Machine learning prediction of neurocognitive impairment among people with HIV using clinical and multimodal magnetic resonance imaging data.

Diagnosis of HIV-associated neurocognitive impairment (NCI) continues to be a clinical challenge. The purpose of this study was to develop a prediction model for NCI among people with HIV using clinical- and magnetic resonance imaging (MRI)-derived features. The sample included 101 adults with chronic HIV disease. NCI was determined using a standardized neuropsychological testing battery comprised of seven domains. MRI features included gray matter volume from high-resolution anatomical scans and white matter integrity from diffusion-weighted imaging. Clinical features included demographics, substance use, and routine laboratory tests. Least Absolute Shrinkage and Selection Operator Logistic regression was used to perform variable selection on MRI features. These features were subsequently used to train a support vector machine (SVM) to predict NCI. Three different classification tasks were performed: one used only clinical features; a second used only selected MRI features; a third used both clinical and selected MRI features. Model performance was evaluated by area under the receiver operating characteristic curve (AUC), accuracy, sensitivity, and specificity with a tenfold cross-validation. The SVM classifier that combined selected MRI with clinical features outperformed the model using clinical features or MRI features alone (AUC: 0.83 vs. 0.62 vs. 0.79; accuracy: 0.80 vs. 0.65 vs. 0.72; sensitivity: 0.86 vs. 0.85 vs. 0.86; specificity: 0.71 vs. 0.37 vs. 0.52). Our results provide preliminary evidence that combining clinical and MRI features can increase accuracy in predicting NCI and could be developed as a potential tool for NCI diagnosis in HIV clinical practice.

Smoking automaticity and tolerance moderate brain activation during explore-exploit behavior.

The adaptive trade-off between exploration and exploitation is a key component in models of reinforcement learning. Over the past decade, these models have been applied to the study of reward-seeking behavior. Drugs of addiction induce reward-seeking behavior and modify its underlying neurophysiological processes. These neurophysiological changes may underlie a behavioral shift from a flexible, exploratory mode to a focused, exploitative mode, which precedes the development of inflexible, habitual drug use. The goal of this study was to investigate the relationship between explore/exploit behavior and drug addiction by examining the neural correlates of this behavior in cigarette smokers. Participants (n=22) with a range of smoking behaviors completed a smoking dependence motives questionnaire and played a 6-armed bandit task while undergoing functional magnetic resonance imaging (fMRI). Exploratory behavior produced greater activation in the bilateral superior parietal and bilateral frontal cortices than exploitative behavior. Exploitative behavior produced greater activation in the bilateral superior and middle temporal gyri than exploratory behavior. fMRI data and orthogonalized smoking dependence motive scores were entered into multiple linear regression analyses. After controlling for nicotine tolerance, smoking automaticity positively correlated with activation in the same bilateral parietal regions preferentially activated by exploratory choices. These preliminary results link smoking dependence motives to variation in the neural processes that mediate exploratory decision making.

Smoking and the bandit: a preliminary study of smoker and nonsmoker differences in exploratory behavior measured with a multiarmed bandit task.

Advantageous decision-making is an adaptive trade-off between exploring alternatives and exploiting the most rewarding option. This trade-off may be related to maladaptive decision-making associated with nicotine dependence; however, explore/exploit behavior has not been previously investigated in the context of addiction. The explore/exploit trade-off is captured by the multiarmed bandit task, in which different arms of a slot machine are chosen to discover the relative payoffs. The goal of this study was to preliminarily investigate whether smokers differ from nonsmokers in their degree of exploratory behavior. Smokers (n = 18) and nonsmokers (n = 17) completed a 6-armed bandit task as well as self-report measures of behavior and personality traits. Smokers were found to exhibit less exploratory behavior (i.e., made fewer switches between slot machine arms) than nonsmokers within the first 300 trials of the bandit task. The overall proportion of exploratory choices negatively correlated with self-reported measures of delay aversion and nonplanning impulsivity. These preliminary results suggest that smokers make fewer initial exploratory choices on the bandit task. The bandit task is a promising measure that could provide valuable insights into how nicotine use and dependence is associated with explore/exploit decision-making.