Evolutionary Thinking 3 Medical AI, focusing on the grassroots

Mondo Technology Updated on 2024-01-19

A 4-year-old boy with a strange illness. I looked at 17 doctors, but I didn't find **. His mom turned to ChatGPT for help. It's ChatGPT, which exploded this year. ChatGPT said that the child had tethered cord syndrome. Then, the doctor confirmed the diagnosis of ChatGPT. Although ChatGPT is not right every time. But this time. Helped a lot.

This is a case shared in "The Power of Evolution", and it actually surprised me, we all know that AI calculates possibilities in a known body of knowledge, it may be faster than humans, it may be "smarter" than most doctors, but its ability to outperform medical experts is incredible.

Sam Altman, the founder of OpenAI, said in a speech that when the model is big, sometimes there are some surprising results. We don't fully understand why either.

If artificial intelligence can really think and explore the unknown, maybe it will not be a surprise but a fright

On November 30, 2022, ChatGPT was born, forming a heated discussion around the world. According to Similarweb,In January this year, an average of about 13 million unique visitors used ChatGPT every day, more than double the number in December last year, with a cumulative total of more than 100 million users, setting a record for the fastest Internet application to break 100 million, surpassing the previous TikTok speed of breaking 100 million in 9 months. This is unmatched by any previous AI application.

The main reasons why ChatGPT is popular are as follows:

Technological innovationThe reason why ChatGPT has been able to explode is due to its strong technical support. ChatGPT uses the Transformer model in deep learning Xi technology, which can effectively handle the semantic and contextual relationships of natural language. Through large-scale pre-training, ChatGPT is able to understand and generate coherent conversations, making user interactions more intelligent and natural. ChatGPT also has the characteristics of a generative language model, which can think independently and generate corresponding responses, bringing users a new interactive experience. Wide range of applicationsThe wide application of ChatGPT has also contributed to its popularity. In the business field, ChatGPT can be used for intelligent customer service and consulting to help enterprises improve service quality and efficiency. In the field of education, ChatGPT can be used as an auxiliary education tool to communicate with students and answer questions, and provide personalized Xi support. In the medical field, ChatGPT can be used for intelligent consultation and health consultation to provide patients with timely and accurate medical advice. ChatGPT can also be used in social entertainment, news recommendations, and other fields to meet people's diverse needs. Economic benefits: ChatGPT-powered chatbots are more cost-effective than traditional customer service channels such as call centers and face-to-face interactions. They are available 24 hours a day** to solve customer problems and improve customer satisfaction at a lower cost. In addition, AIGC can generate articles, write, design and even complete more complex content more efficiently, which greatly improves the efficiency of content creation and output. To sum up, the popularity of ChatGPT and artificial intelligence is due to a combination of factors. As technology continues to advance and innovate, we can expect AI to play a greater role in future developments.

Gartner, a technology research and analyst firm, named it the number one in its Top Strategic Technology Trends for 2022. According to Gartner, generative AI will account for 10% of all generated data by 2025, but this is less than 1% today.

Building on traditional AI and advanced analytics, generative AI could unlock a further $2 trillion worth of economic benefits in China (per year). On the 27th, Violet Chung, senior partner of McKinsey & Company Greater China, said at the same press conference.

The phenomenal performance of ChatGPT has also successfully ignited the enthusiasm of Chinese artificial intelligence to catch up. Once upon a time, we had high hopes for artificial intelligence, which carried China's catch-up and even surpassing that of Western developed countries in the field of high-tech, after all, China's huge population base and vast market have created conditions for the application of artificial intelligence. However, the release of ChatGPT has once again sounded the alarm for China's high-tech enterprises, it is not enough to meet the innovation of the application layer, and artificial intelligence also needs the underlying technology.

Therefore, China's high-tech enterprises are not to be outdone and have joined the battle.

: Wenxin large model Tencent: hunyuan large model Ali: Tongyi large model Huawei: Pangu large model In addition to these top enterprises, unicorns in other industries have also launched their own large models, such as the medical industry in which the author is engaged, is also not to be outdone, and large model products with pattern characteristics have sprung up.

Deepwise Medical: Deepwise Metai.

DeepWise-CIRP Model, a general medical image understanding model, has been independently developed, which can automatically learn Xi understand medical images from raw images and report information, without too much artificial annotation.

Yunzhisheng: a large model of mountains and seas
Relying on the Shanhai model, Yunzhisheng has released three major medical product applications: "Surgical Medical Record Writing Assistant", "Outpatient Medical Record Generation System" and "Commercial Insurance Intelligent Claim System".

Medical Federation: MEDGPT
The self-developed first medical large language model in China based on Transformer architecture - MEDGPT. MEDGPT mainly aims to give full play to the actual diagnosis and treatment value in real medical scenarios, and realize the whole process of intelligent diagnosis and treatment capabilities from disease prevention, diagnosis, and development.

Langma Information: Langma 39ai General Practitioner
Relying on the massive medical and health data accumulated by the Internet medical business of Longma Information for more than 20 years, based on advanced natural language processing technology and deep learning Xi algorithms, and through large model training methods, "Langma 39ai General Practitioner" can solve most of the medical Q&A, health science, health management and other related problems in people's daily life.

In addition to these businesses and products, there are:healthgptHuatuo GPTbsoftgptcloud gptWait a minute.

The author's company, which can be said to be the first enterprise engaged in the application and promotion of medical artificial intelligence, has been committed to applying artificial intelligence to the field of oncology since the very beginning of its cooperation with IBM Watson. Due to many factors such as the Sino-US war, after giving up cooperation with IBM, the company has also invested a lot of resources in independent research and development of tumor artificial intelligence, and has achieved certain research and development results.

