脉得智能
On April 29, 2026, at the 9th Digital China Summit, the China National Intellectual Property Administration released the third batch of ten national typical cases involving data intellectual property. The “Thyroid Nodule Assisted Diagnosis Dataset” registered by MaiDe Intelligent Technology (Wuxi) Co., Ltd. was successfully selected. Jiangsu Province accounted for only 2 of the cases selected as national typical cases. This not only marks national recognition of MaiDe Intelligence’s innovation in the data intellectual property system, but also demonstrates that, in Jiangsu Province, where data is highly developed, MaiDe Intelligence’s high-quality medical data is becoming a driving force for the large-scale clinical application of “artificial intelligence + healthcare” and “artificial intelligence + software,” as well as the creation of a closed loop for the data industry.
The Value of a High-Quality Dataset
In the era of artificial intelligence, high-quality, standardized medical imaging datasets are the core foundation for the commercialization of AI-assisted diagnosis. MaiDe Intelligence’s “Thyroid Nodule Assisted Diagnosis Dataset” brings together multiple leading medical institutions to conduct standardized de-identification and high-precision annotation of ultrasound images from multiple sources and devices, with data accuracy exceeding 99%. The dataset is part of MaiDe Intelligence’s larger “High-Quality Multimodal Ultrasound Medical Dataset.” In August 2025, the project was selected for the National Data Administration’s pilot task for the development of high-quality datasets. Positioned as the first industry-level dataset in the domestic ultrasound field, it covers 15 key organs, including the thyroid, breast, and liver, and contains more than 1 million structured cases and more than 10 million ultrasound images. The thyroid nodule dataset is the most mature core submodule of this data repository and the first to achieve commercialization.
From Data to Products: The Technology Transformation PathThe ITS200 thyroid nodule ultrasound imaging assisted diagnosis software developed on the basis of this dataset uses artificial intelligence deep learning technology to identify thyroid nodules and provide assisted diagnosis results regarding benignity or malignancy. These results are highly consistent with histopathological findings, substantially reducing unnecessary fine-needle aspiration biopsies. The product currently covers more than 200 Class III hospitals nationwide, extends to more than 1,000 institutions, and has obtained the world’s first Class III ultrasound AI medical device certificate benchmarked against the pathological gold standard. Its accuracy in distinguishing benign and malignant lesions exceeds 96%. It can shorten physicians’ diagnostic time by more than 30%, improve diagnostic accuracy at primary-level institutions by 30%+, and reduce reporting time by 80%, enabling non-invasive ultrasound to achieve pathology-level diagnostic precision.
Being selected as a national typical case in data intellectual property is both a milestone and a new starting point. Within its overall intelligent ultrasound diagnosis and treatment solution portfolio, MaiDe Intelligence has already developed applications for the thyroid, breast, cervical lymph nodes, liver, and prostate, among other organs. It is committed to building a “product + service” business closed loop, enabling data to truly become a foundational force driving precision healthcare.
Data intellectual property registration is like a “golden key” that opens the door to unlocking the value of data as a factor of production. MaiDe Intelligence’s practice demonstrates that when high-quality datasets meet institutional innovation, the resulting impact can far exceed expectations. This represents not only commercial monetization, but also systematic improvements in healthcare equity, diagnostic standardization, and the patient experience. References 1.Yangtze Evening News Online, “Two Cases from Jiangsu Selected for the Third Batch of Ten National Typical Cases in Data Intellectual Property”, April 29, 2026 2.China Jiangsu Online, “Two Cases from Jiangsu Selected as Typical Cases of National Data Intellectual Property Registration”, April 29, 2026 Written and typeset by | Luo Chaoran, Huang Qian Designed by | Huang Ruoxi Reviewed by | Chen Yonghong