As early as 1935, science fiction writer Stanley Weinbaum’s novel “Pygmalion’s Glasses” conceived a pair of glasses to realize virtual reality (VR). In recent years, in addition to well-known application scenarios such as games and entertainment, VR has gradually entered various fields such as medical care, education, and manufacturing.
Taking medical care as an example, VR may help alleviate problems such as limited medical resources, uneven distribution, and single diagnostic methods, and realize a series of new business forms such as remote diagnosis and treatment and telemedicine teaching, thereby further becoming an effective supplement to traditional medical methods.
In 2016, an 80-year-old patient with right colon cancer successfully underwent tumor resection, and for the first time broadcast the whole operation live through VR. Operation.
Although after that, many hospitals have carried out research on VR. However, after experiencing ups and downs after 2016, and even being questioned by many, the limelight of VR is far less than that of artificial intelligence entrepreneurship in the same period.
An industry expert said, “We have to admit that VR technology has not yet reached the mature stage of actual production, especially for a special industry such as medical care. Go through constant trial and error and verification.”
However, the outside world’s expectations for VR in the medical industry are slowly moving towards reality. According to a data from Perkins Coie, the application penetration rate of VR/AR in the medical and health field will be the highest in 2021, exceeding 35%, which has a feeling of “although late but arrived”. After the introduction of the VR system, some hospitals that are at the forefront and have the courage to try have experienced the real convenience brought by the technology, which has made a good start for themselves and the industry’s road to smart medical care.
Is the combination of VR and medical treatment a flower in the mirror?
Since the concept of “Virtual Reality” was proposed, various industries have carried out a wealth of research on VR equipment, and the application of VR in the medical field can be traced back to the 1980s.
In 1985, the U.S. National Library of Medicine began digital research on human anatomy images, and used virtual human bodies to carry out computer-aided teaching in subjects such as virtual anatomy, virtual radiology, and virtual endoscopy. In 1994, the Royal Manchester Hospital in the United Kingdom established the first virtual reality development center in Europe for minimal puncture therapy.
At the beginning of 2003, my country’s First Military Medical University announced that it had completed the data collection of the first female virtual human in China.
In the following years, the development of VR technology has experienced a period of dormancy, and it seems that there has not been too much splash in the medical field. Is the combination of VR and medical treatment just a “flower in the mirror” after all?
A related expert from Volcano Engine told Leifeng.com, “The development of any new technology will go through five stages, the impetus period of the birth of technology, the peak period of excessive expectations, the trough period of bubble formation, and the bright period of steady climbing. This is why everyone perceives that the VR concept and technology have gone from boom to trough. Now, VR has come into everyone’s sight again and has been valued by all walks of life. From the perspective of development history, VR has entered A bright period of steady climb has begun.”
It can also be seen from industry data that various cutting-edge applications integrating VR technology are playing an increasingly important role in the medical field.
According to the International Data Corporation (IDC), the five-year compound annual growth rate for AR/VR spending worldwide is 78.3%. Global AR/VR spending is expected to reach $160 billion by 2023, well above the $16.8 billion forecast in 2019. In the medical field, AR/VR medical care is undergoing substantial improvement, with an expected compound annual growth rate of 33.36%, and the market value is expected to increase from US$621.7 million in 2018 to US$3.497 billion in 2024.
Industry giants are also further deploying VR and its products and solutions in the medical field. In February 2020, Johnson & Johnson released a VR helmet developed in cooperation with OSSOVR, which is specially used to train surgeons; in May this year, Siemens Healthcare began to provide immersive virtual reality (VR) training services to help surgeons and technicians perform surgical procedures. Medium quality control and surgical workflow guidance.
What can hospitals get from VR applications?
In addition to the active deployment of medical giants, VR technology has gradually entered the field of vision of the actual users, that is, hospital managers.
In recent years, the release of a series of industry policies has posed a huge challenge to the hospital’s ability to balance the balance of payments and viability. Hospitals need to take the initiative to seek changes, and focus on the intelligent upgrading of hospitals to improve their service capabilities and survival and competitiveness. For higher-level hospitals, in addition to daily diagnosis work, they also undertake important tasks of medical teaching and grassroots assistance.
The president of a tertiary hospital in Beijing said, “VR technology is very important in the medical system, especially medical education. Medical education is not only for medical workers, but also for ordinary people. Through VR technology, let them experience the It is a very good way to experience medical knowledge in the environment.”
He believes that medical care, like all walks of life, realizes digital transformation and intelligent development, which is the core connotation of promoting the high-quality development of public hospitals.
In 2021, Intel and Volcano Engine will help a tertiary hospital in Beijing build a medical emergency education and training center that integrates artificial intelligence, holographic/VR, and Internet of Things technologies based on a brand-new VR medical training application, and effectively improve the hospital’s medical experience. Work effectiveness in the field of training.
