Researchers have devised a novel method to glean more information from images used to train machine-learning models, including those that can analyze medical scans to help diagnose and treat brain conditions.
An active new area in medicine involves training deep-learning models to detect structural patterns in brain scans associated with neurological diseases and disorders, such as Alzheimer's disease and multiple sclerosis. But collecting the training data is laborious: All anatomical structures in each scan must be separately outlined or hand-labeled by neurological experts. And, in some cases, such as for rare brain conditions in children, only a few scans may be available in the first place.
In a paper presented at the recent Conference on Computer Vision and Pattern Recognition, the researchers describe a system that uses a single labeled scan, along with unlabeled scans, to automatically synthesize a massive dataset of distinct training examples. The dataset can be used to better train machine-learning models to find anatomical structures in new scans -- the more training data, the better those predictions.
The crux of the work is automatically generating data for the "image segmentation" process, which partitions an image into regions of pixels that are more meaningful and easier to analyze. To do so, the system uses a convolutional neural network (CNN), a machine-learning model that's become a powerhouse for image-processing tasks. The network analyzes a lot of unlabeled scans from different patients and different equipment to "learn" anatomical, brightness, and contrast variations. Then, it applies a random combination of those learned variations to a single labeled scan to synthesize new scans that are both realistic and accurately labeled. These newly synthesized scans are then fed into a different CNN that learns how to segment new images.
"We're hoping this will make image segmentation more accessible in realistic situations where you don't have a lot of training data," says first author. "In our approach, you can learn to mimic the variations in unlabeled scans to intelligently synthesize a large dataset to train your network."
Although now applied to medical imaging, the system actually started as a means to synthesize training data for a smartphone app that could identify and retrieve information about cards from the popular collectable card game, "Magic: The Gathering." Released in the early 1990s, "Magic" has more than 20,000 unique cards -- with more released every few months -- that players can use to build custom playing decks.
Magnetic resonance images (MRIs) are composed of three-dimensional pixels, called voxels. When segmenting MRIs, experts separate and label voxel regions based on the anatomical structure containing them. The diversity of scans, caused by variations in individual brains and equipment used, poses a challenge to using machine learning to automate this process.
Some existing methods can synthesize training examples from labeled scans using "data augmentation," which warps labeled voxels into different positions. But these methods require experts to hand-write various augmentation guidelines, and some synthesized scans look nothing like a realistic human brain, which may be detrimental to the learning process.
Instead, the researchers' system automatically learns how to synthesize realistic scans. The researchers trained their system on 100 unlabeled scans from real patients to compute spatial transformations -- anatomical correspondences from scan to scan. This generated as many "flow fields," which model how voxels move from one scan to another. Simultaneously, it computes intensity transformations, which capture appearance variations caused by image contrast, noise, and other factors.
In generating a new scan, the system applies a random flow field to the original labeled scan, which shifts around voxels until it structurally matches a real, unlabeled scan. Then, it overlays a random intensity transformation. Finally, the system maps the labels to the new structures, by following how the voxels moved in the flow field. In the end, the synthesized scans closely resemble the real, unlabeled scans -- but with accurate labels.
To test their automated segmentation accuracy, the researchers used Dice scores, which measure how well one 3-D shape fits over another, on a scale of 0 to 1. They compared their system to traditional segmentation methods -- manual and automated -- on 30 different brain structures across 100 held-out test scans. Large structures were comparably accurate among all the methods. But the researchers' system outperformed all other approaches on smaller structures, such as the hippocampus, which occupies only about 0.6 percent of a brain, by volume.
"That shows that our method improves over other methods, especially as you get into the smaller structures, which can be very important in understanding disease," the author says. "And we did that while only needing a single hand-labeled scan."
http://news.mit.edu/2019/training-artificial-intelligence-brain-scan-0619
From one brain scan, more information for medical artificial intelligence
- 1,423 views
- Added
Edited
Latest News
Metabolic rewiring promotes…
By newseditor
Posted 18 Apr
A drug to prevent flu-induc…
By newseditor
Posted 18 Apr
New origin of deep brain waves
By newseditor
Posted 17 Apr
Starving cells hijack prote…
By newseditor
Posted 17 Apr
Miniature battery-free epid…
By newseditor
Posted 17 Apr
Other Top Stories
Tumors shrank with cell cycle-blocking drugs!
Read more
Are very low LDL cholesterol levels beneficial for patients?
Read more
Functional neuronal regeneration from Müller glia in adult mice
Read more
Mechanism of lymphatic vessel growth during infection!
Read more
Patient plays saxophone while surgeons remove brain tumor
Read more
Protocols
MemPrep, a new technology f…
By newseditor
Posted 08 Apr
A tangible method to assess…
By newseditor
Posted 08 Apr
Stem cell-derived vessels-o…
By newseditor
Posted 06 Apr
Single-cell biclustering fo…
By newseditor
Posted 01 Apr
Modular dual-color BiAD sen…
By newseditor
Posted 31 Mar
Publications
How does the microbiota con…
By newseditor
Posted 18 Apr
The integrated stress respo…
By newseditor
Posted 18 Apr
The immunobiology of herpes…
By newseditor
Posted 17 Apr
Circulating microbiome DNA…
By newseditor
Posted 17 Apr
Spindle oscillations in com…
By newseditor
Posted 17 Apr
Presentations
Hydrogels in Drug Delivery
By newseditor
Posted 12 Apr
Lipids
By newseditor
Posted 31 Dec
Cell biology of carbohydrat…
By newseditor
Posted 29 Nov
RNA interference (RNAi)
By newseditor
Posted 23 Oct
RNA structure and functions
By newseditor
Posted 19 Oct
Posters
A chemical biology/modular…
By newseditor
Posted 22 Aug
Single-molecule covalent ma…
By newseditor
Posted 04 Jul
ASCO-2020-HEALTH SERVICES R…
By newseditor
Posted 23 Mar
ASCO-2020-HEAD AND NECK CANCER
By newseditor
Posted 23 Mar
ASCO-2020-GENITOURINARY CAN…
By newseditor
Posted 23 Mar