Grand release! "Blue Book on the Current Development and Future Trends of AI Medical Imaging Industry" (with full-text access methods attached)

In the field of AI medical imaging, software developers are located upstream of the industrial chain. They use advanced algorithms to process large amounts of medical data, develop software systems for specific application scenarios, and achieve commercial transformation through integration with self-produced hardware, cooperation with other device manufacturers, or direct software sales. Established medical hardware manufacturers in the midstream, considering the high cost of building their own AI software development platforms, often choose to join forces with software companies to jointly promote imaging technology innovation. At the downstream of the industrial chain, AI medical imaging software is widely used in healthcare, education, research, and third-party diagnostic institutions, significantly improving the efficiency of disease identification and assessment.

Source: Frost & Sullivan report

With the continuous development of medical imaging technology, its observation accuracy has continuously deepened from the tissue level to the molecular, cellular, and even chromosomal levels. Traditional histological imaging provides a macroscopic understanding of the overall structure and morphology of diseases. The development of advanced staining and imaging techniques enables more precise observation and analysis of changes at the cellular and chromosomal levels.

Medical imaging, pathological diagnosis, and chromosomal testing technologies can all be organically integrated with AI algorithms to improve diagnostic efficiency, enhance diagnostic accuracy, and contribute to the development of the health industry.

The commercialization paths of AI medical imaging software enterprises in our country mainly include direct sales as medical devices and charging service fees based on diagnostic volume or usage. From the current development of the industry, AI medical imaging products are primarily introduced into medical institutions in the form of medical devices. Looking ahead, with continuous enhancement of product functionality and further expansion into the primary healthcare market, the industry will also attempt to adopt a split model based on medical service charges, that is, billing according to diagnostic volume or usage. In addition, the government actively collaborates with AI medical enterprises to optimize regional medical services and improve the public's healthcare experience. The aim is to use AI technology to provide more efficient and personalized medical services, while promoting the balanced distribution and efficient utilization of medical resources within the region.

Source: Frost & Sullivan report

The increasing medical demand brought about by China's large population base, the massive accumulation of medical data due to digitization and informatization, combined with factors such as regional development and uneven personnel mobility, have collectively led to a shortage of medical resources in China and a serious imbalance in their distribution.

According to the China Statistical Yearbook, in 2022, there were over one million various types of hospitals in China, with primary healthcare institutions accounting for nearly 95%. The number of practicing (assistant) physicians in China exceeded 4 million. However, given China's large population base, there are less than 4 physicians per thousand people on average, and this number is even lower in rural areas, with only 2.5 physicians per thousand people. There is a clear shortage of imaging, pathology, and ultrasound physicians in China. Based on the '2022 China Health and Wellness Statistical Yearbook', figures from the National Bureau of Statistics, and combined with industry expert opinions, it is estimated that in 2021, there were only 0.17 imaging doctors per thousand people; there were only 20,400 registered pathology physicians in China, while the actual demand exceeds 100,000; there were only 120,000 registered ultrasound physicians in China, while China has over 200,000 ultrasound devices. The shortage of high-quality medical resources, uneven distribution, and unreasonable layout structure affect the fairness and efficiency of healthcare services, and also make the task of major disease prevention and control arduous and difficult to cope with in the event of public health emergencies.

To alleviate this contradiction, China actively promotes the development of smart healthcare. It supports the development of smart healthcare, telemedicine, and data healthcare at various levels such as infrastructure construction, product approval, and promotion and application, to facilitate the intelligent upgrading of medical equipment. It aims to improve the manufacturing level of digital medical devices, Internet of Things devices, smart health products, and other equipment, as well as promote the upgrading of the smart healthcare equipment industry. At the same time, relevant policies have been introduced to clarify the management norms for AI medical software, strengthen data security and privacy protection, encourage medical institutions to adopt Internet technology, achieve resource linkage up and down, information sharing, and efficient collaborative operations, advance the hierarchical diagnosis and treatment system, advocate the 'grassroots examination, superior diagnosis' model, with the aim of optimizing the distribution of medical resources and improving the efficiency and quality of medical services.