Tan Dai, President of Volcano Engine, revealed that the cooperation has two parts: one is VR remote inspection, which uses technology to solve the limitations of time and space on expert resources; the other is VR emergency training, through simulated and immersive VR scenes, for emergency medical Personnel coaching and training.
From the perspective of the specific implementation path of the system deployment, the overall architecture of the VR medical training program of a tertiary hospital in Beijing consists of different processing devices and display terminals (VR glasses) deployed in the first site (ward), edge cloud and second site .
Among them, audio and video acquisition devices (such as panoramic high-definition cameras) are deployed at the first site, and these data will be synchronized at high speed via the 5G network to the edge cloud deployed near the medical training center for audio and video rendering and data calculation processing.
The rendered VR video stream will be encoded and compressed and pushed to the VR glasses on the second site through the high-speed 5G network, which is convenient for users (doctors, medical students, volunteers, etc. in teaching) to conduct VR remote ward rounds, VR Remote guidance, AED (automated external defibrillator) VR wireless first aid and other behaviors.
The ongoing user behavior data will also be synchronously sent back to the edge cloud for processing, thus forming a closed-loop process for the entire VR medical training.
During the cooperation process, the VR training system faces some common industry challenges:
First of all, VR applications such as remote real-time ward rounds bring massive data collection, processing, and transmission requirements. If all workloads are completed through remote data centers/cloud services, it will inevitably bring huge network pressure. The immersive and interactive experience in VR applications also requires a solution that can effectively shorten the overall end-to-end delay of the system;
Secondly, the particularity of medical scenarios requires that VR applications can provide higher-resolution video formats (such as 8K) to display various details under the existing network bandwidth, which poses a huge challenge to the optimization ability of the video encoding and decoding process.
Facing two major challenges, the core keywords of the solution are two points: faster and more real. Tan Dai said, “Quickness and truth are often contradictory. To solve this contradiction, better algorithms, stronger engineering capabilities and better infrastructure are needed to support it.”
Together with Intel, Volcano Engine introduces Intel® Advanced Vector Extensions 512 (Intel® AVX-512), optimizes the BVC codec, and provides strong support for rendering computing processing in edge cloud data centers to accelerate VR applications and achieve Real-time fast decoding of ultra-clear and high-quality video.
The BVC codec is a self-developed product of the Volcano Engine Multimedia Laboratory. In the VR medical training application scheme, it is used to improve the immersion and interaction when using the VR medical training application, and to support the encoding and transmission brought by 8K video resolution. pressure.
Intel® AVX-512 is a Single Instruction Multiple Data (SIMD) instruction set provided by Intel. Thanks to its advantages in register bit width, the Intel® Xeon® Scalable processor with built-in Intel® AVX-512 instruction set can be used in high-density computing processing scenarios such as audio and video processing, data encryption and compression, and deep learning Has excellent performance.
The latest generation of 4th generation Intel® Xeon® Scalable processors has a series of built-in accelerators, including the new AI accelerator – Intel® Advanced Matrix Extensions (Intel® AMX), covering more tasks including training and fine-tuning. Deep learning use cases that can provide more substantial computing performance for ever-changing and increasingly demanding applications.
Focusing on the needs of deploying VR in hospitals, Intel and the Volcano Engine have optimized the interpolation module and some basic computing modules in the BVC codec, which take up relatively high processor resources, through the Intel® AVX-512 instruction set to improve video performance. Capable of encoding, decoding and compression, so as to meet the complex requirements of hospital VR medical training applications, such as high image compression rate, fast encoding speed, and low bandwidth occupation.
Wu Wenxin, chief technology officer of Intel Cloud and Industry Solutions Group, said, “In general, from video collection, rendering to push, we finally met the hospital’s requirements, with a delay of less than 20 milliseconds and a low bandwidth of 15M. Realize the effect of 8K, effectively solve the picture fineness problem in the doctor’s teaching room, and provide a more scientific on-site reference environment for teaching and consultation.”
At present, with the joint efforts of the three parties, the medical emergency education and training center of the Beijing tertiary hospital has been successfully deployed, and a large number of medical training practices have been carried out, and the training content has been reused in VR, achieving the effect of one-time investment and permanent use. The marginal cost of each use is almost only 5G traffic fee, which effectively helps to reduce training costs and improve economic benefits. ( Interested readers can view the case and program details )
In addition to greatly improving training efficiency, this set of VR application solutions can also provide doctors with a more scientific and realistic on-site teaching environment, and provide an unprecedented way for trained doctors in remote and grassroots areas to improve their professional level. More importantly, unlike traditional remote consultations, with the help of VR medical training applications, experts can provide remote guidance to front-line doctors and communicate with patients in real time, bringing more possibilities for remote smart medical care. The successful implementation of VR technology in the medical and health field will have a profound impact on the construction of regional medical alliances and the strengthening of grassroots diagnosis and treatment capabilities that the country has been advocating.
Cloud-edge collaboration, how does medical VR benefit “Healthy China”?
