Ultrasound imaging is based on the detection and interpretation of scattered ultrasound waves from tissues, and is used worldwide used for diagnostic purposes. Unfortunately, the traditional B-scan imaging technique used in 2D-imaging systems is not able to clearly differentiate fine details of inhomogeneous tissues. Additionally, measurements of 3D anatomy with ultrasound have poor reproducibility primarily due to user factors and the predominance of 2D-imaging systems. Alternative diagnostic technologies for assessing and quantifying the progress of disease states include MRI, PET-CT and biopsy examinations, but these are either invasive and painful or too expensive for frequent use. Supplementing the traditional B-scan images with additional quantitative information about the underlying tissue and cellular structures has therefore been a longstanding goal which has now been achieved in the H-scan.
Researchers at the University of Rochester have developed a methodology for providing additional information about imaged tissues. Based on the theory that different classes of tissue reflectors or scatterers produce different echoes, the H-scan utilizes matched filters to identify different tissue types of expected scatterers and encode corresponding echoes with colors, giving high-resolution information about the nature of the underlying tissue. The H-scan technique has shown sensitivity to scatterer size and morphology and can be adapted to conventional imaging systems (Figure 1).
In addition, the researchers have developed trajectories to represent the progression of disease and trained a support vector machine (SVM) to track the trajectories of progressive liver disease, fibrosis, steatosis, and pancreatic ductal adenocarcinoma metastasis. The approaches have been validated with 100% accuracy across the three different animal studies.
The technology provides a convenient, accurate and less expensive method for diagnosing disease states and monitoring their progression with ultrasound imaging without requiring surgical measures such as biopsies or injection of imaging agents. Using animal models, the H-scan technique has showed highly accurate tracking, classification, visualization, and quantification of the progression of multiple classes of diseases.
Diagnostic medical imaging using ultrasound