A New Way to Read Medical Imaging with All-Optical Smart Hospital
Huawei's all-optical hospital solution is helping develop smart hospitals. It enables all-optical medical imaging, improving storage, networks, and terminals for healthcare providers.
"The first step in building a smart hospital is enabling information connectivity. Networks are like highways, so when information runs smoothly along networks, we can begin to embrace new technologies such as cloud computing and artificial intelligence," explained Lv Zhouping, Deputy Director of the Network Technology Department of Huazhong University of Science and Technology Union Shenzhen Hospital, in an interview with Shenzhen Special Zone Daily.
As the world goes digital, new medical technologies are being developed — for example, digital imaging systems and all-optical hospitals — bringing the healthcare industry a step closer to building fully smart hospitals.
Doctors use a range of imaging systems to diagnose patients, including magnetic resonance imaging (MRI), digital subtraction angiography (DSA), and positron emission and computed tomography (PET-CT). These technologies have made incomparable contributions to increasing people's life expectancy and health. Medical imaging is widely used in medical examination, disease screening, diagnosis and identification, treatment results evaluation, and prognosis.
Today, imaging systems represent some of the most complex systems in the hospital. For example, the radiology information system (RIS) generates a queue and starts imaging devices to take radiographs of patients in sequence based on the information in the queue. The collected information is stored in the hospital's database by the picture archiving and communication system (PACS). Doctors then access patients' imaging files through PACS.
Hospital imaging devices are increasingly complex and precise, so image reading services must rely on medical staff with extensive expertise. The reading process can be time-consuming and complex. Therefore, doctors now need an easy-to-use imaging system to solve the following challenges.
Challenge 1: Hospitals Struggle to Store the Increasing Amount of Data
More advanced imaging devices produce higher quality imaging and take multiple shots in one go. In the past, one digital radiography (DR) image took up about 30 MB. Today, a set of PET-CT images includes anywhere between 2000 and 5000 shots and one image package can take up about 2.5 GB of storage space. Soon, we expect to see 3D images and videos with massive data, requiring even more storage space.
While this precision helps medical outcomes, hospitals are simply running out of storage space. Plus, they need to store the data over long periods of time for quality assurance. And, of course, hospitals must ensure database protection against ransomware to protect patient privacy and data.
Challenge 2: Traditional Networks Fail to Transmit Images
Doctors need to easily access the images they've taken and stored. However, regular networks are too slow — retrieving a set of PET-CT images can take up to 5 minutes, which is very inefficient for a busy hospital.
For example, before a surgery, the orthopedic department needs to carefully study many medical images and plan out the surgery. Normally, they would need to copy the images onto a flash drive from the imaging department and then head back to their hospital wing. Going back and forth can take upwards of thirty minutes at a large hospital, which is very time-consuming.
Challenge 3: Abstract 2D Images and Poor Communication
Regular 2D images can be very abstract. It takes a keen, experienced eye to evaluate them. If the doctor isn't experienced enough, they may misdiagnose the patient. Also, surgery plans are more difficult to discuss and prepare without having very detailed and clear information about the patient's condition.
In addition to diagnosis, 2D images hinder the communication between doctors and patients. It's difficult for patients to fully comprehend their condition and surgery plans. They simply have to trust the doctor to make the right decisions. Plus, during the pandemic, a lot of medical services went online. During online consultations, the traditional imaging system couldn't provide quick or precise imaging, affecting the quality of medical treatment.
The medical imaging system is extremely complex, involving multiple systems and terminals. We need systematic projects and end-to-end collaboration to supplement image reading. Addressing the three challenges discussed above, Huawei launched an all-optical medical imaging solution, which helps doctors optimize image reading, supporting smart hospital transformation.
Active-Active Data Center: Ensures Data Security and Accelerates Read Speed
The active-active data center is an important component of the all-optical medical imaging solution. It enables faster image reading and improves the user experience for other hospital systems. The data center offers the following features.
･ Service continuity: The solution assures that hospital services are available around the clock without latency, data loss, or single points of failure (SPOF). Patients can enjoy smooth registration, access imaging services, spend less time waiting to pick up medication, and rest assured that their medical records are safe and sound.
･ Simplified architecture: This solution integrates OceanStor SAN and NAS, costing 15% less than adding gateways to multiple storage devices.
･ Flash memory acceleration: This solution helps doctors access core data faster — they can retrieve images from PACS in less than half the time it once took.
･ Efficient backup: This solution integrates backup software, servers, and storage. A single storage system will be sufficient for the next five years, leaving room for future data surges.
All-Optical Hospital: A Hospital Information Highway, Facilitating Image Reading
A single fiber can enable a bandwidth of 10 Gbit/s for an all-optical hospital network. As technologies evolve, the bandwidth can be upgraded to 40 Gbit/s or even higher without redeploying optical fibers. This means that the network will no longer be a bottleneck for image reading. It takes only 0.4s to load a 1-GB CT group diagram, about the same time it would take a person to respond, so image reading becomes completely seamless. And it takes just 2.5 seconds to load a 2.5-GB PET-CT group diagram. Plus, the all-optical smart hospital solution offers the following features.
･ Simple architecture: The solution features a simple point-to-multipoint network architecture, so it's suitable for dense deployment in hospitals. Optical fibers can centrally carry all systems, including office computers, printers, videoconferencing systems, wireless backhaul systems, voice systems, information release systems, and room information points.
