Computer aided diagnostic method and device

Abstract

A method for characterizing regions on a map or image generated from diffusion image data of a region of a patient's body, including: loading into a data processing device image data achieved by calculation from MR diffusion imaging; providing a first threshold level to distinguish pixels with image data values that are above or below the values found that are not representative of pathological portions of the region; providing a second threshold level higher than the first threshold; deriving a measure related to the gradient pattern of those pixels that lie within a margin zone defined by those pixels that satisfy a criterion of the second threshold but do not satisfy a criterion of the first threshold and; outputting the measure related to the gradient as a parameter indicative of the likelihood that the pixels that are within the margin zone and / or a region of the body part are cancerous.

Description

[0001]This invention was made with U.S. Government support under contract number HHSN261201100075C awarded by National Institutes of Health (NIH). The U.S. Government has certain rights in the invention.BACKGROUND OF THE INVENTION[0002]The present invention relates to a method and device for estimating the likelihood that an image of a body or body part obtained through Diffusion Tensor Imaging (DTI) or Diffusion Weighted Imaging (DWI) contains a lesion and further estimating the likelihood that the lesion is cancer.[0003]The known technique of diffusion imaging is disclosed in U.S. Patent Application Publication No. 20100298692A1, the disclosure of which is incorporated herein by reference.[0004]Background in Breast MRI Methods and Terminology:[0005]The invention is applicable to a plurality of parts of the body. For the sake of clarity, the background and description will be described in terms of detection and diagnosis of breast cancer. It is to be understood that this descriptio...

AI Technical Summary

Benefits of technology

The present invention relates to methods for generating and characterizing maps or images of regions of a patient's body using diffusion MR image data. These methods can detect benign or malignant lesions in various parts of the body and provide valuable information about the nature of the lesion(s). The technical effects of this invention include improving the accuracy of detecting and characterizing lesions, which can aid in the diagnosis and treatment of cancer.

Problems solved by technology

However, ADC maps are not sufficiently sensitive for establishing a stand alone method for breast cancer detection.

Despite its excellent contrast resolution, DWI has the disadvantages of susceptibility and chemical shift artifacts.

This is probably because the parallel imaging techniques are capable of reducing the number of phase encoding steps and the time required to fill the k-space, which may lead to the suppression of susceptibility and chemical artifacts.

One inherent problem with breast DCE, the current standard method of magnetic resonance imaging, is that enhancing agents are routinely used off-label.

Non-labeled use can result in inconsistency in how the contrast agent is administered.

Images