As of June 2024, China has had 92 artificial intelligence medical imaging diagnostic software approved for Class III medical device certificates by the National Medical Products Administration (NMPA). These include cardiovascular diseases (27), pulmonary diseases (24), cerebrovascular diseases (13), orthopedic examinations (10), fundus diseases (9), breast diseases (2), and more. These products cover multiple disease areas and various types of equipment, capable of improving image quality, lesion recognition and reconstruction, disease staging and grading, disease progression prediction, and other functions. They enhance medical diagnostic efficiency, reduce the workload of doctors, and promote the balanced distribution of medical resources.

Source: Frost & Sullivan report

According to data analysis from MedMolecule, from January 2014 to June 2024, there were more than 166 financing events in the domestic AI imaging diagnosis field, with a financing scale reaching one hundred billion yuan. Since 2022, primary market investments have shown a more cautious trend, with companies such as Yizhun Intelligence, Shenrui Medical, Shukun Technology, Fosun Xingmai, United Imaging Intelligence, TuiXiang Medical, and Desheng Biotech emerging in the industry through extensive and in-depth business layouts.

The contradiction of insufficient and uneven distribution of medical resources continues. In the future, with the stabilization and recovery of the Chinese economy, continuous breakthroughs in AI diagnostic technology, and further facilitation of commercial pathways, market performance is expected to recover.

In terms of policy, regulatory authorities have clearly indicated their policy orientation towards encouraging the development of artificial intelligence and advancing smart healthcare, as well as promoting hierarchical diagnosis and treatment to balance medical resources.

In March 2009, the State Council issued the 'Opinions on Deepening the Reform of the Medical and Health System', which clarified that the main goal of healthcare reform is to establish a basic medical and health system that comprehensively covers urban and rural residents, providing safe, effective, convenient, and economical medical and health services. To achieve this goal, China has successively introduced a series of reform measures, including the hierarchical diagnosis and treatment system, the construction of medical consortia, and 'Internet + healthcare'.

In the wave of healthcare system reform, the application of AI technology has undoubtedly played an extremely important role. AI can not only improve the operational efficiency and service quality of healthcare systems but also effectively narrow the technological and service gaps between different hospitals. For example, AI diagnostic and treatment technologies can help primary care doctors improve diagnostic accuracy and work efficiency. At the same time, with the help of hospital information facilities and internet platforms, remote medical systems and medical data sharing mechanisms have been implemented. This has greatly strengthened communication and cooperation between leading hospitals and primary healthcare institutions. It not only makes medical information transmission more convenient but also demonstrates great potential in medical decision-making, remote medical services, etc., thereby contributing to the rational allocation of medical resources and providing solid technical support for the smooth progress of healthcare reform.

In addition, regulatory authorities have also introduced guiding policies for multiple aspects of AI medical products, including clinical trials, review and approval. While generally encouraging innovation, these policies regulate disorderly competition in the industry and guide the direction of real and effective AI medical product innovation. For example, the 'Guiding Principles for the Registration Review of Artificial Intelligence Medical Devices' issued in 2022 has standardized the registration declarations for Class II and Class III artificial intelligence standalone software and medical devices containing artificial intelligence software components.

With the continuous influx of capital, many enterprises are joining this field. They cooperate closely with medical institutions, academia, etc., continuously refining algorithms and expanding application areas, promoting the application of artificial intelligence in the medical field. Especially, significant progress has been made in image enhancement, lesion recognition, 3D reconstruction, and automated reporting. AI medical imaging algorithms are expected to cover more diseases and solve cross-device data compatibility issues, thereby comprehensively improving medical quality and efficiency.

1) Pulmonary diseases

Currently, the AI medical imaging software for lung diseases listed on the domestic market mainly covers diseases such as pulmonary nodules, pneumonia, and tuberculosis. Especially in the early screening of pulmonary nodules, AI plays an important role in nodule identification, benign and malignant judgment, pathological classification, and follow-up due to its rapid interpretation ability, which has been recognized by industry experts. At the same time, AI has achieved a high accuracy rate in the diagnosis of pneumonia and tuberculosis, and relevant products have been launched for application. AI research on diseases such as pneumothorax and emphysema is also underway, demonstrating the potential for comprehensive application of AI in the field of lung diseases.

2) Breast diseases

Breast diseases, especially breast cancer, pose a significant health threat to women in China. To protect the health of women, departments such as the National Health Commission of China have vigorously promoted the 'two cancers' screening program for breast cancer and cervical cancer. Artificial intelligence technology has been implemented in multiple key areas such as breast ultrasound, mammography, and breast pathology analysis, significantly improving the accuracy and efficiency of screening and providing strong technical support for the early detection and treatment of breast diseases.

