Computer Vision News - August 2021

2 Summary Co puter Vision Tool 10 What is InnerEye? InnerEye aims at providing “automatic, quantitative analysis of three-dimensional medical images” delivered as an open-source solution. The project is a collaboration between Microsoft AI and NHS as well as researchers. By using state of the art and automatic deep learning on medical imaging on the cloud, it provides a safe, coherent workflow which will enable “researchers, hospitals, life science organizations, and healthcare providers to build medical imaging AI models using Microsoft Azure ”. One of the issues in the field of medical AI is the growing demand and the delivery of a coherent workflow, with great data management. That’s important for multiple reasons: workflow acceleration, patient safety, data safety and data integrity. By using a cloud service (in this project that would be Azure), that has a very high level of security, enabled with extra secure measures for institutions, researchers and hospitals, the hope is to achieve great coherency in all those. Introduction A recent paper published by Microsoft Research Cambridge and its evaluation shows how AI can augment and accelerate clinicians’ ability to perform radiotherapy planning 13 times faster . There is an interesting blog on the website about how the image segmentation model ( convolutional neural network based on a 3D U-Net architecture ), which has approximately 39 million trainable parameters, is used in the radiotherapy practice. I hope you had a good month and you’re again eager on learning about a new tool in the great landscape of medical AI. This month, I am going to present a tool, developed recently by Microsoft Research Cambridge , which aims at “democratize medical imaging AI”. Sounds like a good bet, doesn’t it? Now, read further to find out how this is going to happen! I hope, I already have you with me Tool Analysis: “Project InnerEye: Democratizing medical imaging AI