BEIJING, April 27, 2018 /PRNewswire/ — On the morning of April 21, the 2018 Vascular Innovation Symposium was held in Beijing. The symposium focused on conceptual and technological innovation, presented the innovative applications of next-generation information technologies such as big data and artificial intelligence in the medical field, and discussed the application of big data and artificial intelligence on vascular surgery.

Xie Shaofeng, General Director of the Department of Information and Software Services of the Ministry of Industry and Information Technology of the People’s Republic of China; Prof. Yu Qilin, Department of Surgical Medicine, People’s Liberation Army General Hospital; Prof. Wang Shenming, Chief Expert of the Department of Vascular Surgery, First Affiliated Hospital of Sun Yat-Sen University; Jiang Feng, Executive Vice President of the China Association for the Medical Device Industry; Wang Faqiang, President of Chinese Research Hospital Association; Shi Xinli, Head of Center for Medical Device Evaluation; Prof. Chen Zhong, Department of Cardiovascular Surgery, Beijing Anzhen Hospital; Prof. Liu Changwei, Department of Vascular Surgery, Peking Union Medical College Hospital; Prof. Guo Wei, Department of Vascular Surgery, People’s Liberation Army General Hospital; Prof. Liao Hongen, Deputy Dean of the School of Biomedical Sciences at Tsinghua University; Wu Wenxin, General Manager of the Intel Healthcare Industry Solution Group; Chai Xiangfei, CEO of Huiyihuiying; and Guona, COO of Huiyihuiying were among those that attended the symposium.

At the symposium, Huiyihuiying and the Department of Vascular Surgery of the PLA General Hospital (the 301 Hospital) jointly announced the AORTIST 2.0 artificial intelligence research cloud platform. According to Guo Wei, director of the Department of Vascular Surgery at 301 Hospital, it is the first time that an artificial intelligence automatic segmentation method for type B aortic dissection has been developed anywhere in the world.

Big Data and AI could let China pull ahead

Xie Shaofeng, Director General of the Department of Information and Software Services of the Ministry of Industry and Information Technology of the PRC, addressed the meeting, saying: To promote a smart future for the industry, medicine and health care need to be people-centric; positioning people’s yearning for a better life as the target while strengthening data, opening up data security and reinforcing cooperation between other aspects.

The first step is to promote openness and the sharing of medical health data for comprehensive utilization. In response to the development trends present in new technologies and new formats, we will build a unified, comprehensive and interconnected medical and health platform to achieve data and resource co-construction and sharing, eliminating data barriers and facilitating datasharing channels between industries.

The second is cooperation between health information technology companies and medical departments, strengthening the coordination ability between supply side and demand side, to promote big data and AI technology and make them demandoriented, going deep into the application of medical technology and promoting the transformation of technological achievements and industrial upgrade.

The third is to effectively ensure the safety of medical and health data, strengthen the management of identities in medical and health data, and construct a safety system, ensuring the safety of medical and health data through standard implementation of safety precautions.

The deep integration of big data and artificial intelligence with the health care field will optimize resource allocation, integrate social forces, and maximize medical and health benefits. It has great significance for propelling technological advancement, and building a better society for the welfare of the people.

Prof. Liao Hongen, Deputy Dean of the School of Biomedical Sciences at Tsinghua University, said that “With artificial intelligence becoming part of the national strategy, algorithms, computing power, and big data have developed rapidly. In recent years, artificial intelligence has made breakthroughs in disease monitoring, disease early warning, medication monitoring, personalized diagnosis, precise treatment, smart pension and so on. Especially in the field of microinvasive treatment, artificial intelligence has given medical research and clinics a powerful new tool.” When he shared on the topic of ‘Artificial Intelligence in Medical Imaging and Microinvasive Diagnosis and Treatment’, Prof. Liao stated that he thinks that with regard to AI, people should pay more attention to the efficiency of machines and the improvement of people’s effectiveness and cognition. That is not to say that people should let AI completely replace doctors, but that the best model is a complementary one. These kind of diagnoses require a high degree of long-term coordination and cooperation between even some fields merely related to medicine in order to ultimately achieve the goal of accurate personalized medicine.

As a leading medical imaging and artificial intelligence company, Huiyihuiying provides a complete set of medical image data analysis and mining tools and a medical imaging intelligent auxiliary diagnosis service. It has already been put into use in more than 700 hospitals and its intelligent auxiliary diagnosis system based on deep learning has already been applied in hundreds of hospitals.

