One potential tool to objectively measure how health messages are received is neuroimaging under the umbrella of neuroscience. Neuroscience can be subdivided into several areas of interest. Communication neuroscience offers insight into understanding health communication (Falk, 2013) and helps bolster knowledge in nonverbal communication (Todorov et al., 2008);
offering additional but supplementary information to traditional methods (e.g. focus groups).
Communication neuroscience can be explored using an increasing number of available neuroimaging tools. These neuroimaging techniques can help identify what regions of the brain become activated whilst completing specific functions (or tasks) or processes; including emotion and affection, attention, social cognition, reasoning and language (Bookheimer, 2007;
Cacioppo, 2002). The most commonly used neuroimaging techniques are electroencephalography (EEG), functional magnetic resonance imaging (fMRI), eye tracking and functional near infrared spectroscopy (fNIRS).
Neuroimaging tools
Electroencephalography, functional NIRS and event-related potentials
Despite the abundance of neuroimaging tools available, each neuroimaging tool offers respective positive and negative attributes. Electroencephalography records brain signals from the scalp and identifies changes in these signals along the spectral bands of delta, theta, alpha, beta and gamma. Because people are often unable or unwilling to justify or explain preferences when prompted, in part attributable to human behaviours driven by unconscious awareness (Calvert & Brammer, 2012), neuroimaging techniques can be crucial. Eye tracking, in contrast, measure eye movements which are considered the most frequent human behaviour (Bridgeman, 1992). Because the visual system provides an enormous amount of information, eye tracking is a key tool to try and understand what motivates people to act in certain ways.
In previous years, eye tracking relied on direct observation of eye movements; limiting the measurement accuracy to the memory and accuracy of the observer (Dodge, 1906). The need to implement a better, objective record of eye movements using non-invasive methods was acknowledged (Dodge & Cline, 1901) and subsequent efforts evolved (Taylor, 1971). EEG and eye tracking are considered the least invasive neuroimaging tools yet still provide high
32
temporal resolution. fNIRS, despite measuring brain activation in a different way, does offer greater portability and relatively lower costs which may encourage investigations in more naturalistic environments.
Magnetic resonance imaging
Magnetic resonance imaging (MRI) has made a substantial contribution to neuroscience by permitting imaging of the brain. Its history lies in nuclear magnetic resonance (NMR) spectroscopy which largely relies on the angular momentum possessed by subatomic particles (i.e. protons, neutrons and electrons). Over time, NMR produced images using detection coils to align with the resonance frequency of hydrogen which was used to calculate water density.
Given the prevalence of water in the human body, hydrogen is the most commonly studied element with MRI. Sir Peter Mansfield subsequently developed this approach further, producing methods to analyse these images and the approach later became known as MRI.
Magnetic field strengths are typically 0.1-10T (tesla) and apply a strong magnetic field to align the spinning proton. The MRI scanner, such as the one shown in Figure 1.7 (MR750w 3T scanner [General Electric Healthcare, Chicago, IL, USA]), then produces a series of radio frequency currents to create a varying magnetic field. The protons absorb this energy and flip their direction of spin, which is maintained until the radiofrequency field is switched off. Upon switching off, the protons return to their normal spinning motion, and in the process, produces a radio signal. This returning to normal phase can be measured and subsequently made into an image through radio frequency coils. In relation to imaging anatomy, the MRI scanner distinguishes between differing tissue types by the speed at which the protons return to their normal spins. The loud noise anecdotally aligned to MRI scans is due to the constant flipping motion of magnetic fields. Unfortunately, because MRI uses magnets, individuals who possess any metal implants cannot go inside as they are not magnetic resonance safe and pose a hazard.
MR images can be acquired with a range of image contrasts, such as T1, T2, or diffusion weightings, which indicate underlying structure. They may also be acquired with contrast dependent on physiological processes, such as blood flow, which can be manipulated to reveal angiographic information or functional MRI via the BOLD effect, as described below.
