Submitted for "Moratuwa Engineering Research Conference (8th International Multidisciplinary Engineering Research Conference)"
"Many people suffer from lower back pain, and disc herniation is the most common cause. Medical
imaging can be used to diagnose spinal pathology as well as to analyze structural features,
surgical operations, and evaluate alternative treatments.
This study describes a framework that uses MRI data to automate lumbar intervertebral disc
segmentation. At the pixel level, we can detect small changes that are invisible to the naked
eye. To achieve semantic segmentation, we used convolutional neural networks with UNet
architecture. The Jacquard index and the dice coefficient were used to evaluate the
segmentation."
Received the Best Paper Award in Big Data, Machine Learning, and Cloud Computing.
Authors
Dilshani Karunarathna
Yuwin Siriwardhane
Imesh Ekanayake