Artefact detection in MR images with transfer learning approach

What is it about?

This study aimed to develop a deep learning method for automatic detection of Gibbs artefacts in magnetic resonance imaging (MRI) using transfer learning. A magnetic resonance image dataset was created and annotated to include various intensities of Gibbs artefacts. The pre-trained VGG16 convolutional neural network was modified and re-trained for Gibbs artefact detection. The results showed that the created model had high accuracy and AUC values above 93% and 0.98, respectively. The study contributes to incorporating deep learning into MRI quality control and demonstrates its potential for effective automatic detection of artefacts, improving and speeding up the quality assessment process in medical imaging.

Featured Image

Photo by Accuray on Unsplash

Photo by Accuray on Unsplash

Why is it important?

The automatic detection of Gibbs artefacts in magnetic resonance imaging using deep learning is important because it can improve and accelerate the process of quality control in medical imaging. By automating the detection of artefacts, the process of quality control can be made more efficient, reducing the time and resources required for manual inspection. Additionally, the use of deep learning in medical imaging has the potential to provide more consistent and objective results, leading to improved patient outcomes.

Perspectives