Thorough Imaging Secures Tens of Millions in Pre-A Funding to Advance AI-Powered Pathology Products and Hardware Integration

VCBeat has learned thatThorough Images (Beijing) Technology Co., Ltd. recently announced that it had secured tens of millions of yuan in Pre-A round financing. The round was led by Puhua Capital, with Turing Ventures and Panlin Capital participating as co-investors, and BMD Capital serving as the exclusive financial advisor. Thorough Images stated that the proceeds from this financing will be used for AI product development, multi-center clinical trials, NMPA regulatory approval, and hardware integration.

Thorough Images, established in 2017, is a medical technology company dedicated to leveraging artificial intelligence to provide intelligent diagnosis and prediction for pathological images, as well as delivering optimized solutions.

Good technologies and products are inseparable from “talent.”

At Thorough Images, there are more than ten top-tier technical experts, including Ph.D. graduates from Tsinghua University and Nankai University, overseas-returnee algorithm engineers, and high-tech professionals with many years of experience at leading companies such as Philips, Sinovation Ventures, and JD.com. Meanwhile, Thorough Images maintains a close collaboration with the Institute for Interdisciplinary Information Sciences at Tsinghua University.Our top-tier talent across the three dimensions of algorithms, products, and systems has laid a robust foundation for the technological caliber of our products.“Thorough Images General Manager Liu Yanbin said in an interview with VCBeat.”

Liu Yanbin stated, “The product philosophy of Thorough Images is to strive to deliver the most accurate pathological diagnoses.”

In just two years since its establishment, Thorough Images has builtTThorough Insights (透彻洞察) AI-Assisted Diagnostic Platform、Thorough ESD (Thorough ESD) AI Pathology Diagnosis Platform、Thorough Wisdom (Touche Huizhi) Pathology Image Annotation Systemand other industry-leading products.

The deep learning-based AI-assisted diagnostic platform, Thorough Insights, can automatically identify lesion areas, assist physicians in locating lesions on slides, and enter the automatically generated auxiliary diagnostic reports into the LIS system.

(Figure Note: ThoroughInsights | In-Depth Insights)

“Thorough Insights can significantly enhance diagnostic efficiency and prevent missed diagnoses.“Thorough Images’ General Manager, Liu Yanbin, introduced. Currently, Thorough Insights can be applied to the pathological auxiliary diagnosis of organs such as the stomach, intestines, esophagus, prostate, and lungs. He stated that auxiliary diagnostic models for more organs are under development.”

ESD specimens typically comprise dozens of pathological slides, posing significant challenges for the construction of pathological lineage maps. Traditional manual methods require pathologists to spend 4 to 8 hours and lack precision.Thorough Images Launches AI-Powered ESD Pathology Diagnosis Platform, Thorough ESD, Enabling Automated Generation of Pathological Lineage Maps and Reducing Overall Diagnosis Time to Under 15 Minutes, and provide the location distribution of cancerous tissues, the severity of malignancy, and whether metastasis has occurred., this information is critical for clinicians to formulate further treatment plans.

Figure Note: ThoroughESD | Thorough ESD

“In simple terms, our platform not only reduces the time cost for physicians but also precisely identifies lesion locations, significantly improving their work efficiency,” said Liu Yanbin.

Data is a critical gateway to intelligence. To build supervised deep learning models, high-precision, pixel-level pathological annotations are required. “Pathologists have rated Wisdom as the most user-friendly pathological image annotation system,”Annotation efficiency can be more than 10 times that of PC-based annotation tools, and it provides AI-assisted annotation features.，The system currently supports over 1 million high-precision annotations.”introduced Liu Yanbin.

ThoroughWisdom | Thorough HuiZhi

He pointed out,Currently, leveraging these two products and a labeling system, we have established collaborations with multiple renowned hospitals in China.Collaborating hospitals include: Chinese PLA General Hospital, Cancer Hospital of the Chinese Academy of Medical Sciences, Peking Union Medical College Hospital, China-Japan Friendship Hospital, Beijing Chest Hospital, Beijing Friendship Hospital Affiliated to Capital Medical University, and Peking University People's Hospital, among others.

When discussing the reasons for the successful fundraising, Liu Yanbin stated that the success was mainly attributed to the three advantages of Thorough Images.

I. Data Aspects:Thorough Images has established close partnerships with multiple top-tier Grade 3A hospitals, and has designed and developed an annotation tool based on the iPad and Apple Pencil to accelerate physicians’ annotation workflows.

More than 40 collaborating annotating physicians, all attending and chief physicians with over 10 years of experience, are involved. Each slide must undergo a four-tier review process, including initial review, secondary review, chief physician re-review, and expert spot-check, to ensure the accuracy and authority of data annotation.

II. Algorithmic Aspects:Thorough Images has accumulated technical expertise over two and a half years, with its deep learning models progressing through three R&D stages: customization based on existing models, fully independent development, and self-evolving neural networks.

The currently employed models are developed by leveraging large volumes of annotated data in conjunction with automated machine learning techniques, achieving industry-leading accuracy and computational efficiency.

III. Implementation System:Thorough Images has independently developed a distributed pathological image storage and analysis system. The entire system is built on a distributed microservices inference architecture, supporting scalable expansion across any number of machines and GPUs.

