Based on its core medical imaging and fluid dynamics simulation technologies, Yueying Technology is dedicated to developing auxiliary diagnostic and decision-making products for cardiovascular and cerebrovascular diseases to assist doctors in achieving more accurate diagnoses. Currently, the company has several products under development, including pulse imaging, silhouette imaging, and cranial imaging.
"We don't want to create Superman to replace the doctor, but rather to create armor to help the doctor become a superhero."
The R&D team contributed significantly, with 15 years of experience in medical image processing, including Jin Yueying.
According to the "China Cardiovascular Health and Disease Report 2019", there are 330 million people with cardiovascular diseases in my country, including 13 million stroke patients and 11 million coronary heart disease patients.
Backed by millions of patients and targeting the domestic demand for cardiovascular and cerebrovascular disease treatment, Yueying Technology was established in Shenzhen in October 2017. The company focuses on precision diagnosis in the cardiovascular and cerebrovascular field, centering on "artificial intelligence + medical imaging," providing a full-process auxiliary diagnostic and decision-making platform from diagnosis, preoperative planning, intraoperative navigation to prognosis, striving to improve the current cumbersome manual processes.
Entering the precision diagnostics market for cardiovascular and cerebrovascular diseases, Yueying Technology's confidence stems from a team of AI and fluid dynamics simulation talents. The company highly values independent research and development, with its R&D team divided into three main sections: a high-precision simulation team led by CEO Zhang Chao, a machine vision team led by Chief Scientist Wang Chunliang, and an engineering team.
The company's founder, Zhang Chao, holds a PhD jointly earned from the Department of Mechanical Engineering and the School of Medicine at Johns Hopkins University. He participated in the Hopkins Heart project, which involved medical imaging computational technologies, including CT-FFR. During his work, he accumulated expertise in simulation algorithms and computational functionalities, which laid the foundation for his decision to return to China and start his own business.
Since its establishment in 2017, Yueying Technology has attracted a group of overseas-returned PhDs and talents from domestic and foreign universities. At a time of technological change and product line expansion, Dr. Wang Chunliang chose to return to China full-time to serve as the chief scientist of Yueying Technology.
Wang Chunliang is an associate professor at the Royal Institute of Technology in Sweden and has 15 years of experience in medical imaging, making him a senior expert in the field. Speaking about his return to China, he expressed his hope to apply the technology and experience he has accumulated over the years to clinical practice, developing several products that can be used in real-world settings.
With a background in clinical medicine, Wang Chunliang has a deep understanding of clinical needs and believes that AI is a technology that can change clinical practice.
"A large portion of the information contained in medical images is wasted."
For example, a chest CT scan contains information such as lung volume and cardiac morphology. If doctors are willing to spend the time, they will certainly find a lot of information. However, in areas with limited medical resources, comprehensive screening of chest CT images can significantly increase the burden on doctors. Utilizing AI technology to automate the analysis of images can provide doctors with more accurate diagnostic information without increasing their workload, enabling a comprehensive assessment of the patient's condition and thus improving the overall level of medical care.
Fully automated CT image processing, using deep learning based on limited data.
FFR (fractional flow reserve) is the "gold standard" for the functional diagnosis of coronary artery stenosis.
Pressure guidewire measurement of FFR is commonly used clinically and is recommended as a Class I, Level A procedure in both the 2018 European Society of Cardiology Guidelines for Myocardial Revascularization and the Chinese Guidelines for Percutaneous Coronary Intervention (2016). However, traditional pressure guidewire measurement of FFR needs improvement in terms of safety, cost-effectiveness, and practicality, and its application rate in China is less than 1%.
CT-FFR can serve as an important alternative to invasive FFR, non-invasively measuring FFR values based on CT images to provide doctors with auxiliary diagnostic support. Globally, HeartFlow pioneered CT-FFR services, while domestic companies such as Keya Medical and Shukun Technology are also developing related products.
Yueying Technology seized the opportunity presented by the development of AI medical imaging and precision diagnosis to launch its CT-FFR product, Pulse Image. Based on coronary CT images, it uses deep learning to automatically analyze image data to calculate FFR values, thereby assessing the patient's vascular function.
The first stage of the Pulse CT-FFR product's function is 3D modeling, which transforms CT image information into a three-dimensional model of blood vessels. During this process, even a slight deviation in the vessel diameter can exponentially affect the blood flow calculation results.
Led by Dr. Wang Chunliang, an image processing expert with 15 years of experience, Yueying Technology uses image segmentation and graphics simulation technologies to accurately represent the morphology of blood vessels. Simultaneously, the introduction of artificial intelligence in this process significantly accelerates image segmentation and 3D modeling.
