On Wednesday, local time, Google Cloud announced that it and Mayo Clinic officially launched a new service to help medical experts quickly query patient information through various latest artificial intelligence robots. Mayo Clinic is one of the more advanced medical institutions in the United States, with dozens of branches. Google, which is trying to promote "generative AI" in its healthcare system, was one of the first hospitals to adopt the technology.
Google is playing catch-up
This search tool can effectively help medical professionals search a large number of different internal data within the organization through chatbots. This means that doctors, nurses, etc. can more quickly and simply understand a patient's medical history, imaging records, genomes, and other data, even if this information is stored in different formats and locations.
Mayo Clinic will test different uses of the search tool in the coming months. May Clinic chief technology officer Vish Anantraman said it was "very satisfying" in helping clinicians with tiresome administrative tasks. For example, if a doctor needs to view information about a group of female patients between the ages of 45 and 55, including their mammograms and medical records, they only need to type the query into the search tool. Similarly, if doctors want to know which tests a patient needs to match, they can also do a search match. Anantraman said: "This will greatly save time and reduce a lot of administrative burden for doctors. ”
Since OpenAl released its chatbot ChatGPT to the public, generative AI has been the hottest topic in the tech world. Google has jumped to catch up, launching Bard AI chat earlier this year and working to embed the underlying technology into more products.
The healthcare industry is a particularly challenging industry because it is virtually zero fault tolerance. Ashima Gupta, global head of healthcare strategy and solutions at Google Cloud, said in an interview with CNBC: "We want to be thoughtful and responsible when leveraging such a powerful tool as generative AI in healthcare. ”
In 2019, Google Cloud and Mayo Clinic signed a 10-year partnership. Mayo Clinic said at the time that it chose Google Cloud as a "cornerstone of digital transformation."
The partnership announced Wednesday is the first step in the two companies' expanded agreement to work together on the use of AI in healthcare.
According to the analysis of Forbes News Network in the United States, Google Cloud's announcement of a partnership with the world-renowned "Mayo Clinic" clinic is a milestone. It is important for medical professionals to go through a large number of patient records, to organize multiple databases and academic collections, or even to gather information about treatment strategies and the latest therapies and symptoms.
Big manufacturers are seizing the AI+ medical field
After realizing the breakthrough significance of generative AI, the global medical industry has rapidly introduced generative AI into the medical field with unprecedented speed and enthusiasm, and has already achieved preliminary results.
According to Artery Network, the recent big news about the exploration of generative AI in the medical field is the cooperation between the famous medical software development organization EPIC and Microsoft. As the world's leading healthcare informatics giant, EPIC is firmly the leader in the electronic medical record (EMR) market in the United States and the first choice for large medical institutions.
At the HIMSS23 exhibition in April 2023, EPIC, a well-known medical software development organization, announced its cooperation with Microsoft to integrate AIGC in EHR systems. This is also the first time EPIC has embraced generative AI. At present, EPIC has launched two solutions. One is the introduction of generative AI in communication scenarios. It will automatically draft responses to some of the most common and time-consuming information for doctors. Of course, it does not replace the doctor's judgment, and the doctor can accept it or reject the recommendation to write it himself. The second option is to combine generative AI with EPIC's SlicerDicer tool. SlicerDicer is a self-service reporting tool that provides physicians, department managers, and other users with intuitive and customizable data search capabilities.
In March 2023, less than a week after the release of GPT-4, Microsoft's Nuance announced that it would introduce GPT-4 in its product Dragon Ambient eXperience Express.
A former leader in voice AI, Nuance turned its focus to healthcare after being challenged by the tech giants. In April 2021, Nuance was acquired by Microsoft for a huge $19.7 billion. The deal greatly strengthens Microsoft's presence in the healthcare vertical. Nuance's products primarily provide speech recognition and transcription services to doctors. Voice AI can intelligently identify the content of doctor-patient conversations, perform contextual analysis, and then input data into electronic medical records to automatically create clinical records to improve the efficiency of doctors' diagnosis.
Google also announced in mid-April that it would test its large medical-specific model, Med-PaLM2, in a limited user base. In the past few years, Google has been conducting research on large language models in medicine and has released the first generation of Med-PaLM to cope with the professionalism and specificity required in the medical field. The Med-PaLM model has achieved good results, and it is the first AI to achieve a "passing score" (>60%) on US medical licensing questions. Med-PaLM2 goes further. In the medical exam, Med-PaLM2's performance is basically close to that of "expert" doctors, scoring 85%.
In the combination of generative AI and medical treatment, NVIDIA has also long had a layout. In 2022, NVIDIA and King's College London will use a supercomputer to create a dataset of 100,000 synthetic brain images to train AI applications to accelerate the understanding of dementia, Parkinson's disease and other brain diseases.
In terms of synthetic data, the Academic Health Center of the University of Florida has also collaborated with NVIDIA to develop a SynGatorTron generative AI model that generates synthetic clinical data. It is trained on a decade of data from more than 20,000 patients and synthesizes patient profiles that researchers can use to train other AI models in healthcare. In addition, NVIDIA has partnered with a number of companies to provide generative AI assistance for its research and development of new drugs.