Seven Years of Growth: Shukun Technology Sets a Benchmark in AI-Driven Healthcare

At the end of last month, SHUKUN celebrated a milestone birthday.

On the eve of its seventh anniversary, SHUKUN’s self-developed chest CT image processing and analysis software (Lungdimension Doc) officially received the NMPA Class III medical device registration certificate, marking its 13th Class III certification for medical AI.

Unlike AI for assisted diagnosis, chest CT image processing and analysis software is an AI tool designed for preoperative lung planning. This product can automatically identify and analyze plain and contrast-enhanced CT imaging data, rapidly generate 3D reconstructed images of pulmonary vascular structures, and reconstruct normal and variant lung segments, thereby guiding physicians in formulating more precise and personalized surgical plans to improve the success rate and safety of surgeries.

Since its inception, medical AI has aimed to integrate comprehensively into the entire disease diagnosis and treatment workflow. The recent approval of SHUKUN’s AI solution for preoperative lung planning has enabled the company to successfully tackle the most challenging aspect of tumor care: treatment. Combined with its previously approved AI products for pulmonary nodules and pneumonia, SHUKUN’s AI portfolio now provides comprehensive coverage across the full cycle of disease diagnosis and treatment.

Reviewing the numerous achievements attained today, SHUKUN has clearly realized the goals it set during the initial surge of the AI industry. Having shed the “imaging” label, it has ultimately become a true “medical AI” company.

When SHUKUN entered the medical AI sector in early June 2017, it faced an emerging yet crowded market. Prior to achieving commercial validation, numerous companies had already established their presence in scenarios where imaging data was readily accessible, such as fundus photography and pulmonary nodule detection. At that time, SHUKUN adopted a differentiated R&D strategy, entering the field through the less-explored area of cardiology, using coronary CTA as the foundation to explore medical artificial intelligence.

Selecting coronary CTA as the target is driven not only by the high market potential and significant barriers to entry in the cardiac sector, but also by its ability to generate tangible health economic value. After all, AI-enabled coronary CTA can reduce examination time from hours to just three to five minutes, provide physicians with high-precision image reconstruction, guide precise surgical decision-making, and significantly lower the per-disease treatment costs for hospitals.

This approach not only aligns closely with clinical needs but also avoids the risk, common in internet-driven development of medical AI, of creating solutions that are “visually appealing but practically unusable.” For a considerable period, SHUKUN has been expanding its product pipeline along this logic. Covering major body regions from the heart and brain to the chest, abdomen, and musculoskeletal system, the company achieved comprehensive coverage of cardiovascular and cerebrovascular diseases as well as common oncological conditions in less than two years, securing its position among the industry’s leading tier.

However, the aforementioned logic still has its limitations. Since physicians’ “clinical needs” are complex and fragmented, AI developed along these lines mostly amounts to “single-disease AI” that addresses only one specific problem. In clinical practice, however, physicians always take a comprehensive view of potential issues across the patient’s organs, rather than first assuming the presence of a particular disease and then using a corresponding AI tool to validate that hypothesis.

Therefore, SHUKUN has come to realize that an ideal medical AI must break through the limitations of single departments and single modalities. By focusing on patients’ physiological functions and adopting a holistic perspective, it can deliver more precise and convenient smart healthcare services.

Guided by this core philosophy, SHUKUN launched its groundbreaking medical AI technology platform, “Digital Human,” at the 2021 World Artificial Intelligence Conference. This platform integrates digital products for various anatomical regions—including “Digital Heart,” “Digital Brain,” “Digital Chest,” “Digital Abdomen,” and “Digital Musculoskeletal System”—and distills their key technologies into industry-wide infrastructure. It enables intelligent understanding of human anatomical structures and diverse pathological conditions, thereby better delivering patient-centric solutions that cover the entire continuum of disease screening, diagnosis, and treatment decision-making.

As a result, SHUKUN’s Digital Doctor suite has begun to expand beyond CT to additional imaging modalities, extend from radiology to other clinical departments, and broaden its scope from image post-processing and computer-aided diagnosis to encompass more comprehensive workflows. It is applicable to a wide range of clinical scenarios, including coronary artery disease, stroke, oncology, trauma, chest pain centers, stroke centers, and cardiology specialties. This advancement transcends the previous fragmented, single-disease AI pipelines, establishing a multidisciplinary diagnostic and treatment model supported by an integrated imaging platform, thereby achieving a transformation in AI applications from isolated points to comprehensive coverage.

Taking the most mature “Digital Heart” solution of that year as an example, SHUKUN integrated its existing products—including coronary AI, intelligent CT-based calcium scoring assessment, CT-FFR, FAI, and TAG intelligent assessment—to form the “Digital Heart” family. Following this integration, patients can undergo a single scan of the target organ to screen for a wide range of common conditions. This approach effectively reduces diagnostic and screening costs while comprehensively covering all stages of disease management: screening, diagnosis, treatment, and ongoing care, thereby leveraging AI to realize the traditional Chinese medical philosophy that “the superior physician prevents disease before it occurs.”

