In two new studies, scientists from the University of California at Los Angeles (UCLA) used a neural network for reconstruction of holograms. Both works not only demonstrate the level of development of holographic technologies, but also promise to open this door technology in medicine, where they will be able to produce a real revolution.
In the first study, the results of which were described in the journal Light: Science & Applications, researchers used a technology called deep learning to create images of biological specimens, blood smears, PAP tests, and some other biological samples.
The study proved that the use of neural networks significantly accelerates and simplifies the process of creating holographic images, compared to more traditional methods of creating similar images that require for the reconstruction of the test object of the preliminary physical calculations and manual computer data entry.
In the second study, the team used their framework deep learning to improve the resolution and quality of microscopic images that help doctors to determine the smallest, barely noticeable anomalies in large samples of blood and tissue.
One of the problems with the current holographic methods is that the rendering process of the hologram may lose some information, which in turn can cause “artifacts” in the final image. Sometimes these losses are very significant. For example, the image may appear black spots, which the doctors mistakenly may be the growth of cancer cells. Such artifacts are often found during radiological scans, especially if the patient starts to move while the scanner does its job.
A system, deep learning at the University of California have demonstrated the effectiveness in solving this problem. Once the system is properly trained, neural network will be able to separate the spatial characteristics of a real image from any external interference (which often is the light).
The multilayered artificial neural networks allows deep learning algorithms to analyze the data offline. The technology has already demonstrated its effectiveness on the example of translation of speech from one language to another in real time, capture images, and many other tasks that before had to deal to a person who, by the way, loses the algorithms even in the speed of these tasks.
Since machine learning systems acquired the ability to sort and analyze vast amounts of information much faster than people, it is not surprising that these technologies are beginning to show their interest in a variety of areas, including medicine. The algorithms find their application, for example, in diagnostic radiology, where they demonstrate their effectiveness in reading x-ray images, as well as finding cancer cells that might be missed by physicians during scanning.
Holographic technologies are considered now not as it was before, when they were considered rather an object of science fiction than a practical tool. Now scientists are confident in the prospects of this direction.
Methods deep learning, in turn, can help in this direction, said Aydogan Ozcan, lead researcher. In his view, these technologies will open new possibilities of visualization. In the published press release of the University of California Ozkan noted that such technologies may even lead to the development of a completely new coherent imaging systems. The scientist believes that practices at UCLA can be used for further improvement of technology and introduction of it support other parts of the electromagnetic spectrum, e.g. x-ray and optical radiation.
If we expect a future that we could see in science fiction the last 40-50 years, holograms will play precisely not last role. UCLA research in this direction, in turn, not just trying to support this fantastic technology, they offer a real environment for its application.
AI will help holographic technology to reach a new level
Nikolai Khizhnyak