Overview

Bru™ was designed specifically for AI projects with focus on 3D medical images. It expedites the following two stages in the AI project lifecycle:

1) Data labeling
Bru™ has a web viewer for labeling of 3D medical images. The viewer employs various algorithmic tricks, including AI, to increase speed of labeling and provide convenient means for labels overseeing.

2) No or low code deployment of a trained Neural Network (NN)
Deployment of a NN can be as straightforward as uploading a file. Once uploaded, the NN becomes available for the Bru™ web viewer, allowing evaluation by AI team members and facilitating faster labeling. Furthermore, the NN becomes available for prediction through RESTful API.

Features

• Available where you store your data: AWS, Azure, on-premises
• Supported types of labels: polygon, segmentation mask, 3D bounding box
• Easy integration of your NN for assisting in labeling and model evaluation
• CT/MRI/PET DICOM compatibility
• Support fused modalities as PET-CT
• Reformatting of images

Use Cases

Create labeled MRI dataset for an AI project on abdominal organs segmentation

Create labeled PET-CT dataset for an AI project on horse limb segmentation and analysis of uptake and shape of body parts