C11328: A Method for the Classification and Segmentation of MRI ImagesNovelty:
A novel method for segmentation and classification which also combines multi-parametric MRI and/or other radiological imaging into a single dataset for increased specificity of diagnosis.
Value Proposition:
Early detection is the key to survival for cancer patients. Currently, there is limited technology with computer aided diagnosis (CAD) systems that integrate multi-parametric MRI and/or other radiological imaging procedures into highly specific datasets. Moreover, the CAD systems use only small portions of the data for diagnosis. This limits the ability for a radiologist to be confident in gauging benign and malignant lesion boundaries and the extent of disease. The technology is a method for development of diagnostic tools for early detection and classification of different tissue types with high specificity and sensitivity. Advantages of the present invention include:
• The ability to combine multiple-input images into a single unit for increased specificity of diagnosis.
• Creation of embedded images for fully automatic segmentation of different tissue types
.
• Easy development and fast implementation into CAD systems.
Technical Details:
Johns Hopkins researchers have developed a novel method for segmentation and classification of radiological images using tissue signature vectors from different tissues in multiple images and creating an embedded image for potential targets of disease. For example, in early detection of breast cancer, multi-parametric MRI data consists of fat suppressed T2-weighted (T2WI), T1-weighted (T1WI), Dynamic Contrast Enhanced (DCE) and diffusion weighted imaging (DWI). These methods provide critically functional information that is currently not captured in existing CAD systems and are not combining or visualizing the data. This technology improves the following diagnostic radiological imaging techniques, which are powerful non-invasive tools for identification of normal and suspicious regions within the body. The use of multi-parametric imaging methods that incorporate different functional parameters for quantitative diagnosis has been increasing.
Looking for Partners:
To develop and commercialize the technology as a method for classification and segmentation of MRI/radiological image processing.
Stage of Development:
Prototype
Data Availability:
Technology successfully tested using MRI data from patients
Publications/Associated Cases:
C11421