3) Cardiovascular diseases

Cardiovascular diseases have become a major threat to the health of Chinese residents, ranking among the top causes of mortality. Artificial intelligence has great potential in the field of cardiovascular diseases, capable of performing functions such as coronary artery stenosis assessment based on coronary CT angiography, calculation of flow reserve fraction and quantitative flow fraction, automatic interpretation of electrocardiogram data, and assisting ultrasound doctors in obtaining high-quality images. This provides patients with timely and effective treatment plans.

4) Cerebrovascular diseases

Cerebrovascular diseases can be divided into acute cerebrovascular diseases (stroke) and chronic cerebrovascular diseases (such as unruptured intracranial aneurysms, atherosclerosis, etc.). Since cerebrovascular diseases often occur rapidly and the treatment window is narrow, rapid diagnosis and classification of the disease are crucial. In the diagnosis and treatment of cerebrovascular diseases, artificial intelligence technology can accurately and quickly automatically segment and quantify the hemorrhagic or ischemic areas of stroke, while assisting in detecting complex lesions such as intracranial aneurysms and head and neck artery stenosis, thus saving precious time for the diagnosis and treatment of cerebrovascular diseases.

5) Orthopedics

With the acceleration of the aging population process, the number of patients with fragility fractures caused by osteoporosis is increasing year by year, becoming an increasingly focused area of public health. In orthopedic diagnosis and treatment, artificial intelligence technology is playing an increasingly important role, especially in the immediate analysis and three-dimensional reconstruction of fractures, significantly improving diagnostic efficiency and accuracy. Moreover, artificial intelligence technology has also been widely applied in the field of pediatric bone age assessment, helping to monitor and intervene in children's abnormal bone development in a timely manner and ensuring their healthy growth.

6) Fundus diseases

Fundus image interpretation based on fundus cameras not only plays a key role in the detection of ophthalmic diseases such as diabetic retinopathy, glaucoma, and macular degeneration, but also indirectly reflects the status of systemic diseases through subtle changes in the fundus blood vessels, such as early signs of diabetes, hypertension, chronic kidney disease, and even Alzheimer's disease. Fundus image analysis algorithms based on artificial intelligence technology have been applied in multiple fields of fundus images, enabling rapid and accurate identification of multiple diseases.

7) Other diseases

In multiple medical fields, including thyroid diseases, digestive tract diseases, kidney disorders, oral health, cervical cancer pathological screening, and chromosomal karyotype analysis, artificial intelligence technology is demonstrating its wide applicability and profound impact. Through precise analysis, assisted diagnosis, and personalized treatment recommendations, AI not only accelerates the process of disease detection but also improves diagnostic accuracy and treatment effectiveness, providing patients with more comprehensive and refined medical care.

Yizhun Intelligence ® is a technology company dedicated to artificial intelligence-assisted medical imaging diagnosis. It is the world's first AI company to apply both two-dimensional image detection technology and three-dimensional dynamic video real-time detection technology to the field of medical imaging, and it is also the industry's first national-level specialized, refined, characteristic, and innovative 'little giant' enterprise recognized by the Ministry of Industry and Information Technology. In the 'Technical Expert Consensus of the China Cancer Screening Center', Yizhun Intelligence ® is an AI technology provider that fully meets the technical requirements and functions proposed by the expert consensus.

Shanghai Xingmai Information Technology Co., Ltd. is an internationally leading and domestically first-class medical artificial intelligence enterprise established in Shanghai in 2018. It was co-incubated by Fosun International and Fosun Pharma, and since 2022, it has introduced state-owned assets, industry, and capital shareholders such as Jiading Anting, Chongqing Banan, and Yaojin Capital. The company was awarded the title of National High-tech Enterprise in 2019 and recognized as a 'Specialized, Refined, and Innovative' enterprise in Shanghai in 2023. Fosun Xingmai has achieved a multi-department layout including 'radiology, pathology, ultrasound, cardiology, respiratory medicine, orthopedics, neurology, and thyroid and breast surgery.' Its product matrix covers multiple fields such as imaging, pathology, and ultrasound, including 5 imaging modalities and 8 systematic diseases. The technology is at the international leading level, adhering to the strategic route of high-quality development and strengthening grassroots healthcare. It collaborates deeply with top medical centers to shape future intelligent healthcare scenarios.

