PereDoc Carves Out Differentiation in AI Medical Imaging with China's First AI-Powered Pulmonary Nodule MDT Clinic
Medical imaging is one of the fields most likely to see AI implementation, and 2017 was a year of concentrated breakthroughs for medical imaging + AI.
Riding the wave of AI, PereDoc (Beijing Qingyan Xiangyun Technology Co., Ltd., hereinafter referred to as “PereDoc”) entered the field of AI medical imaging in August 2017. After two years of development, it launched five major product lines: the PerePathos™ Intelligent Pathology Solution, the PereBrain™ AI Research Solution, the PereMagic™ Storage Compression Solution, the PereAir™ Cloud Service Solution, and the PereSmart™ Portable Device Series. These solutions have been deployed in more than 200 medical institutions across China.
Currently, PereDoc Qingyan Xiangyun has achieved seamless integration with mainstream PACS/RIS vendors. Selected as a strategic partner by China Electronics Data, the national team in health and medical big data, it has successfully connected to the national-level Medical Consortium platform. By combining artificial intelligence technologies with health and medical big data, PereDoc has launched a globally leading CT imaging screening cloud service, which has been successfully applied in scenarios such as low-dose lung CT health screenings, user health profiling, and cross-hospital intelligent follow-up.
In the closed and rigorous healthcare industry, two years is not a long time; however, for AI-based medical imaging, it has been sufficient to transition from a state of distrust to the threshold of commercial implementation.
In the new phase of AI medical imaging development, how is PereDoc exerting its efforts? At the 26th National Congress of Radiology of the Chinese Medical Association (CCR2019), Lian Jing, CEO of PereDoc, shared her insights with VCBeat (WeChat ID: vcbeat).
PereDoc’s core technical team, led by Chief Scientist Hu Fei, hails from Tsinghua University. Mr. Hu previously served as a specially appointed expert at the Ministry of Science and Technology–Tsinghua University Technology Innovation Research Center, and has led or participated in multiple national key research projects on medical artificial intelligence. He has extensive expertise in image recognition and medical imaging analysis. CEO Lian Jing brings rich experience across the healthcare, insurance, finance, and internet industries. Vice President He Tao holds a degree in Computer Science from the University of Trier in Germany and has long focused on data acquisition and application in the pharmaceutical industry, backed by solid theoretical foundations in technology and robust project management experience.
From the perspective of PereDoc’s core team, the decision to enter the AI medical imaging industry was driven by the healthcare sector’s need for AI.
Taking the radiological examination of pulmonary nodules as an example, in a densely populated city like Shanghai, a single physician may sometimes consult 100–200 patients per day. Each patient generates 200–300 CT images during the examination process, resulting in a daily workload of up to 40,000 CT images for the physician to interpret.
The lung nodule-assisted diagnostic solution developed by PereDoc enables AI to accurately identify nodules and their types within one minute. Furthermore, leveraging its proprietary patented technology, the system easily compares follow-up cases and tracks nodule progression, providing critical evidence for clinical diagnosis and treatment.
Lian Jing stated, “China’s medical resources are extremely unevenly distributed. Large hospitals are overwhelmed with patients; although their physicians are highly skilled, they struggle to meet the demand. Primary-care facilities face an even greater scarcity of senior specialists, and primary-care physicians bear substantial workloads. We must leverage technological means to empower healthcare, freeing physicians from burdensome routine tasks so they can devote more time to patient care and scientific research.”
From the perspective of technical applicability, with the maturation of artificial intelligence, big data, and deep learning technologies, both the speed and accuracy of image recognition have significantly improved. AI has gradually risen to the level of national strategy, and supported by relevant national policies, “AI + Healthcare” has emerged as a new trend. Moreover, the healthcare industry possesses vast amounts of data and standardized imaging records. In compliance with national regulations, medical institutions such as hospitals are required to retain imaging data for 30 years, which makes it possible to build rich, resource-abundant databases.
Although the introduction of technology can add weight to the side of disadvantaged doctors on the medical balance, the primary issue that AI technology must address to achieve genuine implementation in clinical settings is earning physicians’ trust.