33
Figure 1.7. An MR750w 3T scanner (left) with a participant being prepared to go inside (right) Functional magnetic resonance imaging
fMRI is a non-invasive neuroimaging technique that studies the brain whilst an individual completes a cognitive task inside an MRI scanner (Figure 1.7). More specifically, fMRI provides images that show the location of magnetic resonance signal changes associated with neural activity. fMRI works on the basis that a vascular change occurs when neural tissue is activated (Ogawa et al., 1990). Using a method called blood oxygen level-dependent (BOLD) contrast imaging, fMRI provides an indirect measure of neuro-electric activity (Logothetis et al., 2001). An early observation advocated that changes in neural activity resulted in signal changes that can take seconds to develop and decay (Bandettini, 1993). The theory behind BOLD contrasts is supported because deoxyhaemoglobin is paramagnetic in nature meaning it causes reductions in signal strength in the vasculature and surrounding tissue. Cerebral blood volume and blood flow increase when an area of the brain is activated; resulting in a lower oxygen extraction fraction of the blood. Blood supply demand is subsequently exceeded which causes a reduction in deoxyhaemoglobin. As deoxyhaemoglobin decreases, the paramagnetic properties are removed which results in a greater signal intensity. Therefore, an activated region of the brain demonstrates a more intense signal which can reveal a temporal measure of neural activity. This occurs after a haemodynamic filter has smoothed the pattern of activation which, in the process, can slightly delay signal production (Aguirre et al., 1998). While arterial blood is similar in its magnetic properties to tissue, deoxygenated blood is paramagnetic and so induces in-homogeneities within the magnetic field in tissue. As a result, the magnetic resonance imaging signal decays faster but signals from activated regions of cortex increase as the tissue becomes more magnetically uniform. Dynamic increases in volume and flow of
34
blood to an activated region of the brain, accompanied by changes in oxygen consumption, occur shortly after cognitive stimulation (Leniger-Follert & Lübbers, 1976). To localise these neural activations, low resolution images are acquired in rapid succession to produce mapped brain volumes every few seconds. In combination, these volumes produce a time-series of activation intensities for each voxel. fMRI produces relatively good spatial resolution with whole brain coverage, but the technique suffers from poor temporal resolution and issues surrounding reverse inference. Reverse inference suggests that activations of specific brain regions infer the engagement of a specific cognitive process which is not fully valid (Poldrack, 2006). There has been a recent expansion of interest in using fMRI as a neuroimaging tool bringing forward both scepticism and enthusiasm (Aue et al., 2009). However, having the capacity to measure specific cerebral structures in social cognition and behaviour has been noted as an outstanding achievement in contemporary neuroscience (Eisenberg, 1995).
Functional MRI, health messages and behaviour change
A systematic review conducted by Kaye and colleagues (Kaye et al., 2016) identified a variety of neuroimaging studies focusing on key human behaviours including smoking (Chua et al., 2011), nutrition (Kessels et al., 2011), sun safety (Falk et al., 2010), narcotic substances (Weber et al., 2015), safe sex (e.g. Seelig et al., 2014) and blood donation (Falk et al., 2010). The review identified twenty studies that employed event-related potential, functional near infrared spectroscopy or fMRI and demonstrated a growing body of research assessing visual stimuli.
However, expanding the scope of the review to identify studies on other lifestyle behaviours may have revealed further studies. Lifestyle behaviour health messages are often disseminated on packaging to deter individuals or billboards to highlight health implications. Studies to date have compared persuasive and unpersuasive messages (Falk et al., 2010), tailored and untailored messages (Noar et al., 2007) as well as images of lifestyle behaviour (Jackson et al., 2014). Functional MRI studies often examine neural activation patterns in response to stimuli;
offering insight into how people cortically respond. Understanding how messages can be made more potent (or persuasive) to encourage behaviour change is likely crucial. fMRI studies can also investigate human behaviour following exposure to a stimulus, such as health messages.
For instance, after viewing anti-smoking advertisements, participants were subsequently measured at follow up for smoking consumption using exhaled carbon monoxide (a proxy indicator). Findings revealed that neural activity in response to anti-smoking advertisements accounted for 20% of the variance in how much exhaled carbon monoxide was recorded (i.e.
how many were still smoking) (Falk et al., 2011). Moreover, the medial prefrontal cortex, a
35
region of the brain, acted as a surrogate marker for subsequent smoking cessation. In another study, using EEG, Versace and colleagues examined rates of smoking cessation and identified that neural patterns to emotional and smoking-related pictures had a role in predicting subsequent smoking cessation (Versace et al., 2011). More specifically, smokers with lower levels of neural activation in response to pleasant stimuli were less successful at ceasing smoking habits. These encouraging findings support the suggestion that social neuroscientists should examine new forms of media, such as social network sites and smartphones, to assess the role of technology in health communications (Cascio et al., 2013). The authors emphasise that people are affected differently by health message communications and subsequently act differently after exposure. However, in combination, neuroimaging tools (quantifying neural activity) and self-report surveys explain some variation related to behaviour change (Cascio et al., 2013). As a result, conducting more studies that employ objective measurements (using techniques such as fMRI) may be warranted.