Shenrui Medical is a national high-tech enterprise established in March 2017. It has independent operating companies in Beijing, Shanghai, and Hangzhou, dedicated to providing multi-modal general medical AI solutions for medical institutions at all levels through computer vision (CV), natural language processing (NLP), and new-generation artificial intelligence technologies such as large models and AIGC, helping them achieve high-quality and high-level development. Shenrui Medical has completed 7 rounds of financing in 5 years and has now become a leading enterprise in the field of medical artificial intelligence in China. In recent years, Shenrui Medical's business has achieved explosive growth, with products distributed across more than 30 provinces and cities across the country, providing thousands of medical institutions with full-link artificial intelligence services ranging from AI-assisted diagnosis, intelligent screening, clinical decision-making, patient services to medical big data governance, scientific research, doctor training, capacity building, and more.

Airdoc was established in September 2015, dedicated to providing comprehensive and multi-faceted artificial intelligence solutions for early screening and management of chronic diseases, myopia prevention and control, etc. It is a global leader and pioneer in the field of retinal image artificial intelligence, as well as a national-level specialized and innovative "little giant" enterprise. On November 5, 2021, Airdoc successfully listed on the Hong Kong Stock Exchange, becoming the "first stock in medical AI." Airdoc's diabetes retinopathy auxiliary diagnostic software Airdoc-AIFUNDUS was the first to obtain a Class III medical device certificate from the National Medical Products Administration among similar products. Moreover, Airdoc is actively expanding overseas markets; its Airdoc-AIFUNDUS has been approved for the EU CE MDR certification, allowing it to fully enter 27 EU countries including Belgium, Germany, Italy, France, etc.; while its portable fundus camera has received regulatory approval and registration in Indonesia, South Africa, Saudi Arabia, the United Arab Emirates, and Thailand. At the Geneva International Invention Exhibition, Airdoc's portable fundus camera and myopia treatment products won the "Special Commendation Gold Award." As of the end of 2023, Airdoc's retinal image artificial intelligence products have been distributed across more than ten countries on five continents, serving nearly 30 million users in total.

Hangzhou Deshi Biotechnology Co., Ltd. (hereinafter referred to as 'Deshi') was established in 2016, focusing on the core technological innovation of medical imaging artificial intelligence. The company has independently developed a globally leading general medical imaging large model, which has been trained on over 1000 million medical imaging data sets, covering 10 clinical specialties including X-rays, CT scans, MRI scans, pathology, and chromosome analysis. The implementation application of chromosome karyotype analysis based on the medical imaging foundation large model combines AI imaging core technology with precision optical equipment technology, breaking and surpassing foreign monopolies. It has raised the industry accuracy benchmark from 50% to 99.45%, shortened the time for clinical reports to be issued from nearly 30 days to 4 days, established a new paradigm for global chromosome karyotype analysis, obtained international certification for the first (set) of equipment, and been rated as having core technology leading internationally by a national expert group. It has been sold into more than 400 top-tier hospitals across the country.

DeShi will continue to increase the parameter volume of its self-developed medical imaging backbone large model. In five specific medical imaging scenarios, it aims to completely break foreign enterprises' technological monopoly and achieve overtaking on a curve. DeShi Biotechnology is a national-level specialized, refined, characteristic and innovative 'little giant', national technology-based enterprise, and national high-tech enterprise. It holds more than 60 core independent intellectual property rights in China, the United States, Europe and other places. Its scientific and technological achievements have been widely recognized. The core products of its medical imaging artificial intelligence large model applications have obtained a total of 27 certificates including Chinese NMPA registration certificate, EU CE certification, and US FDA registration certificate. It has won a series of honors such as the first prize of the Scientific and Technological Achievement Award of the China Birth Defect Intervention and Rescue Foundation, and ranked first in the reproductive field on the list of 'Artificial Intelligence Medical Devices' by the Ministry of Industry and Information Technology of China.

The 'Blue Book' introduces some companies in the AI medical imaging product industry, including company profiles, business segment layouts, listed products, competitive advantages, and technology platforms. For more detailed content, please refer to the full report.