Medical schools falling short in educating students on machine learning
Without educating med students, the advancement of AI will be limited in use and less likely to impact patient care, say researchers.
While healthcare is experiencing an explosion of interest in artificial intelligence and machine learning, medical schools must do a better job of educating future clinicians about the technology.
That’s the contention of researchers at the Boston University School of Medicine, who published a perspective article on Thursday in the journal NPJ Digital Medicine.
“Educating the next generation of medical professionals with the right (machine learning) techniques will enable them to become part of this emerging data science revolution,” state the authors. “Yet, the medical school curriculum as well as the graduate medical education and other teaching programs within academic hospitals across the United States and around the world have not yet come to grips with educating students and trainees on this emerging technology.”
Researchers did a PubMed search with “machine learning” as the subject heading term and discovered that the number of published machine learning (ML) papers has increased since 2010. However, they also found that the number of publications related to undergraduate and graduate medical education have remained relatively unchanged during the same timeframe.
“As medical education thinks about competencies for physicians, ML should be embedded into information technology and the education in that domain,” said co-author Priya Sinha Garg, MD, associate dean ad interim for Academic Affairs at Boston University School of Medicine.
"If medical education does not begin to teach medical students about AI and how to apply it into patient care, then the advancement of technology will be limited in use and its impact on patient care," added co-author Vijaya Kolachalama, assistant professor of medicine at Boston University School of Medicine.
According to Peter Elkin, MD, chair of the Department of Biomedical Informatics at the State University of New York at Buffalo’s Jacobs School of Medicine, over the last few years, the better U.S. medical schools are beginning to teach more of their academic informaticians to leverage machine learning technology.
Elkin contends that machine Learning is a technology that has been around for two decades, but the advent of big data in healthcare has served to enable these algorithms. When used in conjunction with increasingly powerful graphical processing unit cards, he says they have the potential to develop highly accurate predictive analytics.
“One caveat that we should all keep in mind is that poor quality data can limit these methods’ ability to create highly accurate predictive analytics,” Elkin observes. “We teach machine learning with healthcare examples in our Department of Biomedical Informatics at the University at Buffalo. We make sure that our students understand concepts such as data provenance, cleaning, imputation and data validation strategies. High-quality data when connected to powerful machine learning algorithms have the potential to bring cutting-edge care to patients all over the world.”
Last year, the Buffalo medical school received a five-year $2.5 million grant from the National Library of Medicine to train researchers to analyze and leverage the explosion of healthcare data that is transforming the industry.
Also See: Buffalo med school gets $2.5M grant to train informatics researchers
“Medical students who take our clinical informatics course learn machine learning,” says Elkin. “All other medical students are given an intersession course—mid-December through the first week in February—where they are exposed to informatics methods. This is so that they know what the methods can do to assist them in their practice and their research. That exposure is not aimed at training them to work in informatics.”
Currently, his school offers a master’s degree, a doctoral program, postdoctoral training and a clinical informatics fellowship that’s Accreditation Council for Graduate Medical Education-approved toward board certification in clinical informatics.
“To study toward a career in clinical informatics where you master technology such as machine learning, physicians are encouraged to apply for an ACGME-accredited fellowship in clinical informatics which leads to board eligibility,” Elkin adds.
That’s the contention of researchers at the Boston University School of Medicine, who published a perspective article on Thursday in the journal NPJ Digital Medicine.
“Educating the next generation of medical professionals with the right (machine learning) techniques will enable them to become part of this emerging data science revolution,” state the authors. “Yet, the medical school curriculum as well as the graduate medical education and other teaching programs within academic hospitals across the United States and around the world have not yet come to grips with educating students and trainees on this emerging technology.”
Researchers did a PubMed search with “machine learning” as the subject heading term and discovered that the number of published machine learning (ML) papers has increased since 2010. However, they also found that the number of publications related to undergraduate and graduate medical education have remained relatively unchanged during the same timeframe.
“As medical education thinks about competencies for physicians, ML should be embedded into information technology and the education in that domain,” said co-author Priya Sinha Garg, MD, associate dean ad interim for Academic Affairs at Boston University School of Medicine.
"If medical education does not begin to teach medical students about AI and how to apply it into patient care, then the advancement of technology will be limited in use and its impact on patient care," added co-author Vijaya Kolachalama, assistant professor of medicine at Boston University School of Medicine.According to Peter Elkin, MD, chair of the Department of Biomedical Informatics at the State University of New York at Buffalo’s Jacobs School of Medicine, over the last few years, the better U.S. medical schools are beginning to teach more of their academic informaticians to leverage machine learning technology.
Elkin contends that machine Learning is a technology that has been around for two decades, but the advent of big data in healthcare has served to enable these algorithms. When used in conjunction with increasingly powerful graphical processing unit cards, he says they have the potential to develop highly accurate predictive analytics.
“One caveat that we should all keep in mind is that poor quality data can limit these methods’ ability to create highly accurate predictive analytics,” Elkin observes. “We teach machine learning with healthcare examples in our Department of Biomedical Informatics at the University at Buffalo. We make sure that our students understand concepts such as data provenance, cleaning, imputation and data validation strategies. High-quality data when connected to powerful machine learning algorithms have the potential to bring cutting-edge care to patients all over the world.”
Last year, the Buffalo medical school received a five-year $2.5 million grant from the National Library of Medicine to train researchers to analyze and leverage the explosion of healthcare data that is transforming the industry.
Also See: Buffalo med school gets $2.5M grant to train informatics researchers
“Medical students who take our clinical informatics course learn machine learning,” says Elkin. “All other medical students are given an intersession course—mid-December through the first week in February—where they are exposed to informatics methods. This is so that they know what the methods can do to assist them in their practice and their research. That exposure is not aimed at training them to work in informatics.”
Currently, his school offers a master’s degree, a doctoral program, postdoctoral training and a clinical informatics fellowship that’s Accreditation Council for Graduate Medical Education-approved toward board certification in clinical informatics.
“To study toward a career in clinical informatics where you master technology such as machine learning, physicians are encouraged to apply for an ACGME-accredited fellowship in clinical informatics which leads to board eligibility,” Elkin adds.
More for you
Loading data for hdm_tax_topic #better-outcomes...