A：Tex Technology, led by founder and chairman of the board of directors Song Ning, was listed in Hong Kong IPO on March 30 this year, known as "the first stock of medical imaging model". The company focuses on pushing the global medical imaging industry into a new era of artificial intelligence, successfully developing the world's first and only medical imaging AI base model, and has the world's largest medical imaging data platform MID loop. Since its establishment in 2016, Texi Technology has accumulated more than 20 years of experience in medical clinical and computer disciplines. It is in an international leading position in medical imaging large model technology and has the fastest market advancement. It has cooperated with many well-known companies in medical care. Strategic cooperation in the field of imaging intelligence. The image of MI developed by Tex Technology is trained from scratch and has significant performance advantages. Compared with the model development required for traditional disease diagnosis and treatment, only a few hundred medical image photos of specific diseases can be used to train relevant models, which greatly reduces data cost and computing power expenditure and shortens the development cycle. In addition, the company also launched the IMI loop, through data labeling quality control and special training, so that clinical experts can quickly train proprietary models on the basis of large models, forming an intelligent ecological scene in the medical imaging industry. Tex has successfully launched products that have been certified for medical device registration, such as chromosomal AI-assisted diagnostic software, and has enhanced the value of medical imaging data assets through AI technology.

A：Tex Technology has accumulated more than 28.95 million cases of gold-labeled medical imaging data, ranking first in the market share of specific medical imaging fields, and some segments quickly surpassed international well-known manufacturers to become the first in market share within one year. and recognized by the international academic community. The company has established a wide range of cooperative relations with top hospitals in China, and developed AI auto vision products applied to test tube baby parents and the gold standard detection of blood tumor morphology. In May of this year, Desshi Technology obtained the three types of medical device registration certificates issued by the State Food and Drug Administration, becoming the first approved medical imaging AI auxiliary product in the past three years, reflecting its technical advancement and clinical effects. Highly recognized. At the same time, Tex Technology not only has the sales of model authorization and medical device registration certificate products, but also has intelligent supporting assembly lines, which further enhances the competitiveness of products. In the first half of last year, revenue increased by 21% year-on-year, of which model revenue increased by 101 year-on-year. The profit margin maintains a relatively high level in the industry. With continued research and development investment and industry ecological co-creation, the company is expected to achieve sustained high-speed growth.

A：Medical imaging covers B- ultrasound, CT, nuclear magnetic resonance, pathological images, fundus endoscopy and other macro and micro detection items inside and outside the human body obtained through medical equipment. At present, there are more than 3000 kinds of medical imaging in the world, and there are more than 14000 different categories according to the indications and the number of testing items. It is estimated that by 2030, the number of medical imaging tests in AI will reach more than 30000, and the market size will reach more than 150 billion RMB, which is about 15 times higher than the current 10 billion RMB size in 2026. Around this development trend, Tex Technology is committed to the research and development of the underlying technology, and has constructed the image of the large model MI of medical imaging to meet the needs of accurate diagnosis in the medical field.

A：Well, in response to this problem, our short-term strategic goals focus on core technologies, including the collection and construction of high-quality data sets for medical imaging, and the continuous improvement of the underlying capabilities of the model. At the level of commercial market landing, we will pay more attention to cooperation with top hospitals at home and abroad, and promote doctors to build thousands of medical imaging model training based on a small amount of their own data, so as to improve the intelligent research and development level of the whole industry. At the same time, we are also actively communicating with the competent national ministries and commissions to integrate medical imaging large model technology into the country's artificial intelligence medical strategy to accelerate industrial development.

A：In the medium term we think the medical imaging industry will go through three phases. The first stage is the transition from manual expert identification and analysis to AI-assisted diagnosis in the next few years of more than 14000 existing 3000 kinds of medical imaging diagnosis projects, which greatly improves the accuracy of diagnosis and treatment and the level of primary medical diagnosis. The second stage will begin about two years later. With the development of artificial intelligence technology, more and more imaging diagnostic technologies that have not been developed or do not exist in clinical practice will appear, such as predicting the risk of premature delivery by B- ultrasound, which is expected to reach more than 16000. In the third stage, about 4 to 6 years later, the in-depth development of AI technology and the deep integration of images and language models will promote the efficient, accurate and inclusive development of the global human health management system.

