Machine-Learning Technology Devised to Control Drug Cravings

Novel machine-learning imaging technology has been created to anticipate neurocognitive changes, similar to predictive text-entry for cell phones and Internet search engines.

In the study, researchers reported on several key developments in this field, employing functional magnetic resonance imaging (fMRI) and machine-learning techniques to image the brains of smokers experiencing nicotine cravings.

At University of California, Los Angeles’ (UCLA; USA) Laboratory of Integrative Neuroimaging Technology, the researchers used functional MRI brain scans to observe brain signal alterations that occur during mental activity. They then employ computerized machine learning (ML) methods to examine these patterns and identify the cognitive state--or sometimes the thought process--of human subjects. The technique is called “brain reading” or “brain decoding.”

The study’s findings were presented in December 2011 at the Neural Information Processing Systems’ Machine Learning and Interpretation in Neuroimaging workshop in Grenada (Spain) and was funded by the US National Institute on Drug Abuse (Bethesda, MD, USA), which is interested in using this technology to help people control drug cravings.

In this study on addiction and cravings, the scientists classified data taken from cigarette smokers who were scanned while watching videos designed to trigger nicotine cravings. The goal was to determine which regions of the brain and neural networks are responsible for resisting nicotine addiction specifically and cravings in general, according to Dr. Ariana Anderson, a postdoctoral fellow in the Integrative Neuroimaging Technology lab and the study’s lead author.

“We are interested in exploring the relationships between structure and function in the human brain, particularly as related to higher-level cognition, such as mental imagery,” Dr. Anderson said. “The lab is engaged in the active exploration of modern data-analysis approaches, such as machine learning, with special attention to methods that reveal systems-level neural organization.”

For the study, smokers sometimes watched videos meant to induce cravings, sometimes watched “neutral” videos, and at times watched no video at all. They were instructed to attempt to fight nicotine cravings when they arose.

The data gathered from fMRI scans taken of the study participants were then analyzed. Traditional machine-learning techniques were supplemented by Markov processes, which use past history to predict future states. By measuring the brain networks active over time during the scans, the resulting machine learning algorithms were able to anticipate changes in the study participants’ underlying neurocognitive structure, predicting with a high level of accuracy (90% for some of the models assessed) what they were watching, and as far as cravings were concerned, how they were reacting to what they viewed.

“We detected whether people were watching and resisting cravings, indulging in them, or watching videos that were unrelated to smoking or cravings,” said Dr. Anderson, who completed her PhD in statistics at UCLA. “Essentially, we were predicting and detecting what kind of videos people were watching and whether they were resisting their cravings.”

Basically, the algorithm was able to complete or “predict” the study participant’s mental states and thought processes in much the same way that Internet search engines or texting programs on cell phones anticipate and complete a sentence or request before the user is finished typing. Moreover, this machine-learning method based on Markov processes demonstrated a large improvement in accuracy over conventional approaches, according to the researchers.

Machine-learning methods, in general, create a “decision layer”--essentially a boundary separating the different classes one needs to distinguish. For example, values on one side of the boundary might indicate that a subject believes various test statements, and on the other, that a subject disbelieves these statements. Researchers have found they can detect these believe-disbelieve differences with high accuracy, in effect creating a lie detector. An innovation described in the new study is a means of making these boundaries interpretable by neuroscientists, instead of a frequently obscure boundary created by more conventional methods, such as support vector-machine learning.

“In our study, these boundaries are designed to reflect the contributed activity of a variety of brain subsystems or networks whose functions are identifiable--for example, a visual network, an emotional-regulation network or a conflict-monitoring network,” said study coauthor Mark S. Cohen, a professor of neurology, psychiatry and biobehavioral sciences at UCLA’s Staglin Center for Cognitive Neuroscience and a researcher at the California NanoSystems Institute at UCLA.

“By projecting our problem of isolating specific networks associated with cravings into the domain of neurology, the technique does more than classify brain states--it actually helps us to better understand the way the brain resists cravings,” added Dr. Cohen, who also directs UCLA’s Neuroengineering Training Program.

Remarkably, by placing this problem into neurologic terms, the decoding process becomes considerably more effective and accurate, according to the researchers. This is particularly significant, the researchers noted, because it is unusual to use prior outcomes and states in order to inform the machine-learning algorithms, and it is especially challenging in the brain because so much is unknown about how the brain works.

Machine learning typically involves two steps: a “training phase” in which the computer evaluates a set of known outcomes--such as, a number of trials in which a subject indicated belief or disbelief--and a second, “prediction” phase in which the computer constructs a boundary based on that knowledge.

In future research, the neuroscientists will be utilizing these machine-learning techniques in a biofeedback framework, showing study participants real-time brain readouts to let them know when they are experiencing cravings and how intense those cravings are, in the hopes of training them to control and inhibit those cravings. But since this distinctly alters the process and cognitive state for the individual, according to the researchers, they may face special challenges in trying to decode a “moving target” and in separating the “training” phase from the “prediction” phase.

Related Links:
University of Michigan
University of California, Los Angeles