DETECTION OF VARICOSE VEINS FROM DOPPLER ULTRASOUND IMAGES USING CNN

  • KRITTIKA I. SHIRKOL, MALLIKARJUNASWAMY M. S., NANDA S., SHAILAJA K.
Keywords: Varicose Veins, CNN, TensorFlow, Color Doppler Image, Back Flow of Blood

Abstract

Varicose veins are the swollen twisted weak veins which are present under the skin which usually occurs in the legs. Varicose veins are often damaged vein valves, which forms whenever there is an increase in blood pressure in veins. The main feature in varicose veins is back flow of blood. Detection of affected veins in early stage and treatment reduces further complications. Ultrasound color Doppler imaging better visualize varicose veins compared to other imaging modalities. Automatic detection of varicose veins helps doctor in decision making and treatment of affected people. In this work, convolutional neural network (CNN) is used for classification of varicose veins using ultrasound color Doppler images. CNN developed for training and testing of data and shown good performance. The accuracy obtained using CNN is high and takes less time for execution. The developed method useful to doctors in detection and treatment of varicose vein affected people.

Published
2021-10-27
How to Cite
NANDA S., SHAILAJA K., K. I. S. M. M. S. (2021). DETECTION OF VARICOSE VEINS FROM DOPPLER ULTRASOUND IMAGES USING CNN. Design Engineering, 2912-2920. Retrieved from http://www.thedesignengineering.com/index.php/DE/article/view/5768
Section
Articles