brace it and engage AI to discover new approaches to imaging
and improve the care of patients.
AI involves many different complex components. Several
components are worth mentioning. Machine learning describes computers that have the ability to learn without explicitly being programmed. Deep learning is part of a broader
family of machine learning methods based on learning representations of data (i.e., objects displayed in images). Big data
is a term for data sets that are so large or complex that traditional data processing software and hardware are inadequate
to deal with them (i.e., radiology image archives). Artificial
neural networks are computational models used in computers
that are based on large collections of simple neural units (i.e.,
artificial neurons) loosely analogous to the observed behavior of biological neurons. Natural language processing is a
field of computer science concerned with the interactions between computers and human languages (i.e., voice recognition software). Learning algorithms are the study and
construction of algorithms that enable computers to learn
from multiple sources (i.e., radiology archives and pathology
data bases) to make predictions on data [i.e., computer-aided
AI systems have already been proven to perform better than
humans in visual recognition of certain objects on digital images.
More importantly, machine learning techniques utilizing artificial neural networks and learning algorithms will continue to enable computers to learn more and faster without explicitly being
programmed. Although processing speeds and computers have
not reached the computing power of humans, many experts are
predicting that machine learning could reach or exceed the ability of some elements of human intelligence by 2030!
However, will AI have the ability to advance medical care
via discovery? AI will certainly have the ability to discover by
Exploring the New Frontier of
Iam very honored and excited to serve as your president of the American Roentgen Ray Society. I first want to thank Mauricio Castillo our 2016–2017 ARRS president for his
great leadership in guiding the society this past year.
Mauricio’s efforts have been many and culminated in one of
the most successful annual meetings in New Orleans!
Our great specialty of radiology was founded with the discovery of x-rays by William Roentgen in 1895. Throughout the
20th century we have added many other significant discoveries including ultrasound, nuclear medicine, CT, and MRI.
Discovery must continue to be a core element of radiology if
we want to meet the challenges of the future.
Discovery is a very unique human endeavor guided by intellectual curiosity and intuition. It is also fostered by exploration,
observation, questioning, and, most importantly, serendipity.
Finally, the innately human components of passion, intensity,
and persistence are key drivers in the process of discovery.
However, there are many challenges facing radiology in
the future. It’s important for all of us to recognize these challenges, meet them head-on, and engage them as opportunities so that new discoveries can continue to improve the care
of our patients.
One of the perceived challenges in the future is artificial intelligence (AI). There are some who believe that AI could replace radiologists someday. Geoffrey Hinton is a cognitive
scientist and professor in the department of computer science
at the University of Toronto and is considered an international
leader in AI, deep learning, and artificial neural networks.
Hinton was recently quoted in the fall of 2016 as saying that he
believes AI will replace radiologists in 5–10 years because of its
deeper and faster learning capabilities in recognizing abnormalities in diagnostic imaging. Of course this was greeted with
consternation by the radiology and medical communities.
However, rather than resist this new technology, we should em-
By Bernard King
2017–2018 ARRS President
There are many challenges facing radiology
in the future. It’s important for all of us to
recognize these challenges, meet them head-on, and engage them as opportunities so that
new discoveries can continue to improve the
care of our patients.
President's Message continues on p. 15