Relying on the successful implementation and effective support of the VR medical training program, the medical emergency education and training center has realized the innovation of the training mode, which is also a true portrayal of many medical policy guidance and intelligent transformation of hospitals in recent years.
On November 3, 2021, the National Health Commission issued the “Thousand Counties Project” County Hospital Comprehensive Capacity Improvement Work Plan (2021-2025) (hereinafter referred to as the “Work Plan”).
The “Work Plan” clarifies that the promotion of high-quality medical resources in provinces and cities to the county area, combined with the county hospital’s standard expansion project, will make up for the shortcomings of the county hospital’s medical service and management capabilities, and gradually realize the integration and sharing of medical resources in the county area. By 2025 , At least 1,000 county hospitals across the country have reached the level of medical service capabilities of tertiary hospitals, giving full play to the role of county medical centers, and laying a solid foundation for the realization of common diseases in cities and counties.
On the other hand, marked by the full launch of “East Counting and West Computing”, 300 billion new infrastructure supplementary investment in the middle of the year, and the acceleration of 5G large-scale application, China’s major computing power and digital infrastructure construction projects in 2022 New steps have been taken to further promote the digital and intelligent transformation of enterprises.
The cooperation between Intel and Volcano Engine has taken an early step towards these two macro-construction goals, and verified the development potential of VR in the medical and health field to a certain extent.
Experts from Volcano Engine believe that VR is gradually demonstrating its irreplaceable advantages and value in innovative applications and experiments in the medical industry, especially in medical training, auxiliary diagnosis and treatment, and clinical assistance.
“We have to admit that VR technology has not yet reached the real mature stage of actual production, especially for a special industry such as medical care, there is still a long way to go to truly move towards the stage of high availability, and it has to go through continuous trial and error and verification Therefore, it is even more necessary to start exploration and co-creation from now on, while the technology itself is developing rapidly, to find the best practice of combining VR and medical care as soon as possible, to promote the development of VR medical care, and to assist the development of overall digital medical care.”
From the perspective of “the practice of integrating technology and medical care”, Wu Wenxin said that considering VR as an important user entrance is also an important part of Intel’s cloud-side collaborative solution, which will promote the accelerated digital transformation of the medical industry. Intel’s cloud-edge collaboration architecture will bring value to medical institutions in at least three aspects.
“First, to make data processing faster while taking data security into account; second, to make medical applications more intelligent, such as using AI technology to assist grassroots doctors in diagnosis and treatment; third, to make medical services more inclusive, through VR The immersive interactive mode allows clinics in remote areas and communities to learn the diagnosis and treatment experience of tertiary hospitals, and allows patients to receive treatment from doctors in large hospitals, promoting the promotion and popularization of smart medical care in China.”
On December 28, 2021, ten departments including the Ministry of Industry and Information Technology jointly issued the “14th Five-Year Plan for the Development of Medical Equipment Industry”, in which it was proposed to promote the development of a new model of “5G+” medical and health:
Promote the innovation of smart medical health equipment and applications, support the accelerated development of remote consultation and remote consultation, lay a solid foundation and support for the construction of an orderly and effective hierarchical diagnosis and treatment pattern, and promote the sinking of high-quality medical resources to serve the grassroots; strengthen clinical application innovation Research. Support in-depth cooperation among medical institutions, scientific research institutions, and manufacturing enterprises, benchmark against international advanced levels, strengthen the coordinated development of medical industry, and carry out innovative research on the clinical application of medical equipment.
“Supporting in-depth cooperation among medical institutions, scientific research institutions, and manufacturing enterprises”, such an industry-university-research combination is particularly important for the industry transformation needs of the intelligent era.
In recent years, Intel has been working with various partners to continuously explore the application of technological innovation in the field of healthcare. For example, China Mobile’s 5G telemedicine, based on Intel® Xeon® Scalable processors and GPUs, launched a 5G cloud-network integration overall solution design, creating a multi-network integration, native security, cloud-network synergy, and wide-area interconnection. Network, to help hospitals achieve remote ultrasound, remote consultation, 5G emergency rescue, etc., and to perform remote surgery through remote operation of micro-robots to help medical poverty alleviation.
Intel has also been active at the forefront of “AI for Science”, with Xeon® Scalable processors with built-in AI acceleration technology and Optane with a capacity several times that of traditional memory 
The combination of persistent memory implements end-to-end high-throughput optimization of AlphaFold2 on the CPU platform, helping AI pharmaceutical companies represented by Jingtai Technology to break through the computing power bottleneck of new drug research and development, and extensively empower more digital and intelligent The efficiency of drug R&D has been upgraded.
This perfect fit of technology and industry will play an important role in promoting the country’s “Healthy China” strategy.
This article is reproduced from: https://www.leiphone.com/category/aihealth/meCh0i2S5L8a1tI0.html
This site is only for collection, and the copyright belongs to the original author.