United Imaging Chairman and CEO Xue Min: AI Empowers a Medical Technology Innovation Ecosystem

Xue Min, Chairman and CEO of United Imaging Healthcare Group Co., Ltd.

Editor’s Note: This article is based on the speech delivered by Xue Min, Chairman and Chief Executive Officer of United Imaging Healthcare Group Co., Ltd.

Artificial intelligence is, in fact, a long-established discipline; more than 60 years have passed since the term “AI” was first coined. However, it is only in the past decade that AI has begun to be comprehensively applied across industries and in people’s daily lives, driven by breakthroughs in deep learning algorithms, the accumulation of big data, and substantial increases in computing power, thereby creating opportunities for disruptive innovation. This naturally includes the healthcare industry.

In the healthcare sector, “AI + medical imaging” has emerged as one of the most promising areas with the highest likelihood of achieving commercialization first. Many technology and internet companies are actively establishing their presence in this field. The number of enterprises and the volume of financing clearly indicate that AI-driven medical imaging has become a hot spot for startups.

However, it is also evident that among the currently deployed AI applications in medical imaging, the majority are focused on the “diagnosis” stage, with a particular concentration in the field of “lung” imaging. In terms of imaging data modalities, most applications primarily utilize CT and XR data. This is largely because pulmonary nodule data, as well as CT and XR datasets, are relatively abundant and easier to acquire.

However, healthcare actually involves many stages, with diagnosis being just one step. Beyond pulmonary diseases, there are 20,000 to 30,000 human diseases that require attention. In particular, the diagnosis of major conditions such as neurological disorders and tumors often relies on multimodal data, including MRI and PET-CT.

Therefore, it can be said that the industry has only taken a small step forward, and “Medical Imaging AI” still has enormous room for development.

How else can AI more comprehensively empower medical imaging?

United Imaging’s approach is to start at the source, using its full line of imaging equipment as the entry point for AI, thereby empowering the entire healthcare process from the ground up.

Imaging equipment is the source of imaging data and directly determines image quality, which in turn directly affects physicians’ diagnosis and treatment.

Furthermore, it is well recognized that data and practical application scenarios are critical to the successful deployment of AI. Medical devices possess inherent advantages in both aspects. Imaging equipment can be utilized across all clinical stages, from image acquisition to prognosis, offering a wide range of application scenarios. Moreover, multimodal devices have accumulated vast amounts of clinical data.

Therefore, it is crucial to ensure quality at the source. Empowering the initial imaging stage with AI provides a solid and reliable foundation for the true implementation of AI.

Specifically, we will carry out our practices from two dimensions: empowering devices and empowering clinical practice.

Empowering Devices: We Integrate AI Across the Entire Imaging Workflow, from Scanning and Post-Processing to Reporting and Treatment.

For example, this is a technology that helps radiologic technologists achieve intelligent and rapid scanning, applied in the scanning process.

Through our latest independently developed 3D imaging technology, the camera can accurately capture 3D dynamic human body information in a 360-degree range. When a patient enters the scanning room, it can precisely capture the patient's movements and posture, identify information such as age, height, and weight, and automatically push this data to the equipment. The device will then automatically generate the necessary scanning parameters based on this information.

With a single click, physicians can complete the scan, eliminating the need to spend time identifying patients, entering patient information, and selecting parameters as was previously required. This significantly reduces physician workload and enhances scanning speed.

Our AI-powered rapid MRI scanning technology, using the head as an example, can complete routine clinical scans in 1 minute and 48 seconds, with all other body parts scanned within three minutes, reducing scan time by 70%.

Furthermore, by leveraging AI-powered intelligent noise reduction algorithms, we enable 1.5T MR systems to achieve imaging signal-to-noise ratios comparable to those of 3.0T MR systems.