The big data intelligent analysis cloud platform based on radiomics, machine learning, big data and other technologies has already been settled in more than 300 hospitals.

In the 2018 Vascular Innovation Symposium, Huiyihuiying and the Department of Vascular Surgery of the PLA General Hospital (the 301 Hospital) announced an artificial intelligence research cloud platform named AORTIST 2.0.

The full name of the AORTIST is “Artificial intelligence Online Research platform Targeting Individualized aortic Stent-grafting Therapy.” According to reports, after a year of cooperation, the two parties have made breakthrough progress in precision measurement, prognosis prediction and follow-up management on surgery of type B aortic dissections, and Guo Wei, director of vascular surgery at 301 Hospital, claimed this is the first time in the world that an artificial intelligence automatic segmentation method for type B aortic dissection has been developed. For Huiyihuiying, this is also a breakthrough, because its AI application for the first time ‘came out’ from the imaging department and began seeing application in clinical departments, also participating in clinical treatment decision-making after making breakthroughs in the diagnosis process, from efficiency improvement to precise treatment.

AORTIST2.0 gives intelligence to the diagnosis and treatment of aortic disease

Aortic dissection is the most dangerous and complex of the aortic diseases. With the evolution of medical equipment and surgical instruments, human beings have been able to treat aortic diseases with endovascular repair, achieving a major shift from massively traumatic aortic surgery to microinvasive methods. The main problem of the microinvasive treatment of aortic diseases has also begun to change from how to complete the surgery more safely to how to treat the disease more effectively. Questions such as ‘how to accurately measure aortic anatomic parameters to reduce the incidence of surgical complications and how to achieve prognosis of patients with aortic disease to develop a personalized treatment plan and follow-up plan have become larger concerns for vascular surgeons.

According to the data, the aorta is the main blood vessel of the body. If there is tearing of the intimal layer that cannot be treated promptly, then when it ruptures, it has a very high fatality rate. The dissection is located only in the area below the distal opening of the left subclavian artery, and the B-type aorta is not involved in the proximal aortic artery. B-type aortic dissection is a very rare but serious disease, where 65% to 70% die of cardiac tamponade, arrhythmia and so on, so it’s absolutely necessary to get early diagnosis and treatment.

According to Chai Xiangfei, Founder and CEO of Huiyihuiying, after the 301 Hospital and Huiyingying jointly designed the project, the hospital provided image data and patient clinical data, while Huiyihuiying provided the artificial intelligence algorithm, radiology algorithm, and follow-up engineered algorithm. Once the platform was fully trained and fleshed out, a limited verification was performed to debug for accuracy, before expanding the application to its full scope.

AORTIST resolves the three core issues of accurate measurement, prognostic prediction, and remote follow-up in previous B-type aortic dissection surgery. Taking precision measurement as an example, B-type aortic dissection surgery requires accurate measurements of the diameter of the proximal & distal landing zone, the location of the breach, and some important distance information. Manual measurement based on the CTA axial position is prone to errors, especially when it comes to the measurement of the diameter of the aortic arch, and it is also difficult to obtain information such as length and distance using the manual measurement method. In this case, vascular surgeons usually need to rely on trained professionals to use commercial software to obtain accurate anatomical parameters, but the accuracy and timeliness of information acquisition cannot be guaranteed.

The AORTIST Cloud Platform achieves breakthroughs in the three-dimensional reconstruction of patients’ aorta, segmentation, centerline extraction, breach analysis, and accurate measurement of diameter and length. According to reports, using the AORTIST cloud platform, the intersection-over-union ratio of arterial diameter measurement reached 98%, the arterial diameter error margin was reduced to within 1.5mm, and the accuracy was improved by more than 50% compared with the manual measurement method. The accurate measurement of the diameter, length, and distance between branching arteries of the patient’s landing zone could be completed within 10 minutes, greatly improving efficiency and accuracy, which is extremely helpful for doctors to develop personalized surgeries.

