Subjects
IA in action
This series of columns examines the biggest data and analysis challenges faced by modern companies and dives deeply in successful use cases that can help other organizations accelerate their AI progression.
More in this series




In a precedent AI action column in actionWe argued that in the health world, the administrative applications of artificial intelligence were the low fruits. Sometimes, however, it is reasonable to reach the upper branches of the tree, and the AI clinical applications enter this category. We expect that one day many decisions of important diagnostics and treatment be taken or increased by AI applications. Today, we are in the early stages of achieving this objective.
Most of the advances currently carried out by the clinical AI come from innovative health care establishments. An important example is Mayo Clinic, a hospital and a research center that was founded in Rochester, Minnesota, over 150 years ago. He is consistent in US News & World Report‘s ranking of the best hospitals And excels in a variety of specialty fields. At any time, approximately 12,000 research studies are underway. According to our experience, there is no organization of health care providers that does more with AI than the Mayo clinic.
AI is explored in a variety of areas there, including the types of administrative use cases that we mentioned in our previous column (including patient planning, prior authorization with payers and income cycle requests). But it is also being developed for clinical uses such as specific specific applications in radiology and cardiology, diagnostics and remote management based on health care sensors and appear patients with clinical trials.
These AI initiatives are part of a major development towards digitization. In 2020, Mayo Clinic appointed a digital director to direct his digital strategy on a business scale and trained the Center for Digital Health (CDH). As Mayo Clinic HUB for digital transformation, CDH uses data, analysis, automatic learning and AI to help improve patient health and provide recommendations for monitoring, diagnostic and processing through digital channels to patients around the world.
IA innovation at the head at the Mayo clinic
Dr. Bhavik Patel is a diagnostic radiologist and director of the Arizona branch of the Mayo clinic, and the chief of innovation and activation of AI at Mayo Clinic as a whole. Many of the first advances of AI, in particular those of computer vision, were in medical imaging, it is therefore not surprising that an expertise in radiology contributes to paving the way for many AI efforts of the organization. Some of his own research, for example, explains how AI can help create and improve the quality of images for radiologists to read.
In an interview, Patel described these efforts as “carrying out clinical trials with AI to work on clinical implementation” – in other words, disciplined experimentation. Patel said that in radiology, deep learning models are entirely able to recognize the image if they have fed enough quality training data. However, he noted that the AI systems currently available – both those of suppliers and those developed internally – are always close in scope (that is to say, specific radiology or cardiology). Efforts are underway, he said, to develop multimodal AI models that affect several specialties and therefore have a wider impact of patient care.
Patel said he considered his main role as concentrating on the moment and the moment when AI and automatic learning are necessary and on the best ways in research. He said that at present, he thinks that many automatic learning models are brittle and may not generalize beyond the research laboratory. Validation and in -depth tests are necessary to guarantee safe and responsible AI solutions for patient care.
Before any AI model entered the clinical deployment, the Mayo clinic and similar service providers must establish who will governing and directs its use, what are the best practices, what is the scope of the model, what are the biases of the model and when to trust the model. Mayo Clinic currently has some models in daily clinical use, one of which analyzes electrocardiogram data. Patel noted that they do not focus on the analysis and diagnosis of autonomous images still; He believes that AI will help human radiologists and other suppliers in the foreseeable future.
Accelerate the development and marketing of the model
Mayo Clinic focuses on how to generate cases of using AI faster with the normalization of software, development processes and technology. It aims to quickly produce automatic learning models with a so-called AI factory approach It was launched in 2021. More than 200 use cases have already been part of the process, which he calls a “research continuum on discovery translation applications”. Mayo Clinic also collaborates closely with Google on AI, including its use of Google Cloud Platform, and has some joint initiatives with Google Health, a unit which is associated with organizations of health care providers to conduct R&D solutions and create solutions.
Mayo Clinic has specific programs that support AI for the three areas of practice, research and education. It is Mayo clinic platformDirected by the long -standing director of the Boston hospital, Dr. John Halamka, serves as a connector of external digital health producers and consumers, allowing the dissemination of the organization’s knowledge on a large scale.
This powerful innovation and technology platform orchestrates an ecosystem of partners, including a portfolio of startups using AI and other resources of the Mayo clinic. A program of startups called Platform_accerateNew in 2022, provides identified training data sets, model validation frameworks, clinical workflow planning and Mayo Clinic expert mentoring in exchange for participation stations. The first four entrants in the 20 -week competitive accelerator were IA companies. The Mayo clinic has also transformed several startups to market its own research on AI.
It is clear from our conversations and our research that the use of AI at the Mayo clinic explodes, but this growth is accompanied by challenges. Patel offered this perspective: “There are so many activities in progress. Identities. “”
The key question, he said, is how Mayo Clinic can continue to support innovative efforts while providing executives for the creation of fair and explanable clinical AI models. Fortunately, Mayo Clinic has long been a collaborative culture, and there is an opening to the sharing of early learning throughout the organization. It is still too early to say how AI will reshape health care, but perhaps not too early to predict that Mayo Clinic will play an important role in transformation.