Here are the [PDF] Segmentation based medical image compression of brain magnetic resonance images using optimized convolutional neural network journals presenting the latest research across various disciplines. From social sciences to technology, each article is expected to provide valuable insights to our readers.
Abstract
Image compression plays a crucial role in the feld of medical imaging, including Magnetic
Resonance Imaging (MRI). The MRI images are typically large and high-resolution, which
results in substantial data storage requirements. Compressing MRI images helps reduce the
storage space needed to store the images, making it more efcient and cost-efective to store
and transmit them. To overcome these drawbacks, this paper proposes an efcient medical
image compression based on hybrid machine learning approaches. There are two main stages
are considered in this proposed methodology, named a segmentation stage The Region of
Interest (ROI) in the image is recognized by the segmentation stage; and it given to the next
stage. Segmentation is carried out by hybrid Grey Wolf Optimization with Fuzzy C-Means
(FCM) is proposed to better balance the exploitation and exploration phases of optimization.
Then, the neural network i.e., optimized convolutional neural network (Op-CNN), compress
the ROI region of the input image depending on the detected segments. Meanwhile, the sec-
ond region (NROI) is compressed by the Recurrent Neural Networks (RNNs). The suggested
method of image compression for medical imaging outcomes and datasets are assessed with
highest PSNR value of 45.502, which is higher than the existing techniques.
https://link.springer.com/article/10.1007/s11042-023-16559-4
https://doi.org/10.1007/s11042-023-16559-4