New research could help AI identify rare conditions in medical images by augmenting existing datasets
Artificial intelligence (AI) holds real potential for improving both the speed and accuracy of medical diagnostics. But before clinicians can harness the power of AI to identify conditions in images such as X-rays, they have to 'teach' the algorithms what to look for.
Identifying rare pathologies in medical images has presented a persistent challenge for researchers, because of the scarcity of images that can be used to train AI systems in a supervised learning setting.
Professor Shahrokh Valaee and his team have designed a new approach: using machine learning to create computer generated X-rays to augment AI training sets.
"In a sense, we are using machine learning to do machine learning," says Valaee, a professor in The Edward S. Rogers Sr. Department of Electrical & Computer Engineering (ECE) at the University of Toronto. "We are creating simulated X-rays that reflect certain rare conditions so that we can combine them with real X-rays to have a sufficiently large database to train the neural networks to identify these conditions in other X-rays."
Valaee is a member of the Machine Intelligence in Medicine Lab (MIMLab), a group of physicians, scientists and engineering researchers who are combining their expertise in image processing, artificial intelligence and medicine to solve medical challenges. "AI has the potential to help in a myriad of ways in the field of medicine," says Valaee. "But to do this we need a lot of data -- the thousands of labelled images we need to make these systems work just don't exist for some rare conditions."
To create these artificial X-rays, the team uses an AI technique called a deep convolutional generative adversarial network (DCGAN) to generate and continually improve the simulated images. GANs are a type of algorithm made up of two networks: one that generates the images and the other that tries to discriminate synthetic images from real images. The two networks are trained to the point that the discriminator cannot differentiate real images from synthesized ones. Once a sufficient number of artificial X-rays are created, they are combined with real X-rays to train a deep convolutional neural network, which then classifies the images as either normal or identifies a number of conditions.
"We've been able to show that artificial data generated by a deep convolutional GANs can be used to augment real datasets," says Valaee. "This provides a greater quantity of data for training and improves the performance of these systems in identifying rare conditions."
The MIMLab compared the accuracy of their augmented dataset to the original dataset when fed through their AI system and found that classification accuracy improved by 20 per cent for common conditions. For some rare conditions, accuracy improved up to about 40 per cent -- and because the synthesized X-rays are not from real individuals the dataset can be readily available to researchers outside the hospital premises without violating privacy concerns.
"It's exciting because we've been able to overcome a hurdle in applying artificial intelligence to medicine by showing that these augmented datasets help to improve classification accuracy," says Valaee. "Deep learning only works if the volume of training data is large enough and this is one way to ensure we have neural networks that can classify images with high precision."
Marit Mitchell | EurekAlert!
Patented nanostructure for solar cells: Rough optics, smooth surface
18.09.2018 | Helmholtz-Zentrum Berlin für Materialien und Energie GmbH
With Gallium Nitride for a Powerful 5G Cellular Network - EU project “5G GaN2” started
17.09.2018 | Fraunhofer-Institut für Angewandte Festkörperphysik IAF
The building blocks of matter in our universe were formed in the first 10 microseconds of its existence, according to the currently accepted scientific picture. After the Big Bang about 13.7 billion years ago, matter consisted mainly of quarks and gluons, two types of elementary particles whose interactions are governed by quantum chromodynamics (QCD), the theory of strong interaction. In the early universe, these particles moved (nearly) freely in a quark-gluon plasma.
This is a joint press release of University Muenster and Heidelberg as well as the GSI Helmholtzzentrum für Schwerionenforschung in Darmstadt.
Then, in a phase transition, they combined and formed hadrons, among them the building blocks of atomic nuclei, protons and neutrons. In the current issue of...
Thin-film solar cells made of crystalline silicon are inexpensive and achieve efficiencies of a good 14 percent. However, they could do even better if their shiny surfaces reflected less light. A team led by Prof. Christiane Becker from the Helmholtz-Zentrum Berlin (HZB) has now patented a sophisticated new solution to this problem.
"It is not enough simply to bring more light into the cell," says Christiane Becker. Such surface structures can even ultimately reduce the efficiency by...
A study in the journal Bulletin of Marine Science describes a new, blood-red species of octocoral found in Panama. The species in the genus Thesea was discovered in the threatened low-light reef environment on Hannibal Bank, 60 kilometers off mainland Pacific Panama, by researchers at the Smithsonian Tropical Research Institute in Panama (STRI) and the Centro de Investigación en Ciencias del Mar y Limnología (CIMAR) at the University of Costa Rica.
Scientists established the new species, Thesea dalioi, by comparing its physical traits, such as branch thickness and the bright red colony color, with the...
Scientists have succeeded in observing the first long-distance transfer of information in a magnetic group of materials known as antiferromagnets.
An international team of researchers has mapped Nemo's genome, providing the research community with an invaluable resource to decode the response of fish to...
03.09.2018 | Event News
27.08.2018 | Event News
17.08.2018 | Event News
20.09.2018 | Earth Sciences
20.09.2018 | Earth Sciences
20.09.2018 | Physics and Astronomy