Neurintel (402) 438-2693 | (402) 679-0849 |    



Tendon Ultrasound Image Analysis

Two U.S. universities with a vested interest in the advancement of rehabilitation and medical imaging.

Much of the population, especially persons who are disabled or advancing in age, can frequently experience pain and reduced function due to the degeneration of tendons, ligaments and joints. Magnetic resonance imaging (MRI) is often used to effectively assess the status of tendons. However, the expenses associated with MRIs can be prohibitive until advanced pathology is present. Neurintel and associates were assigned the task of developing a reliable and valid method for classifying tendon morphology using ultrasonography, a more cost-effective and portable imaging method than MRI.

Key Challenges
The major task was developing an effective and reliable method of image analysis that could accompany the ultrasonographic approach. Success was dependent upon the ability to accurately predict the level of pathology compared to a "gold standard" microscopic examination performed on tendon tissue harvested during surgery.

Advanced signal processing and artificial neural networks (ANNs) were used to determine whether ultrasonic images of tendons could detect a healthy structure versus one with tendinopathy. An ANN is an intelligent learning algorithm that is able to make classifications after "learning" input-output relationships that are defined by users. Key characteristics of the images were extracted using signal processing techniques. These features were used as inputs to a trained ANN and the output was a classification of the pathology of the tendon.

Correct classification rates of 95-99% were achieved using the above proposed methods. The future implementation of these advancements will lead to cost and time savings in the rapidly growing healthcare industry.

P.O. Box 83342
Lincoln, NE 68501-3342
Office: (402) 438-2693
Mobile: (402) 679-0849
Home | Who We Are | What We Do | Who We Work With | Investigation | Contact Us
Copyright © 2020 Neurintel. All rights reserved | Website by Digital House