A：First of all, it is of great strategic significance and economic value to AI the existing diagnostic projects, which can make the cognitive ability and diagnostic ability of top experts quickly digitized and popularized in primary hospitals, so as to benefit more people. Secondly, we note that based on long-term clinical knowledge, top clinical experts have put forward many new diagnosis and treatment propositions, such as predicting circulatory system diseases through plain CT, and accurately predicting other diseases through ordinary image detection beyond the resolution limit of the human eye, which are expected to develop rapidly. Finally, in terms of commercial landing, we have noticed that some friends have trained proprietary models, but this approach is expensive, time-consuming and difficult to obtain evidence. We believe that our model is indispensable, and we also see more possibilities and innovative directions.

A：We believe that from the perspective of the entire medical industry and medical imaging, the market size of the traditional non-AI era is very large. With the advent of the era of artificial intelligence, the global attention to health has increased, and equipment has been rapidly deployed to hospitals, but it is difficult for expert doctors who use high levels and issue diagnostic reports to train in batches. AI becoming a trending technology, the industry is going through a rapid early stage of development. It is estimated that by 2030, the global medical imaging AI market will reach 150 billion RMB, with high growth potential.

A：German is a company with both computer and clinical background, the founder has a deep dual professional foundation. In the past ten years, German has made industry-leading cognitive breakthroughs in technical routes, strategic landing and commercialization, and proposed innovative ideas for processing medical images with large model technology, and achieved results in different fields. German is a professional team with a deep understanding of clinical medicine and basic medicine led by AI science and technology, and is committed to transforming scientific and technological achievements into products and services that can be implemented in the medical industry.

A：The current model service includes two forms of cloud service and offline service. The core is that each service can precipitate the company's underlying IMIH base large model capability and provide reusable base capability. More importantly, focus on the deep deployment of the model into the clinical workflow after training and bring in continuous case payment revenue. The specific path is divided into two stages: the first stage is to cooperate with major hospitals or enterprises to develop a special model to meet their needs.

A：We put forward the solution, that is, through the development of high performance AI technology, it is applied to the detection of medical images. For example, the use of AI-assisted diagnostic technology can sink high-quality medical resources to basic hospitals. With the help of the information channels of the medical association, the grass-level hospitals issue diagnostic reports, which improve the throughput and accuracy, and reduce the pressure of medical insurance at the same time. Under this model, the hospital gets a hierarchical income, the base hospital gets the brand response and testing costs, and the upper hospital.

A：We have many years of experience in the field of medical imaging software and hospital recruitment, AI automation products have entered hundreds of domestic hospitals. AI auction team in the product from the certification to the hospital procurement landing process, will follow the industry cycle law, is expected to have a certain volume growth in the second half of this year, and continue to grow next year, driving the entire medical imaging intelligent industry update iteration. At the same time, with the listing and global awareness, products have begun to be used by top medical institutions around the world, and the global market is developing. Therefore, investors should pay attention to the company's performance in upgrading and purchasing the whole industry from the second half of the year to next year.

A：In the field of medical imaging, what is really scarce is the professionalism and uniqueness of medical imaging. Our core barrier is to reconstruct and redesign the entire model architecture from the underlying framework to the characteristics of medical imaging to meet the needs of medical imaging judgment. In addition, we have also done a lot of work in data processing. Through the intelligent labeling platform-MI's loop and studio, we have accumulated more than tens of millions of high-quality labeled medical image data, which is another important barrier in the industry.

A：The demand for medical images is unique and different from application scenarios such as automatic driving or street navigation. In medical images, these abnormalities are numerous and inexhaustible.

A：The current general image model does not have the characteristics of the underlying technology in processing medical images, especially for abnormal situations that are complex in medical images and require professional doctors to identify. Therefore, we start from the underlying framework and reconstruct the model architecture based on the transformer concept to meet the needs of medical imaging-assisted diagnosis. Li Feifei's point of view mainly emphasizes that large models can also be described abnormally based on text reports without direct observation of pictures, but this does not represent a major breakthrough in the field of medical imaging. In fact, we have been focused on this area for many years, have broken through technically, and have been recognized in the top clinical community. We welcome Fei Li to experience our medical imaging model to gain an intuitive understanding of its latest technological developments.