The second phase involves fluid dynamics simulation, which simulates blood flow based on a three-dimensional model of blood vessels. Dr. Zhang Chao and his research team developed a precise and rapid blood flow simulation technology and established an scalable and modular visualization platform to support the integration of information from multiple modules.
CT-FFR non-invasive measurement involves significant computational demands and requires advanced computer hardware. Traditional CT-FFR systems require hospitals to upload CT data to the cloud, where it is then processed by the company's servers, typically taking several hours or even a day to deliver the data to clinical departments. In contrast, PulseVision products can automate the entire CT image processing process directly in clinical departments, reducing the entire process to less than 10 minutes.
This not only reduces the burden on doctors and shortens the time for FFR measurement, but also increases the enthusiasm for clinical application of FFR technology. Since doctors already have CT images, if they can obtain FFR data effortlessly without having to perform additional image acquisition operations on patients, they will be more willing to use the CT-FFR system.
Furthermore, Chinese hospitals have very strict control over data security. Following this trend, in the future, companies will only be able to collect and manage data if they deploy auxiliary diagnostic technologies in hospitals.
Furthermore, the AI algorithm developed by Yueying Technology only requires a small number of high-quality computational examples to complete the construction of a high-precision convolutional neural network architecture, greatly reducing the reliance on raw medical data. Currently, many deep learning methods offer an end-to-end learning paradigm, where the entire learning process is entirely handled by the deep learning model, which directly learns the mapping from raw data to the desired output.
Taking cerebral hemorrhage as an example, depending on the location of the hemorrhage, it can be classified into subarachnoid hemorrhage, intracerebral hemorrhage, and intraventricular hemorrhage, among others. Only by presenting various types of hemorrhage cases to the neural network in advance can it accurately identify different types of cerebral hemorrhage. This results in the neural network being highly data-dependent and potentially unable to identify sporadic or infrequent cerebral hemorrhage lesions.
This approach differs from the learning process of doctors. Doctors learn the normal anatomical structure of the human body before diagnosing diseases based on abnormalities. Yueying Technology's deep learning network borrows from this learning path, starting with healthy, normal human anatomy and then making a comprehensive judgment based on abnormalities, thus reducing reliance on raw data. In this way, even if a patient presents with abnormalities, such as large-scale calcification or the placement of a stent, Yueying Technology's products can accurately segment blood vessels.
According to clinical application data, the accuracy rate of the Pulse Imaging product in measuring FFR values is 90%, and the accuracy rate is even higher for patients in the FFR gray zone (i.e., FFR between 0.75 and 0.80).
Based on 3D modeling and fluid dynamics simulation technologies, Yueying Technology continues to develop new products that integrate morphological and functional information of the cardiovascular system. These new products will not only be able to determine vascular stenosis rates and blood flow conditions, but also reflect the plaque and its structural composition around blood vessels.
"First you have the target, then you get the gun," expanding into two major product lines: cardiac and intracranial.
Yueying Technology focuses on the field of vascular imaging, primarily developing diagnostic aids and diagnostic decision-making products.
The company has two product lines. One is the development of cardiac imaging diagnostic products, such as coronary artery morphology analysis platforms, diagnosis and preoperative planning for structural heart disease, and preoperative planning for atrial fibrillation radiofrequency ablation. The other is the development of intracranial functional diagnostic products, not only targeting intracranial arteries, but also intracranial multimodal image fusion.
In terms of product strategy, Yueying Technology accurately grasps clinical needs, adopting a "target first, then gun" approach, focusing on in-depth development in the cardiovascular and cerebrovascular field. Recognizing the criticality and importance of cardiovascular and cerebrovascular diseases, Yueying Technology has placed its emphasis on cardiovascular and cerebrovascular imaging. The company plans to expand its business into the peripheral vascular field next.
During its development, Yueying Technology has partnered with Peking University Third Hospital, Fuwai Hospital, and Anzhen Hospital to jointly conduct clinical trials of its Pulse Imaging CT-FFR product. Xuanwu Hospital and Huashan Hospital affiliated with Fudan University have also been designated as centers for the product's implementation. In overseas markets, the company collaborates closely with the Department of Cardiology at Stanford University School of Medicine and Johns Hopkins University to conduct multi-center clinical trials globally.
In the future, Yueying Technology hopes to build a cardiovascular and cerebrovascular auxiliary diagnosis and decision-making platform by connecting cardiovascular and cerebrovascular products, truly replacing doctors' existing diagnostic and treatment methods.
