HDC, DCL Medical to create new imaging tests for digital pathology

Health Discovery Corporation (HDC) and DCL Medical Laboratories have signed an agreement for the collaborative development and commercialization of support vector machine (SVM)-based computer-aided detection (CAD) tests for the independent detection of ovarian, cervical and endometrial cancers.

Tests will be performed on digital images to assist pathologists in providing more accurate diagnoses on biopsy and surgical specimens, according to both companies.

As part of the agreement, the Savannah, Ga.-based HDC will own the intellectual property and the Indianapolis-based DCL will have a sole-use license relating to applications and new mathematical tools developed during the course of the agreement.

Through the advancing technology of pattern recognition, both companies said that the new SVM-based diagnostic imaging systems are expected to improve the sensitivity of detection for endometrial and cervical cancers and significantly improve the specificity of ovarian cancer diagnosis. In addition, images and interpretative data from the new SVM-based systems will be designed for use in web-based applications, thus allowing remote review, second opinions and collaborative pathologist interpretation.

The new CAD-based digital pathology SVM-based algorithms could offer a faster, highly objective, and more accurate interpretation of cells to assist pathologists in making correct diagnoses for physicians and their patients, according to Stephen D. Barnhill, MD, chairman and CEO of HDC.

“We believe that using currently available image capturing technology, our SVM-based image analysis could be made available to assist pathologists in clinical laboratories, hospitals, academic centers and medical/pharmaceutical industries provide an accurate objective diagnostic interpretation to physicians and their patients around the world via transmission and evaluation of these digital pathology images using telepathology techniques over the internet,” Barnhill said.

 

Trimed Popup
Trimed Popup