Faster analysis of medical images

The algorithm makes the process of comparing 3-D scans up to 1,000 times faster.

MIT researchers describe a machine-learning algorithm that can register brain scans and other 3-D images more than 1,000 times more quickly using novel learning techniques.
MIT researchers describe a machine-learning algorithm that can register brain scans and other 3-D images more than 1,000 times more quickly using novel learning techniques.

Until now, MRI technique has been used to form pictures of the anatomy and the physiological processes of the body in both health and disease. This procedure, however, can regularly take two hours or more, as customary systems meticulously align each of conceivably a million pixels in the combined scans.

Now, MIT scientists depict a machine-learning algorithm called VoxelMorph that can enlist brain scans and other 3-D pictures in excess of 1,000 times all the more rapidly utilizing novel learning procedures.

The algorithm learns while registering thousands of pairs of images. During this, it acquires information about how to align images and estimates some optimal alignment parameters. After training, it uses those parameters to map all pixels of one image to another, all at once. This reduces registration time to a minute or two using a normal computer, or less than a second using a GPU with comparable accuracy to state-of-the-art systems.

Co-author on both papers Guha Balakrishnan, a graduate student in MIT’s Computer Science and Artificial Intelligence Laboratory (CSAIL) and Department of Engineering and Computer Science (EECS) said, “The tasks of aligning a brain MRI shouldn’t be that different when you’re aligning one pair of brain MRIs or another. The tasks of aligning a brain MRI shouldn’t be that different when you’re aligning one pair of brain MRIs or another.”

MRI scans involve a stack of 2d images that form massive 3-D images, called “volumes,” containing a million or more 3-D pixels, called “voxels.” Therefore, it’s very time-consuming to align all voxels in the first volume with those in the second.

This process turns out to be especially moderate while breaking down outputs from the huge populace. Neuroscientists examining varieties in cerebrum structures crosswise over many patients with a specific disease or condition, for example, could possibly take several hours.

The VoxelMorph uses a convolutional neural network (CNN), a machine-learning approach ordinarily utilized for image processing. These systems comprise of numerous nodes that process the image and other data over a few layers of calculation.

Scientists trained their algorithm on 7,000 publicly available MRI brain scans and then tested it on 250 additional scans. They found that the algorithm learns information about groups of voxels — such as anatomical shapes common to both scans — which it uses to calculate optimized parameters that can be applied to any scan pair.

In short, the algorithm’s CNN component gains all necessary information during training so that, during each new registration, the entire registration can be executed using one, easily computable function evaluation.

The researchers found their algorithm could accurately register all of their 250 test brain scans — those registered after the training set — within two minutes using a traditional central processing unit, and in under one second using a graphics processing unit.

The papers are being presented at the Conference on Computer Vision and Pattern Recognition (CVPR), held this week, and at the Medical Image Computing and Computer Assisted Interventions Conference (MICCAI), held in September. Co-authors are Adrian Dalca, a postdoc at Massachusetts General Hospital and CSAIL; Amy Zhao, a graduate student in CSAIL; Mert R. Sabuncu, a former CSAIL postdoc and now a professor at Cornell University; and John Guttag, the Dugald C. Jackson Professor in Electrical Engineering at MIT.

Balakrishnan said, “The MICCAI paper develops a refined VoxelMorph algorithm that, “says how sure we are about each registration. It also guarantees the registration “smoothness,” meaning it doesn’t produce folds, holes, or general distortions in the composite image. The paper presents a mathematical model that validates the algorithm’s accuracy using something called a Dice score, a standard metric to evaluate the accuracy of overlapped images. Across 17 brain regions, the refined VoxelMorph algorithm scored the same accuracy as a commonly used state-of-the-art registration algorithm, while providing runtime and methodological improvements.”