an ultrasound sweep using only image information [2]
We offer to leverage our great experience with machine learning to get the most value out of your data, and speed up your developments by integrating machine learning into your workflow.
[1] M. Salehi, R. Prevost, J.-L. Moctezuma, N. Navab, W. Wein. Precise Ultrasound Bone Registration with Learning-based Segmentation and Speed of Sound Calibration.
[2] R. Prevost, M. Salehi, J. Sprung, R. Bauer, A. Ladikos, W. Wein. Deep Learning for Sensorless 3D Freehand Ultrasound Imaging.