･ Flexible evolution: The all-optical hospital solution optimizes existing network cabling, building a highly scalable high-speed hospital bearer network for "one cloud, one network". Hospitals have unlimited capacity to upgrade the bandwidth with one optical network, making it perfect for future data evolution.
･ Simplified O&M: For O&M, the network administrator only needs to perform operations on the optical line termination (OLT). The optical network units (ONUs) in each room can be replaced and become available immediately, reducing the O&M workload and improving efficiency by 60%.
IdeaHub and eViewer: Building Image Reading Terminals for Better Experience
In partnership with eViewer, Huawei has developed improved image reading terminals, replacing 2D imaging.
･ Software: eViewer uses AI algorithms and image processing technologies to automatically generate 3D models and provide doctors with more comprehensive, clear, and vivid imaging.
･ Hardware: IdeaHub — eViewer's hard terminal — uses highly sensitive, low-latency, and ultra-HD 4K smart terminals, so multiple people can interact with images through touch control. This improves the efficiency and accuracy of medical information transmission. The smart terminals come pre-configured with intelligent conferencing, enabling remote video consultation.
eViewer is ideal for teaching-oriented ward rounds, delivering 3D images with rich details and making it easier for medical staff to diagnose and communicate medical knowledge. It helps young doctors learn and develop professionally, saving time and money on learning and training.
The solution can also help doctors discuss more complex cases and even supports remote consultations. eViewer delivers unique functions, such as touch interaction, 3D reconstruction, and annotation synchronization, providing a very clear display of injuries and nearby organs. Users can save measurements and annotations, and then access the information at any time.
In terms of doctor-patient communication, eViewer's unique 3D imaging helps patients and their families understand their condition, treatment plan, and risks even without medical knowledge. This ensures the patient's right to know while preventing medical disputes caused by misunderstandings.
Last but not least, eViewer facilitates remote consultation, including annotation through video, voice, image, and handwriting. Consultation comments can be archived by multiple parties.
As hospitals go digital, they need higher bandwidth and lower latency. This is particularly true from remote supervision, conferencing, imaging, and surgery as well as AR/VR healthcare. All these services require a minimum bandwidth of 200 Mbit/s. Remote surgery and AR/VR healthcare require a bandwidth above 1 Gbit/s and latency below 10 ms. Considering these requirements, Huawei's all-optical hospital solution offers a new way to build smart hospital networks.
In the all-optical hospital solution, passive optical LAN (POL) uses optical fibers as the main transmission media. Only passive optical splitters need to be deployed in weak-current rooms. Fiber to the room, desktop, and medical devices need a lot less cabling and do not require active access or aggregation switches in weak-current rooms. This makes the solution both cheaper and greener.
In addition, the link layer of the passive optical network (PON) supports secure and efficient AES128 encryption and PON link protection. This assures high bandwidth, low latency, security, reliability, and flexible expansion. Plus, it saves energy and reduces carbon emissions. We expect this to be the trend for smart hospital networks.
The International Data Corporation (IDC) predicts that the application of all-optical networks in the healthcare industry will grow at a compound annual growth rate (CAGR) of 47.5% from 2019 to 2024. The industry has released a series of all-optical hospital construction standards for Fifth Generation Fixed Network (F5G), including the Industrial White Paper of F5G All-Optical Application for Smart Hospital released by the Optical Network Alliance (ONA) and Zhuyitai and the Engineering and Technical Specifications of Passive Optical LAN published by China Engineering & Consulting Association.
Some pioneering hospitals have already deployed the F5G all-optical solution for commercial use, including the Puren Hospital affiliated to Wuhan University of Science and Technology (Wuhan Puren Hospital) and Huazhong University of Science and Technology Union Shenzhen Hospital (Shenzhen Hospital).
Take Wuhan Puren Hospital as an example. When building the new outpatient building, the hospital took several future trends into account. It has considered the changing functions of wards and consultation rooms in different periods and the increasing type and number of smart medical terminals.
Numerous terminals require network access and interconnection. Hospitals need an elastic and secure network to support rapid expansion and rollout of new services, massive data transmission, and future smart diagnosis and treatment services. Within this context, Wuhan Puren Hospital chose Huawei's F5G-based all-optical hospital solution. The solution uses optical fibers as transmission media and deploys optical fibers closer to information points for centralized bearing of all services in multiple scenarios through one optical network. With the all-optical network, Wuhan Puren Hospital has improved its medical imaging efficiency and built a high-speed logistics service channel based on the hard pipe architecture. This has improved patient experience and medical staff efficiency.
Shenzhen Hospital has also adopted the F5G solution to promote hospital informatization, building an "Internet+" healthcare model. It has just finished a new building in its international diagnosis center, which has adopted Huawei's all-optical hospital solution as the network infrastructure. This solution has improved the hospital network quality and supports IoT development. It features high bandwidth, low latency, and symmetric uplink and downlink frame structure configurations, helping improve the hospital's imaging system. The F5G technology has become a solid foundation for the smart development of Shenzhen Hospital.
Smart and all-optical hospitals are emerging and we believe that all-optical medical imaging will be one of the drivers for hospital service transformation.
