The large model was built by AI training system, based on hundreds of millions of professional images and a huge database. Bioptimus, a French startup company, is one of those companies that is using artificial intelligence to promote the advancement of medical science.
According to the Wisdom Wealth News app, Bioptimus, a French startup company focusing on the comprehensive integration of cutting-edge AI technology, medical science, and biotechnology, recently launched a new AI large model, which was built by an AI training system and based on hundreds of millions of professional images/huge databases. The startup said that this latest AI large model will fully assist the research process of various diseases and help doctors speed up the diagnosis efficiency.
The company, based in Paris, said in a statement that the new AI model, called H-optimus-0, is able to perform extremely complex medical diagnostic tasks, including precise identification of cancer cells and detection of genetic abnormalities in tumors. Bioptimus describes the AI system as the largest open-source and widely accessible pathological AI model.
This start-up, which focuses on AI and medical science, is one of many companies that are trying to use artificial intelligence to achieve technological breakthroughs in medicine. Google's AI business unit, DeepMind, recently released a milestone AI large model called AlphaFold, which predicts the structures of proteins. US startup K Health has developed an AI chatbot that can talk to patients before they visit the hospital, to gain detailed knowledge of patient symptoms prior to diagnosis, providing important support for doctors' diagnoses. The company raised about $50 million in a transaction and was valued at about $900 million.
Bioptimus CTO Rodolphe Jenatton said the company chose the open source sharing mode to increase transparency and accelerate development in the field of "AI + healthcare." The company said it hopes further development will make the large model applicable to all major aspects of human biology (currently the model is basically focused on human cellular tissue).
People's general concerns about the use of artificial intelligence to diagnose diseases still exist. A 2023 survey by the Pew Research Center found that 60% of patients were concerned about doctors relying on AI systems to provide medical services. Earlier this week, some experts called on the UK's National Health Service (NHS) to focus on basic cancer diagnosis and treatment rather than AI's "miracle cure."
A study published in the top medical journal Nature Medicine in 2021 found that the AI algorithm was initially applicable to women with insufficient chest X-ray diagnoses and some patient groups receiving medical assistance.
Jenatton described the first release of the "H-optimus-0" AI large model as "a part of a longer journey."
Bioptimus was officially established in February of this year with the support and investment of Owkin Inc., a French biotechnology company. The startup received up to $35 million in seed funding, with investors including state-owned investment bank Bpifrance and telecoms billionaire Xavier Niel.
Bioptimus benefits from Owkin's powerful data generation capabilities and multimodal patient data from leading academic hospitals around the world. It also announced a partnership with AWS, a cloud computing giant under Amazon, to provide Bioptimus with scalable and secure cloud computing environments, and cloud computing resources services based on NVIDIA high-performance AI GPU and Amazon's self-developed AI chips, greatly enhancing Bioptimus's research capabilities.
The "golden track" certified by the AI Godfather! "AI + healthcare" can be said to be the most concerned field of NVIDIA.
As the helm of NVIDIA, Huang Renxun, who has been certified as the "AI Godfather," has more than once called "AI + medical science/biology" the next amazing revolution in the global technological field. He even declared at a recent conference that the era when everyone must learn how to operate computers has passed, and that biology and medicine are the future of humanity.
AI has made significant progress in the field of medical application, especially in drug discovery, gene editing, disease diagnosis, patient monitoring, and medical image analysis. For example, Nvidia has provided strong computing hardware and AI software tools support for the medical industry through its BioNeMo cloud computing services and other AI platforms supported by its high computing power, accelerating the process of drug development and disease diagnosis and treatment. Behind technologies such as AI large-model drug discovery and medical imaging, Nvidia's GPUs and software ecosystem provide powerful computing support.
Nvidia's high-performance AI GPU and more importantly CUDA platform, which has become a software/hardware co-working system preferred in the field of high-performance computing, have lifted Nvidia to the throne of 'King of Computing Power' and 'World's highest market cap chip company'. According to statistics, over the past two years, Nvidia's venture capital division has invested a considerable portion of funds in the fields of biotechnology and drug development. For example, in 2023 alone, Nvidia invested in eight startup biotech companies dedicated to drug research and development.
Behind medical technology that incorporates AI large models, Nvidia's GPU+CUDA provides powerful computing support. Nvidia not only provides GPU hardware, but also a complete set of supporting software tools and libraries to support the development of medical AI.
Nvidia's GPU, together with CUDA software tools and libraries, not only comprehensively accelerates research in fields such as medical image processing, drug discovery, protein structure prediction, and gene sequence analysis, but also supports the efficient operation of complex algorithms and large models. Therefore, Nvidia's software and hardware ecosystem plays an indispensable role in helping the medical industry achieve more efficient and accurate diagnosis and treatment solutions. From the perspective of AI large model at the bottom of the medical industry, Nvidia's AI GPU+CUDA is the preferred software and hardware platform to train medical AI large models, which are applied to various applications ranging from disease diagnosis to treatment recommendations in the medical field.
According to the latest research report from GMI Research, the market value of AI-enabled healthcare in 2022 is only about $10.4 billion, and it is expected to reach $189.9 billion in 2030, with a CAGR of 43.7% during this period. GMI Research believes that when analyzing highly recurrent medical data, AI can often perform tasks more quickly and at lower costs, thus greatly optimizing the connections between patients, doctors, and hospital management personnel. The organization also expects AI to be fully applied to medical imaging and rapid diagnosis, drug discovery, medical surgical assistant robots, gene-level precision medicine, drug simulation screening, and protein structure prediction, greatly reducing the product innovation cycle, comprehensively improving the accuracy of modern medicine, and covering a larger range of patients in the same amount of time.