Prof. Guo Wei, Director of Department of Vascular Surgery at the 301 Hospital, said that “Radiomics is closely linked with big data and artificial intelligence.” Director Guo believes that the true strength of big data is not simply a large amount of data, but is actually big data paired with cloud-based data. The most important part of medical data is imaging data. Imaging data is often a manifestation of the objective response of the disease. It is more truthful than a patient’s statement and is more accurate than the medical record. Aortic disease is very suitable for big data, artificial intelligence and radiomics. There are many companies that make images today, but companies that extend it into automated segments are still uncommon. The most important part of a convolutional neuralbased network is automatic measurement. AORTIST 2.0 can assist in accurate measurement. Formerly most patients were manually measured before surgery. The difficulty was to find the plane which is perpendicular to the axis of the aorta, and then select the correct bracket. In addition, AORTIST 2.0 also provides a pre-plan recommendation, which is based on the analysis results of large amounts of data and will improve long-term outcomes.

Report from Japanese market in AI and radiomics

The best proof of the work related to the development of Radiomics and AI platforms has been confirmed during the 77th Annual Meeting of the Japan Radiological Society. During that meeting professor Shuji Yamamoto emphasized the peculiarity of Chinese development of Artificial Intelligence and the current global status quo in the Radiology field. Professor Shuji Yamamoto graduated from School of Medicine in Medical Physics from Osaka University, he is Project Associate Professor in Tokyo Institute of Technology (ACLS) and visiting researcher in the National Cancer Center in Japan, working on research into Cancer prevention and screening, as well as a member of the International Cancer Imaging Society (United Kingdom).

Japanese radiologists have been very interested in opportunities presented by the implementation of Radiomics in routine work and calculation of prediction and forecasting of Cancer. Even though the coverage of PACS systems in Japan exceeds 95%, the level of AI integration is still lacking – though there is great demand. This could be a great opportunity to develop the quality of patients’ data for companies like HY. So the market response for HY AI products was outstanding and the Japanese market, which boasts the highest adoption rates of MRI and CT scanners in the world, is regarded as the most interesting potential market for all companies working in Medical Imaging AI field.

The future of AI and radiomics in vascular surgery

Looking back over the history of the development of artificial intelligence, the earliest artificial intelligence was based on expert systems, and had turned into a statistical model by the end of the 1990s. AI gradually came into the limelight starting from 2012, when data-driven artificial intelligence emerged. In 2016, because of the spotlight shined on AlphaGo, artificial intelligence technology began appearing in a wide variety of fields and industries. Therefore, this came to be known as the “First year of AI”

Chai Xiangfei believes that images and pathology are the most mature technologies and the most widely used ones in the application of AI technology in medical field. In addition to intelligent analysis, AI medical imaging can provide new strategies for diagnosis and treatment. It can also realize intelligent segmentation of medical images to achieve 3D modeling and multimodal image registration. Providing doctors with multidimensional information and radiation-free image tracking techniques helps to reduce the intraoperative radiation damage for doctors.

Although medical images have achieved breakthroughs when it comes to type B aorta, there are still many challenges left to be solved. Chai Xiangfei said “The numbers of incidences of aortic disease in the country is probably 100,000. When we subdivide A and B aortic aneurysms, there are only thousands of cases for each disease, so we can not obtain data from millions of cases as we may wish. Usually, we can only perform calculations in hundreds or thousands of data sets. Handling artificial intelligence calculations in this process for such a small amount of data is the area where AI-enhanced medical imaging most urgently needs improvement.”

From the view of the medical industry, the specificity of this industry determines that many new technologies and new things cannot be “tried and tested” because it there are human lives on the line. At the same time, especially in the subdivision of medical imaging, entrepreneurs with cross-disciplinary abilities will have greater competitiveness. Entrepreneurs not only need medical clinical knowledge, but also computer knowledge and data processing knowledge, as well as needing to understand the market, sales, and methods of management.

As Chai Xiangfei pointed out at the end of the symposium, there is a real lack of such crossover talents in the medical imaging industry, which is one of the factors that restricts its development. Chai added “No matter how products or research might develop, or how commercialized the industry might become, in fact, it is very important to conduct multi-party collaboration, especially in the cultivation of interdisciplinary talents. During the process, no matter whether business, technology, or medical knowledge, they are all indispensable parts”.

Cision View original content:–aiming-to-make-efficiency-improvements-to-precise-treatments-300638084.html

Share this:

Facebook Comments

By Admin, AsiaFitnessToday

Hi! I'm Ursula Lee, writer and community manager at Asia Fitness Today. I love my food, my cat and my fitness regime :)

Creative Commons License
Except where otherwise noted, © on this site is licensed under a Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International License.