Augmented intelligence (AI), also known as artificial intelligence, is helping doctors improve healthcare, from achieving more accurate diagnoses to developing more personalized treatment plans to enhancing overall patient care. There are many ways to help you transform. Doctors know this too. His AMA survey of more than 1,000 physicians found that nearly two-thirds of them saw a potential benefit.
The AMA Ed Hub™ CME series introduces learners to the fundamental principles of AI and machine learning. Machine learning is a subdomain of AI that allows computers to learn patterns and relationships from data without being explicitly programmed by humans. Developed by the AMA ChangeMedEd initiative and the University of Michigan DATA-MD team, it is aimed at medical students and is also suitable for residents, fellows, practicing physicians, and other health care professionals who can earn their CME by completing these modules. I am.
The second module in the series, “AI in Healthcare: Methodologies,” examines three basic machine learning techniques: supervised learning, unsupervised learning, and reinforcement learning. We also discuss deep learning and show how natural language processing is used in medicine.
To help AI reach its full potential to advance clinical care and improve the health of clinicians, the AMA has developed new advocacy principles that build on current AI policy. These new principles (PDF) address the development, implementation, and use of healthcare AI.
Learn more about the emerging landscape of augmented intelligence in healthcare at AMA (PDF).
Three ways machine learning happens
In this module, we will explore these three main methodologies for training machine learning in the medical field.
Supervised learning. This involves a computer learning from examples that accurately predict the outcome of interest, with the goal of generating accurate predictions for new examples. Supervised learning requires labels and input. The labels describe what is expected, such as the presence or absence of a diagnosis, while the inputs consist of electronic medical record data, omics, medical images, and medical text.
Unsupervised learning. This is a type of machine learning in which an algorithm learns from unlabeled data without a predefined outcome or target variable. The goal of unsupervised learning is to uncover common patterns, structures, or relationships in your data, such as distinct clusters of patients, disease subtypes, and outliers. Unsupervised learning models can be used, for example, to help cluster patients with autism spectrum disorders and discover typical usage and disease progression trajectories.
Reinforcement learning. It uses data about a set of interventions and their outcomes or rewards to identify the optimal set of interventions that maximizes rewards. For example, when treating sepsis, doctors can monitor a patient's health status, including vital signs, test results, and comorbidities. Next, determine the amount of fluids and vasopressors to administer to the patient. After treatment, check to see if the patient has fully recovered or died. The reward is 1 in the former case and 0 in the latter case. Using this data, the model outputs the optimal sequence of actions to follow depending on the patient's condition.
Challenges common to each methodology are highlighted.
This module also provides an example of how to apply supervised learning techniques to a dataset to predict the presence or absence of cardiovascular disease. Additionally, you will learn how deep learning, which processes data using artificial neural networks with multiple layers, is applied to each of the three methodologies, and how natural language processing allows computers to analyze text data. , and how it will be possible to generate human language.
Regular knowledge checks and review sections test users' vocabulary and understanding of how concepts are applied.
The CME module “AI in Healthcare: Methodology” is a permanent resource and is rated by the AMA at a maximum of 0.50.AMA PRA Category 1 Credit™.
It is part of the AMA Ed Hub, an online platform with high-quality CME and education that supports the professional development needs of physicians and other healthcare professionals. It also provides an easy and streamlined way to search, capture, track, and report on your educational activities using topics that are relevant to you.
Learn more about AMA CME certification.
