As the initiator and organizer of the International Symposium on Image Computing and Digital Medicine (ISICDM), Professor Li highlighted the pain point of “the integration of science, technology and medical science”. The combination of medical science and technology, production and study will propel the collaborative innovation and implementation only by cooperation and exchanges among experts with interdisciplinary backgrounds.

It also brings about new opportunities and challenges in the analysis of medical images with the AI rush.

It is absolute for medical and IT communities that difference in profession makes one feel worlds apart. Various problems caused by different minds, judgments and fuzzy leadership when two groups decide to design a product together. AI-engergized Medical Care is actually in the initial stage of the whole market. Players from diverse fields all have access to the market one after another, whereas medical AI products are still in the research phase.

In the context, ISICDM 2019, jointly hosted by ISDM and Tianyuan Mathematical Center in Northwest China, sponsored by the School of Mathematics and Statistics of Xidian University and the School of  Biomedical Engineering of Fourth Military Medical University, was held in Xi'an from August 24th to 26th, 2019.

Honorary chairmen of ISICDM 2019 include Professor Xu Zongben of Xi'an Jiaotong University, the member of the Chinese Academy of Sciences (CAS), and Professor Yu Mengsun of the Air Force Military Medical University, the academician of the Chinese Academy of Engineering (CAE). Professor James Duncan of Yale University, the founder of MICCAI, Professor Terry Peters of the Royal Canadian Academy of Sciences, Professor Liu Shiyuan, Director of the Department of Imaging and Nuclear Medicine of Shanghai Changzheng Hospital, all appointed as chairmen in the forum.

This seminar invited scholars from home and abroad in the fronts of information science, mathematics and medical science to give special reports. It also organized a group of professionals in science and technology, as well as doctors equipped with rich research and clinical experience to popularize the science and make cutting-edge academic reports. The Symposium fully embodies the interdisciplinary features and promotes the exchanges and cooperation of research results.

At present, AI technologies are widely used in biomedical engineering, but he certification is required and prerequisite for AI-assited medical care.

In September 2017,the National Institutes for Food and Drug Control issued a new edition of the Medical Device Catalogue. It defines the software for diagnosis, sets standards for companies in medical imaging and puts up with new requirements to promote the normalization of the industry. In the Medical Device Catalogue, medical AI with strong momentum recently is listed in "Medical Software" among 22 catalogs.

With the increase of startups in medical AI, higher requirements have been brought on the development and regulation. Professor Ren Haiping was invited by ISICDM to deliver a speech of Quality Assessment Trend of Medical AI.

She pinpointed that even though there are rich applied scenarios in medical AI, dynamic quality evaluation caused by rapid iteration results in great pressure.  Food and Drug Control system is preparing for rules and regulations, drawing up a universal standard (ISO/IEC WD 22989 Artificial intelligence - Concepts and terminology). The standardization of AI-assisted medical devices in China is on the right track, and the quality assessment of clinical application of AIMD should be performed at the same time.

In the roundtable forum, pioneers from fields as medicine, mathematics, and business are invited to discuss “Medical AI in Post-Deep Learning Era” in the interview. Lu Yao, the professor and doctoral supervisor in Sun Yat-sen University and the chairman in PVmed, attended the forum. Excerpts from Q&A session:

Q: The neural network has been hailed from the beginning to what was once called the ‘neuropathy network’, and now it is popular again, how do you think about this trend?

A: I'll explain it from the perspective of mathematics since its my major. The essence of neural network mapping is actually very simple. In my opinion, there is no 'pathy' in neural network. The only doubt is the lack of data training. It's impossible for doctors in the institute to have access to much data for training. But what our company's doing now is to "train" our product so as to assist doctors and predict problems in clinical practice.

Q: The AI-assisted medical imaging is now in the heyday. As an adept, what will you consider opportunities and challenges confronted with?

A: AI-assited medical imaging is meant to facilitate doctors. It represents an opportunity as well as a challenge for us to concern demands, pain points and to solve problems of doctors.

Q: Please look forward to the development of AI medical care in the coming five and even ten years?

A: I think AI is just a tool from researches to production, but I hope AI can be fully applied into clinical practice. It absolutely will become the most convenient and popular scalpel for doctors.

Medical care itself is a technology-intensive industry with strict thresholds for supervision. AI products require a large amount of high-quality data and experts in medical field to achieve data labeling and quality control, which are more rigorous than that of other industries. On the other hand, medical science poses challenges of data collection and management, algorithm verification, standardization of clinical applications since it's evidence-based.

Different from other startups that rely only on technologies, PVmed recruits a team of professionals with more than 20-year experience in medical images and AI led by Professor Lu. PVmed acts as the pioneer and supplier of technologies, equipped with multi-modality image databases in organs of head and neck, lungs and breasts.

Q: The AI-assisted medical imaging is now in the heyday. As an adept, what will you consider opportunities and challenges confronted with?

It has developed solutions of intelligent radiotherapy contouring, medical image processing and other medical AI products, which focus on doctors' realistic needs and difficulties from design to practice.

PVmed's exhibits received great attention at ISICDM 2019. Professor Xu Zongben, the academician of CAS, and Professor Li Chunming, the organizer of ISICDM visited PVmed's booth for more details, and then they took a group photo with Professor Lu.

At present, PVmed boasts the advanced strengths in medical AI of image processing. It has built cooperative relations with National Cancer Center Singapore (NCCS), Sun Yat-sen University, Chinese PLA General Hospital and other medical institutes around the world, enjoying highly appraisal from hospitals, image device suppliers, medical software developers, inspection organizations and  other partners in medical ecosystem.

PVmed will continue to remain true to its original aspiration of ‘Assisting Doctors & Serving Medical Care’, and fulfill demands from patients and doctors, forming a strong synergy, breaking industrial barriers and energizing medical AI in the coming years.