Credit: IEEE deals on ultrasound, ferroelectricity and frequency control (2022). DOI: 10.1109/TUFFC.2022.3219401
DGIST Department of Electrical Engineering and Computer Science Professor Hwang Jae-yoon’s team has developed a deep learning-based ultrasonic hologram framework technology that can freely configure the converging ultrasound geometry in real time. real-time based on holograms. It is expected to be used as a fundamental technology in the field of brain stimulation and treatment that requires precision in the future.
Ultrasound is a safe technology even used for prenatal testing. Since it can stimulate deep areas without surgery, ultrasound methods for brain stimulation and treatment have recently been investigated. Previous research has found that ultrasonic brain stimulation can help treat conditions like Alzheimer’s disease, depression, and pain.
The problem, however, is that it is difficult to selectively stimulate relevant areas of the brain, where several areas interact at the same time, because current technology focuses ultrasound waves on a single point. most or a large circle to stimulate. To solve this problem, a technology capable of focusing the ultrasound freely on the desired area using the hologram principle has been proposed, but has limitations, such as low accuracy and long computation time to generate holograms.
Professor Hwang Jae-yoon’s group of DGIST has proposed a deep learning-based learning framework that can demonstrate free and accurate ultrasonic focusing in real time by learning how to generate super holograms. negative to overcome limitations. As a result, Professor Hwang’s group has demonstrated that it is possible to focus the ultrasound on the desired form more precisely in near-real-time holograms and up to 400 times faster than the algorithmic approach. create existing ultrasound holograms.
The deep learning-based learning framework proposed by the research team learns to generate ultrasonic holograms through self-supervised learning. Self-supervised learning is a method of learning to find answers by finding patterns for data with no answers. The team proposed a learning method to generate ultrasound holograms, a deep learning network optimized for ultrasound. holograms and a new loss function, while demonstrating the validity and excellence of each component through simulations and real-world tests.
Professor Hwang Jae-yoon of DGIST’s Department of Electrical Engineering and Computer Science said: “We applied deep learning technology to the proposed ultrasonic hologram relatively recently. As a result, we developed a technology that can freely, quickly and accurately generate and change the ultrasonic shape of beams,” and added, “We hope that the results of this study can be used with great accuracy.” body for the patient brain stimulation field of technology and general ultrasound (supersonic imaging, heat therapy, etc.).
Article published in the magazine IEEE deals on ultrasound, ferroelectricity and frequency control.
Moon Hwan Lee et al, Deep learning-based framework for fast and accurate audio hologram generation, IEEE deals on ultrasound, ferroelectricity and frequency control (2022). DOI: 10.1109/TUFFC.2022.3219401
Provided by DGIST (Daegu Gyeongbuk Institute of Science and Technology)
quote: Development of real-time ultrasonic hologram generation technology based on deep learning (2022, December 29) retrieved December 29, 2022 from https://techxplore.com/news/2022-12- deep-based-real-time-ultrasonic-hologram.html
This document is the subject for the collection of authors. Other than any fair dealing for private learning or research purposes, no part may be reproduced without written permission. The content provided is for informational purposes only.