System performance scales linearly with the number of GPUs, enabling a single server to complete the intelligent interpretation of a 1 GB pathology image within 20 seconds.

Based on the above technological achievements, Thorough Images currentlyHolds over 20 software copyrights, has filed nearly 10 invention patents for AI in pathology, with 3 already granted.

“In addition to invention patents, we have also achieved notable success in academic research.Research on weakly supervised learning in pathological imaging was recently accepted by ICCV 2019, a top-tier international conference on computer vision.“, and has garnered extensive attention from researchers both in China and abroad,” said Wang Shuhao, CTO of Thorough Images.

(Figure caption: Frontier research by Thorough Images was published at the top international computer vision conference ICCV 2019)）

Through this financing round, VCBeat also interviewed several institutions that invested in Thorough Images to hear their reasons for investing.

Ms. Li Yang, Investment Director at Puhua Capital, stated:

We began strategically building a comprehensive ecosystem for pathology products in 2016, making AI-powered pathology a key focus area. AI companies typically progress through six stages: prototype development, principal investigator (PI) collaboration, academic promotion, database establishment, regulatory communication, and distributor-based sales. Each stage requires the reserve of key personnel and resources.

The Thorough Images team is a multidisciplinary group that combines technical and clinical expertise, demonstrating strong strategic acumen and operational pacing. We look forward to the synergies between Thorough Images and the Puhua Pathology ecosystem.

Mr. Huang Longbo, a member of the Investment Committee at Turing Venture Capital and a tenured associate professor at the Institute for Interdisciplinary Information Sciences, Tsinghua University, stated:

China, with the world’s largest population, boasts an immensely promising healthcare market. Among various diagnostic approaches, pathological examination for cancer stands out as one of the most critical methods for diagnosing major and severe diseases.

Thorough Images, with its multidisciplinary team, robust artificial intelligence technologies, and deep understanding of and resources in the healthcare industry, provides integrated hardware-software solutions for pathological diagnosis of diseases affecting the gastrointestinal tract, lungs, prostate, and other organs, thereby creating substantial value for healthcare institutions and patients. The Institute for Interdisciplinary Information Sciences at Tsinghua University has licensed certain AI technology transfers to Thorough Images. We are highly optimistic about and supportive of Thorough Images’ long-term development.

Mr. Li Yuhui, Founding Managing Partner of Panlin Capital, stated:

The rapid advancement of computer vision AI has positioned the healthcare sector as one of its key future growth areas. Pathology, being the foundation of medicine, faces significant challenges in China, including high demand, a shortage of professionals, heavy workloads, and elevated risks. With the growing adoption of digital pathology in clinical practice, it has become an ideal application scenario for AI. This technology helps reduce repetitive tasks and improve operational efficiency, effectively driving cost reduction and efficiency enhancement on the supply side.

Thorough Images’ team entered the field of pathological AI by focusing on gastrointestinal biopsies, which are characterized by a high rate of repetitive labor. This strategic choice allowed the company to overcome data barriers in product development while actively exploring viable commercialization models. We are optimistic about the company’s successful application of technological innovation to integrate into and enhance medical workflows.

Finally, Liu Yanbin, General Manager of Thorough Images, told VCBeat that after the financing, they will continue to focus on the development of artificial intelligence products, accelerate hardware integration, and penetrate the market.

[About Puhua Capital]

PuHua Capital, founded in 2004, specializes in early- and mid-stage venture capital investment. Over the years, it has deeply cultivated three major sectors: healthcare, new technology, and pan-culture, with investments in more than 400 companies to date. As a key focus area, PuHua Capital has comprehensively invested and strategically positioned itself across the broader healthcare industry, covering precision medicine, biopharmaceuticals, medical robotics, innovative medical devices, and healthcare services. The firm has been repeatedly ranked among the Top 20 Healthcare Investment Institutions in China by Zero2IPO (Qingke) and among the Top 10 Best Medical Device Investment Institutions in China’s Healthcare and Health Services Industry by CVSource (TouZhong).

[About Turing Venture Capital]

Turing Ventures, established in 2018, specializes in early-stage equity investments in artificial intelligence and information technology. Leveraging the premier technological capabilities, talent pool, and government and industry resources of Tsinghua University’s Institute for Interdisciplinary Information Sciences, led by Academician Andrew Chi-Chih Yao, as well as the Nanjing Turing Artificial Intelligence Research Institute, and backed by extensive fund management experience and a professional investment team, Turing Ventures is committed to facilitating the commercialization of scientific research outcomes. By investing in and incubating next-generation information technology startups, the firm aims to accelerate the adoption of AI and IT solutions in traditional industries.

[About Panlin Capital]

Panlin Capital is among the earliest venture capital firms in China to explore and practice the theme of technology-driven innovation leading consumption upgrading amid the development of China’s new economy. Following the logic of consumption upgrading, Panlin Capital focuses on the prevention, diagnosis, and treatment of major diseases, as well as the intelligent transformation of the consumer supply side. It continuously identifies and invests in technologically innovative enterprises and entrepreneurial leaders who leverage biotechnology or artificial intelligence innovations to effectively meet emerging consumer and diagnostic/treatment needs, thereby significantly enhancing supply efficiency and service quality, and accelerates their rapid growth.