Radiation dose is another major concern for both physicians and patients during imaging procedures. Many people harbor fears about the radiation generated during CT scans. Although technological advancements have already reduced radiation doses to a safe range, the integration of AI technology can further lower radiation exposure by up to 85%, providing greater peace of mind for patients.

In the post-processing stage, AI-powered intelligent rendering technology can generate high-resolution, detailed images of organs, assisting physicians in making more accurate imaging diagnoses and surgical plans.

Currently, AI is rarely applied in the treatment phase, yet this area has a substantial demand for AI. We know that radiotherapy is one of the most commonly used methods for tumor treatment. In China, due to the long training period and high technical difficulty, there is a severe shortage of medical physicists specializing in radiotherapy, particularly senior physicists. The application of AI technology is expected to alleviate this predicament.

For instance, prior to radiotherapy, the organs at risk and target volumes must be delineated. Traditionally, physicians spent considerable time performing manual delineation. With AI-powered intelligent delineation technology, this process can be accelerated from 15 minutes to under 0.3 seconds, significantly saving time while enhancing precision. Physicians need only perform a brief verification of the results. This serves as an extremely powerful tool for medical physicists with limited experience.

Beyond empowering devices, we also enable AI to better empower clinical practice, creating an AI solution that spans the entire clinical diagnosis and treatment workflow and covers multiple disease types.

Leveraging high-quality imaging results, we apply AI to diagnosis, particularly for complex and rare conditions, to better assist physicians.

For instance, on MRI equipment, our intelligent brain structure assessment system can automatically segment 112 brain substructures and quantify the data to generate a structured report within three seconds with a single click. By leveraging these quantitative data, physicians can more intuitively observe specific changes in various brain structures, enabling early detection of pathological changes and timely intervention for major conditions such as Alzheimer’s disease, autism, Parkinson’s disease, and brain tumors, thereby significantly improving patients’ quality of life. This is undoubtedly of great significance for China, which is rapidly aging. It is worth noting that this technology will also play a pivotal role in frontier explorations in brain science.

Leveraging low-dose, high-definition images generated by our CT systems, we also employ AI to detect and diagnose multiple pulmonary diseases in a single scan.

The above represents only some of United Imaging’s initial innovations and practices. In the future, I believe medical AI will not be confined to the fields of diagnosis and treatment; it will inevitably expand to the broader consumer market, reach the general public, and permeate the entire health and wellness sector. By leveraging intelligent medical-grade wearable devices, smart surgical robots, and rehabilitation robots, we will ultimately establish a fully intelligent healthcare ecosystem that spans prevention, diagnosis, treatment, and rehabilitation.

United Imaging has also been gradually expanding its presence in these fields.

Leveraging the United Imaging Group, our subsidiaries and business segments can share resources and platforms to mutually empower one another.

We have not only established United Imaging Intelligence, a subsidiary dedicated to AI research and development, but also founded United Imaging Microelectronics to independently develop customized medical chips and AI chips. These initiatives will enable our future equipment to become more miniaturized and intelligent. Furthermore, United Imaging Zhiron (also known as Wuhan United Imaging) will help us expand our business comprehensively, extending from the B2B sector to the B2C market.

The maturation of new technologies has created a “window of opportunity” for enterprises to accelerate their development; only by seizing this historical moment can they achieve leadership in new competitive arenas.

United Imaging is honored to be part of the current wave of rapid artificial intelligence development. We hope to contribute our utmost to shaping the future of the entire intelligent healthcare ecosystem.

To this end, we are building an open medical AI ecosystem to enable more like-minded individuals and entities—including a wide range of companies and motivated, skilled physicians—to participate in AI innovation and R&D based on our innovative platform. Through cross-disciplinary openness and sharing across various fields, we aim to facilitate the rapid and widespread technological translation and commercial implementation of these applications and innovations, thereby accelerating industry development.

Building an innovative ecosystem requires the joint efforts of industry, academia, clinical practice, and capital. By maintaining patience and taking a long-term view, we will work together to usher in a new era.