Manteia Secures Largest Overseas Order to Date with Contracts from 10 Korean Hospitals

“Recently, Manteia was invited to attend the 2022 Korean National Cancer Congress and the 2nd Asian Oncology Symposium (AOS). Mr. Zhu Yongjin, Executive Vice President, delivered a detailed presentation on the company’s research progress in accelerating and optimizing adaptive radiotherapy pathways. Additionally, procurement contracts were signed with ten hospitals in South Korea.”

Manteia’s AccuContour, an AI-assisted radiotherapy planning workstation, and AccuLearning, an autonomous algorithm fitting system, entered the South Korean market in 2020. As of June 2022, their market coverage reached 75%, with renowned medical centers such as Seoul National University Hospital and Korea University Anam Hospital among Manteia’s users. Approximately 30 hospitals initiated procurement processes, and the recent signing of contracts with 10 hospitals constitutes the largest single overseas order to date.

Manteia leverages the combination of AccuContour (hereinafter referred to as “AC”) and AccuLearning (hereinafter referred to as “AL”) to expand its services globally, serving medical centers worldwide.

As an auxiliary system for Treatment Planning Systems (TPS), AC primarily features independent and fused automatic segmentation of organs at risk and tumor target volumes across multiple imaging modalities (CT/MR/PET-CT), multi-modal image fusion and registration, superposition of physical and biological doses, plan review, and image analysis. It significantly enhances planning efficiency; for instance, it reduces segmentation time from the original 2–3 hours to just 1 minute, with a Dice similarity coefficient ranging from 0.85 to 0.95. The automated plan review function enables second-level verification of radiotherapy plans, achieving a 95% concordance rate with clinical expert scoring.

The diversification of data standards constitutes a significant barrier to the global expansion of products based on data-fitting models. Traditional approaches typically involve deploying engineers for on-site or remote training; however, these methods are clearly suboptimal in scenarios involving stringent data security requirements and a large customer base. To address this, Manteia has developed “robots” that simulate engineer workflows. AccuLearning features code-free operation, high precision, small-sample learning, and one-click automatic fitting. It enables medical centers to conveniently establish their own standard models or specialized target area models, thereby enhancing the applicability and generalizability of AC algorithm models. This effectively resolves the issue of poor adaptability of AC systems in overseas markets.

Adaptive radiotherapy can deliver clear clinical benefits, with daily precise and rapid contouring being a critical component of its implementation. Current generalizable models struggle to meet the demands for high-precision contouring, particularly for MR images that are significantly affected by field strength. Manteia combines Active Contour (AC) with Active Learning (AL), employing a One/Few-Shot training strategy using only 1–5 minimal samples to “fit today’s model with yesterday’s own images,” thereby achieving high-precision automatic contouring during MR-Linac (MRI-guided linear accelerator for radiation therapy) treatment. This approach has also yielded excellent clinical outcomes in overseas hospitals.

Note: 1→2* indicates that the model trained on data from the first treatment session was used for automatic segmentation of the daily images during the second treatment session; 1,2→3* indicates that the model trained on data from the first and second treatment sessions was used for automatic segmentation of the daily images during the third treatment session. The daily results represent the average Dice coefficient across five image pairs. The Dice coefficient ranges from 0 to 1, with values closer to 1 indicating better model performance.

Manteia currently serves approximately 1,100 healthcare institutions worldwide through its five core product lines: AC, AL, Adaptive Planning and Design System, Information Data System, and Online QA Verification System. Of these clients, 300 are overseas medical centers, while approximately 800 are located in China. The ten procurement contracts signed this time represent the largest single order to date in the overseas market. This milestone underscores the significant success of the company’s “global expansion” strategy launched in 2019, demonstrating a strong alignment with national policies promoting “technological innovation and self-reliance.” It also marks the company’s further entry into the international market, enhancing the global influence of Chinese brands and technologies.

Run All the Way, Grow All the Way!