In fact, artificial intelligence also has this wide range of applications in the medical field, including but not limited to the following:

Intelligent diagnosis and treatment: Diagnosis of diseases using artificial intelligence technology. Computers can help doctors to carry out statistics on pathology and physical examination reports, analyze and mine patients' medical data through technologies such as big data and deep mining, and automatically identify patients' clinical variables and indicators. The computer simulates the doctor's thinking and diagnostic reasoning through the professional knowledge related to "learning Xi", so as to give a reliable diagnosis and a good plan.

As early as November 6, 2017, the pass line of the "2017 Comprehensive Written Examination for Clinical Practitioners" was announced, and the iFLYTEK "Intelligent Medical Assistant" robot achieved a score of 456 points, exceeding the qualification line of clinical practitioners (360 points), and belonging to the intermediate and senior level among the 530,000 candidates in the country.

Intelligent recognition of medical imagesAI can help doctors locate the lesion area by learning a large number of medical images through a large number Xi of medical images and reduce the problem of missed diagnosis and misdiagnosis.

Tencent Miying is Tencent's first AI product applied in the medical field. It contains 6 AI systems covering diseases including esophageal cancer, lung cancer, diabetic retinopathy, cervical cancer, and breast cancer. Among them, its intelligent screening system for early esophageal cancer is the most mature, with a laboratory accuracy rate of 90%, and has now entered the preclinical experimental stage.

Medical robotsRobots are widely used in the medical field, such as intelligent prosthetics, exoskeletons and assistive devices to repair damaged human bodies, and healthcare robots to assist medical staff in their work. At present, the research on the application of robots in the medical community is mainly focused on surgical robots, ** robots, nursing robots and service robots.

On November 6, 2023, the domestically produced da Vinci XI surgical robot was officially announced at the CIIE, and Intuitive Fosun successfully realized the localization of the da Vinci surgical robot.

Personalization**Artificial intelligence can provide personalized solutions based on the patient's genetics, medical history and other information. For example, based on a patient's genetic information, it is possible to respond to a certain drug and thus make a better choice of regimen.

On July 23, 2022, at the opening ceremony of the Beijing International Forum on Thoracic Surgery, the scientific research result jointly cooperated by Professor Zhang Yi's team from the Department of Thoracic Surgery of Xuanwu Hospital of Capital Medical University and BAHEAL Intelligent Technology Co., Ltd. - SmartCDS for Lung Cancer was officially released. The system aims to use AI to help the sharing of expert resources, build a new ecosystem of medical interconnection with algorithms, technology and experience, and jointly create a new experience of diagnosis and treatment of lung cancer patients.

Medical management: Artificial intelligence can be used in medical management, such as case management, medical resource allocation, etc., to improve medical efficiency and quality.

The speech recognition technology and information summary generation ability of the surgical medical record writing assistant of the Yunzhisheng Shanhai model are used to automatically record the speech during the operation and convert it into valuable information. The outpatient medical record generation system uses the knowledge graph and language processing capabilities of the Shanhai model to help doctors quickly generate standardized outpatient medical record documents.

The application of artificial intelligence in the medical field can improve the efficiency and quality of medical care, improve the patient** experience, and also help doctors better process large amounts of data and information to improve the level of medical care. Especially in the field of primary care, artificial intelligence has more application value. Why?

First of all, what is the biggest advantage or value of artificial intelligence?"The Power of Evolution" says that all the great opportunities of the future are hidden in the word "productivity". Replacing repetitive low-level labor, improving work efficiency, and reducing human error rate are the first things that artificial intelligence needs to solve, in other words:Do more with less. Replacing people to make decisions and doing higher-level things is the process of continuous iterative upgrading of artificial intelligence technology.

For primary medical institutions, there is no complex clinical decision-making, but more basic diagnosis, tedious and repetitive work. However, the level of grassroots doctors is uneven, and the medical technology is relatively low. Artificial intelligence is now more learning and Xi evolution within the existing knowledge framework of human beings, and may be stronger than most ordinary people in decision-making ability, but it is not necessarily stronger than experts. If the level of grassroots doctors is regarded as 60 points, and the level of medical experts is regarded as 90 points or even higher, it is obvious that it is easier for artificial intelligence to increase 60 points to 80 points, and it is relatively difficult to increase 90 points by 1 point. Therefore, primary medical institutions, primary doctors need more auxiliary tools, it is easier to improve productivity, and in some ways it can even completely replace general practitioners.

Like iFLYTEK's "Intelligent Medical Assistant", Langma's 39ai general practitioners are aiming at this field, and the cognitive level of "Langma 39ai general practitioners" at this stage can reach the medical level of primary and intermediate general practitioners. In the future, it will be widely used in primary medical and health institutions to carry out clinical auxiliary diagnosis services for medical staff, provide artificial intelligence services similar to "general practitioners", carry out auxiliary diagnosis, health consultation, and health management, form a district-level disease spectrum, improve the diagnosis and treatment level of medical personnel, and provide health management consulting services for ordinary people throughout the life cycle.

The author is responsibleBAHEAL Health CloudAfter completing the layout of basic products, it will definitely transplant the capabilities of artificial intelligence to the platform through integration or self-development, so as to help grassroots medical institutions and doctors improve productivity and provide customers with a better product experience.

Caigen Lao Tan (headline number, Douyin account, ** number with the same name), experienced programmer, architect, R&D director, and now the head of a subsidiary of a group, from technology to product, won the 2021, 2022 popular author of the Renren Product Manager Community.

Related Pages