Huiying Medical's Radcloud Completes Four-Year Deployment, Establishes Clinical-Research-Clinical Closed-Loop Ecosystem

From April 23 to 25, the 2021 China Hospital Information Network Conference (CHINC) was grandly held at the Hangzhou International Expo Center. As the inaugural year of the “14th Five-Year Plan,” this CHINC conference, themed “Digital Health: Co-construction and Sharing,” brought together nearly 50,000 participants from health authorities, hospitals, research institutions, and emerging technology enterprises to jointly explore new approaches to smart hospital development and share latest achievements.

Value Mining of Medical Big Data Has Been a Key Topic at Recent CHINC Conferences. How to Store and Collect Standardized Medical Data and Empower Clinical Practice Has Remained a Focus for Both the Industry and Government in Recent Years.

Intense demand has given rise to numerous challenges, such as: How to manage the massive volumes of continuously generated data? How to integrate fragmented and siloed data? How to address the technical barriers in mathematics and algorithms faced by physicians in scientific research? Under these circumstances, the vision of establishing a closed-loop system linking clinical practice, research, and back to clinical practice is encountering significant obstacles.

Returning to CHINC after a one-year hiatus, HY Medical presented its solution to the aforementioned challenges: the Radcloud® Big Data Research Platform. In this era of big data-driven medicine, HY Medical leverages AI to assist physicians in uncovering the secrets hidden within the pixels of medical images.

Radcloud®, an intelligent big data analytics platform for scientific research, was launched four years ago. Driven by imaging data and integrating deep learning with radiomics methodologies, the platform incorporates multi-modal imaging data along with non-imaging datasets—including genomic, proteomic, pathological, and follow-up data—to enable effective management and mining of latent imaging features, thereby unlocking data value. It assists physicians in conducting high-quality research efficiently, facilitating the publication of high-impact papers, enhancing institutional research capabilities, and strengthening disciplinary expertise.

Specifically, Radcloud® has introduced a medical imaging big data analysis method developed on the basis of big data analytics and high-throughput computing. It provides researchers with a one-stop healthcare big data research solution that leverages imaging expertise while integrating multidisciplinary approaches from clinical medicine, pathology, and other fields, marking a significant step forward in the realm of intelligent scientific research.

By extracting feature values from regions of interest, Radcloud® converts imaging information into digital data, generating a database of feature values that describe signal intensity, lesion shape, lesion margins, and lesion texture characteristics within the regions of interest. Combined with image-related clinical information, this approach facilitates research into clinical issues such as lesion type, disease grading, and prognostic treatment outcomes in patients.

To date, it has provided one-stop intelligent research services to over 700 hospitals and nearly 50,000 physician users both domestically and internationally, boosting research efficiency by 80%. It has successfully secured more than 20 research grants, including those from the National Natural Science Foundation of China. The acceptance rate for high-quality papers has reached 80%, with achievements recognized at top international radiology conferences such as SCI, RSNA, and ISMRM.

Behind the widespread user recognition lies HY Medical’s continuous and meticulous refinement of Radcloud®.

Over the past four years of productization and commercial implementation, Radcloud® has undergone six major iterations. HY Medical told VCBeat that scientific research based on medical big data is grounded in clinical practice, which is a complex and dynamic process. During the refinement of its big data platform, the company maintained continuous communication with physicians to clarify the data analysis characteristics of different clinical trials, learn the operational habits of various doctors, and maximize the platform’s personalization capabilities accordingly. In summary, the advantages of Radcloud® can be categorized into three aspects.

In addition to core features such as zero-code, one-stop intelligent research analysis and full-process visualized operations, Radcloud® offers a data de-identification mode and seamlessly integrates with PACS, RIS, and HIS systems. It provides leading support for the management and analysis of diverse medical data types, including multi-modal imaging, case records, follow-up data, and text, while conveniently generating English analytical reports and manuscripts. With simple mouse clicks, users can easily perform rapid, intelligent joint analysis of imaging data alongside clinical, pathological, and genetic data, significantly streamlining the research workflow.

Radcloud®’s ability to consistently generate high-impact publications stems from its dual advantages in technology and service. By integrating over 50 cutting-edge algorithms in deep learning, machine learning, and radiomics, and featuring more than 1,000 built-in feature analysis tools, the platform significantly expands the dimensions of research data analysis. This enables physicians to uncover insights that are beyond the reach of human visual perception and empirical summarization. Furthermore, HY Medical has assembled a dedicated team of nearly 50 PhD-level experts with over a decade of experience in medical image processing. Providing specialized, project-specific research support, this team safeguards physicians’ scientific endeavors, facilitates the efficient translation of research projects, and helps produce high-quality scientific outcomes.

Radcloud® enables flexible control of project permissions, meeting diverse research needs such as individual and departmental studies, while also supporting major multi-center collaborative projects at hospital and regional levels. It enhances the research capabilities and output of individuals and departments, and promotes comprehensive data management for multiple diseases within departments.

Meanwhile, Radcloud® leverages technologies such as blockchain and homomorphic encryption to provide multi-layered security for data interoperability in multi-center collaborative research on major projects. Its robust data middle-platform architecture enables the platform to respond efficiently and rapidly to user needs, delivering high-quality services and fully unlocking the exceptional value of data. This facilitates regional patient data management and scientific research output, fosters collaboration and knowledge exchange among top-tier experts, generates premier research achievements, and contributes to the health of the entire population.

Currently, HY Medical is collaborating closely with the Second Affiliated Hospital of Zhejiang University School of Medicine to establish a hospital-wide big data research platform. This platform enables multi-disciplinary integration of imaging and clinical data from radiology, ultrasound, pathology, nuclear medicine, and other clinical sources for research project management, covering various scenarios including hospital-wide, multi-departmental, single-departmental, and individual projects. Furthermore, HY Medical’s AI tools assist in image interpretation and perform automatic annotation, while also enabling automated model training and validation through various artificial intelligence algorithms.

To further advance research and development in medical big data, HY Medical has built a data middle platform for independent data storage, analysis, and processing, providing data support for Radcloud®. This platform goes beyond mere data storage; it aims to uncover the exceptional value of healthcare data by leveraging artificial intelligence, deep learning, and radiomics algorithms. Through data governance and cleansing, it transforms raw data into a valuable “treasure” for hospitals, facilitating the conversion of research outcomes and contributing to the future development of medicine.

This year marks the fourth year of Radcloud®’s practical application. The platform has helped hundreds of hospitals establish a closed-loop ecosystem spanning “clinical practice–research–clinical practice,” empowering physicians with tools to excel in both clinical care and scientific research, while enabling hospitals to build their own valuable research assets. Although the three-day CHINC conference has concluded, the development of medical digitalization and informatization continues to advance at a rapid pace. HY Medical’s path toward exploring this closed-loop ecosystem is becoming increasingly clear...