Undoubtedly, to earn the trust of physicians, the foremost priority is to ensure the accuracy of the product. To this end, unlike medical imaging AI companies that primarily rely on open-source data, PereDoc adopted gold-standard imaging data from its inception, focusing on algorithmic accuracy while also addressing the product’s generalizability and scalability.
Lian Jing stated, “By leveraging the Joint Research Institute for Intelligent Healthcare and its extensive panel of experts, and through scientific research collaborations with leading hospitals, we have accumulated a wealth of medical imaging data, thereby earning recognition from experts in various blind tests.”
Based on extensive data, PereDoc has launched a suite of AI imaging products, offering comprehensive solutions across radiology, pathology, scientific research, and cloud services.
In independent testing conducted by CEC Data, competing products required over one minute per case for diagnosis. In contrast, PereDoc completes the entire process—from data ingestion to returning diagnostic results—within six seconds. This performance distinguishes it from industry peers and, as a flagship achievement among CEC Data’s dozens of application outcomes, marks a new breakthrough in real-time screening scenarios within the industry.
In optimizing physicians’ workflows, PereDoc has achieved seamless integration with leading domestic PACS/RIS vendors, whose products cover the imaging information systems of most hospitals, thereby reducing integration costs and accelerating market adoption. Meanwhile, the company has also achieved seamless integration with physician workstations, truly embedding its solution into clinicians’ daily workflows.
2019 was not an easy year for AI-based medical imaging. The AI industry experienced several ups and downs, and after addressing physicians’ trust issues, it still needed to tackle the challenges of commercial implementation.
In this regard, Lian Jing stated bluntly, “Currently, everyone is exploring business models for the practical implementation of ‘AI + healthcare.’ To achieve sustained success after the market matures, the biggest bottleneck is policy-related.”
Although policy barriers remain intact and the healthcare industry remains relatively cautious, PereDoc has made many beneficial attempts to accelerate the implementation of AI.
In 2018, PereDoc collaborated with Beijing Shijitan Hospital, Capital Medical University, to establish China’s first artificial intelligence (AI)-powered multidisciplinary team (MDT) outpatient clinic for pulmonary nodules. The AI-based imaging auxiliary diagnosis platform provides initial screening, the results of which are further evaluated by experts to formulate subsequent diagnostic and treatment plans. This approach enables early, detailed, and comprehensive detection of lung diseases, significantly improving diagnostic accuracy and treatment outcomes. This marked the earliest clinical deployment in hospital outpatient settings of medical imaging AI screening technology represented by deep learning.
In 2019, PereDoc was successfully selected for the Beijing Comprehensive Pilot Program for the Expanded Opening-up of the Service Sector. As a key enterprise in this program, it has been assigned dedicated service stewards established by both municipal and district-level governments in Beijing. These stewards facilitate government-enterprise coordination to dismantle policy barriers constraining industry development, with efficient governmental collaboration safeguarding corporate growth. Bottlenecks arising from industry regulatory policies lagging behind corporate development are expected to achieve policy breakthroughs through the Beijing Comprehensive Pilot Program for the Expanded Opening-up of the Service Sector. This enables enterprises to take the lead in establishing innovation centers with medical institutions or jointly building laboratories, thereby fostering the research and development, commercialization, and incubation of new medtech products and innovative business models for medical services.
Medical imaging is the area where AI can be most easily implemented, but it is not the end point of the integration of AI and healthcare. In the future, artificial intelligence technology will be widely used in the medical field.
In the future, PereDoc will expand and refine its product portfolio, striving to meet the demands of diverse application scenarios, particularly in specialized medical fields such as assisted diagnosis, precision medicine, medical imaging, and genomics. By leveraging artificial intelligence in healthcare, we aim to enhance the accuracy and efficiency of medical diagnoses. Through technology empowerment, we will extend high-quality medical resources to grassroots levels, enabling millions of people to access premium healthcare services. With advanced AI technologies, we are committed to supporting the "Healthy China" strategy.