After identifying the right direction, SHUKUN has iterated its Digital Human product annually, further expanding the range of AI-addressable conditions and improving diagnostic and treatment efficiency. Now, in the 3.0 era, the Digital Human proposes the integration of scenarios with technology, algorithms with devices, and medical technologies with general-purpose technologies, further strengthening AI’s empowerment of healthcare scenarios.

The AI-powered preoperative lung planning system, recently approved during the anniversary celebration, is a product of the integration of scenarios and technologies within Digital Human 3.0. In addition, SHUKUN has strategically positioned itself around two surgical categories—"Intelligent Liver Surgery" and "Intelligent Urological Surgery"—to drive surgical procedures toward greater intelligence, precision, and safety.

The integration of algorithms and devices represents another innovation by SHUKUN. Leveraging AI, SHUKUN has independently developed native ultrasound hardware solutions, such as the “Turing Brain” and “Turing AR,” which enable deep integration with intelligent algorithms. These systems not only consolidate information from all organs during ultrasound diagnosis and treatment to reflect lesion characteristics in real time, but also optimize the physician experience by eliminating the need for a “second screen.”

From SHUKUN’s perspective, the synergistic integration of software and hardware not only reconstructs user experience but also serves as the key to AI’s continuous breakthroughs in technological boundaries. In the future, every piece of hardware will transition from the industrial era into the AI era, with AI empowering every step and every second of the interaction between doctors and medical devices. The new quality productive forces emerging during this transformation will define a second growth curve for improving quality and efficiency in healthcare scenarios.

Although the emergence of digital humans has disrupted the application models of imaging AI, a broader healthcare perspective reveals significant room for optimization. After all, medical imaging data accounts for nearly 90% of healthcare information, yet the remaining 10% of non-imaging data also holds substantial value.

In the past, medical imaging AI and natural language processing (NLP) were two independent technologies. Enterprises had to invest substantial capital and effort to develop both simultaneously, yet it remained difficult to generate novel innovations through technological integration. Today, the emergence of large language models has reshaped the trajectory of AI development, positioning SHUKUN at the forefront of this transformative wave.

At this spring’s CMEF, SHUKUN’s AIGC innovation—the multimodal medical large model ShuKunGPT—moved from the backstage to the forefront.

According to SHUKUN, ShuKunGPT represents a breakthrough innovation in the integration of its medical technology with general-purpose technologies. It is capable of not only understanding text, images, and video information but also providing complex recommendations based on comprehensive, multi-dimensional data across modalities.

Currently, the image and video modality capabilities of ShuKunGPT are primarily applied in scientific research and education. The former can overcome the limitations of imaging types, supporting segmentation and recognition of lesions across all human organs to assist physicians in making precise diagnoses; the latter enables movie-level rendering of human organ structures, helping physicians better understand organ function.

In the rankings released by CMB, the largest Chinese medical evaluation benchmark, SHUKUN GPT outperformed its competitors across the vast majority of metrics. In terms of commercialization, SHUKUN has achieved early implementation through research collaborations and partnerships with health administrative authorities.

Overall, the value of ShuKunGPT can be divided into two parts. On one hand, large language models can help SHUKUN strengthen its capabilities in the two non-clinical phases—pre-diagnosis and post-diagnosis—further optimizing the completeness of the entire disease diagnosis and treatment lifecycle. On the other hand, previous medical AI development relied heavily on manual annotation, whereas now, large language models can automatically segment and recognize the vast majority of imaging data, significantly reducing data governance costs and accelerating the generation of medical AI research outcomes.

Perhaps in the future, AI-assisted diagnosis will no longer be exclusive to high-volume diseases.

SHUKUN’s seven-year journey is not only a brilliant chapter of technological innovation, but also a vivid microcosm of the development of the medical AI industry.

From its initial differentiated R&D strategy to the comprehensive deployment of "Digital Human" 3.0, SHUKUN has consistently remained at the forefront of the industry, continuously pushing technological boundaries and advancing patient-centric smart healthcare services to new heights.

However, the imaging diagnosis and surgical scenarios in which SHUKUN is now deeply involved represent only the tip of the iceberg within the healthcare industry. AI holds immense potential to empower the entire medical service process and full-lifecycle health management. SHUKUN is actively exploring this frontier, striving to cover more clinical scenarios and departments, while its regional smart health management model has been deeply implemented in areas such as Changping District in Beijing and Wuzhong District in Suzhou.

Looking ahead, SHUKUN must continue to uphold its spirit of innovation, further explore the development and application of medical AI technologies, constantly expand the range of diseases covered by AI, and improve diagnostic and treatment efficiency. In this process, SHUKUN is well-positioned to provide physicians with more convenient and intelligent tools, deliver higher-quality and more personalized healthcare services to patients, and thereby accelerate the advent of the era of intelligent healthcare.