Automatic segmentation of articular cartilage from MRI

US8706188B2 · kind B2 · utility

Assignee

• The Board of Trustees of the Leland Stanford Junior University · US

Inventors

• Seungbum Koo · Palo Alto, US
• Brian A. Hargreaves · Menlo Park, US
• Garry E. Gold · Stanford, US

Key dates

• Filing date: Jun 4, 2008
• Grant date: Apr 22, 2014
• Expiry date: Oct 2, 2031

Classification

• Technology area (CPC G): Physics
• CPC primary: G16H50/70
• WIPO field: Medical technology
• WIPO sector: Instruments

Abstract

A method for musculoskeletal tissue segmentation used in magnetic resonance imaging (MRI) is provided. MRI image data is collected using at least two different contrast mechanisms. Voxel values from data from each contrast mechanism are used as elements of a feature vector. The feature vector is compared with classification boundaries to classify musculoskeletal tissue type of the voxel. The previous two steps are repeated for a plurality of voxels. An image is generated from the classified musculoskeletal tissue types for the plurality of voxels to provide a musculoskeletal segmentation image.