Brain Imaging Techniques

A modern approach in studying the human brain is neuroimaging which allows scientists to look at the structural or functional aspects of the nervous system. There are two dimensions to be considered when using functional measures that are available for neuroscience research – spatial resolution which is how specific the source of signal can be localised and temporal resolution which is the time scale of the particular measurement. None of the brain imaging techniques is the ‘magic answer’ since each has its own advantages and shortcomings. This essay will focus on the studies of music perception and reading in particular to demonstrate the uses of these techniques in a complimentary fashion. Due to the complexity of these cognitive capacities as well as the different features of each imaging technique, results obtained across multiple experiments using different brain imaging techniques have to be analysed to address the specificities of the brain structures involved in both processes.

In studying the human brain, early researchers has studied patients with brain damage such as lesion and stroke to determine the parts of the brain that are responsible for different human capacities from the patients’ specific abilities are disrupted. However, since brain damages can be extensive and hard to localize, and there are individual variability, this method provides only a simplistic and rudimentary mapping of function to structure and offers limited views of normal brain function. With the advent of technology, there is an increasing accessibility to brain imaging techniques which provide non-invasive ways to look at the brain with more precision and across a wide range of subjects. Neuroimaging can be divided into two broad categories – recording of brain structure which images anatomy of the nervous system, such as magnetic resonance imaging (MRI) and computed tomography (CT), and functional measures which investigates brain activities during cognitive processes, such as electroencephalogram (EEG), functional magnetic resonance imaging (fMRI) and positron emission tomography (PET). Still, none of these imaging techniques on its own can provide an all-inclusive answer to a particular cognitive process due to the complexity of brain functioning and respective strengths and limitations each technique. Thus, neuroimaging techniques are often used in a complementary fashion to study brain function. This essay will focus on the functional neuroimaging techniques EEG and fMRI, and demonstrate their use in studying music perception and reading, which are both high-level cognitive functions of the human brain.

In the area of music perception, a melody is an organized series of individual tones on the physical level. However, it is more complex on a perceptual level. A musical melody can be defined as the succession of tones which is processed in terms of multiple structured relationships. Melody perception thus emphasizes the active role of human brain, meaning that it relies on the listener’s perception to take a sequence of sounds and transform them into a meaningful mental experience. The multifaceted nature of melody signifies that even simple tunes engage multiple sets of mental operations used for interpreting tonal relations. Therefore, the neuroscience of melody aims to ‘tap into the moment-to-moment history of mental involvement with the music’. Apart from studying melodic perception deficits in individuals with localized brain damage, one primary approach to identifying brain regions involved in melody perception is the haemodynamic approach based on techniques such as Functional magnetic resonance imaging (fMRI). FMRI uses a large magnetic field to measure differences in blood flow (hemodynamic response) across the brain and images in excellent spatial resolution. Active neurons consume oxygen and convert oxyhemoglobin into deoxyhemoglobin, thus an increase in oxygen level in a particular neural structure indicates activity in that brain area. The change in the concentration of deoxyhemoglobin in the blood is known as the blood-oxygen-level dependent (BOLD) contrast. For example, in a study using whole-brain fMRI data, subjects were played a series of tones and told to discriminate between two pitches. (Binder et al, 2007) This task required participants to not only perceive the pitch but also to maintain that information in their working memory. The data was compared to one that involves semantic categorization of words and showed that the tone judgement task was related to stronger activation of right temporoparietal areas such as the right posterior middle temporal gyrus and right superior parietal cortex. However, blood flow and metabolism are grows and decay over many seconds in response to the physiological demands of the underlying neural tissue, and thus produce sluggish signals. Therefore, fMRI has relatively poor temporal resolution and unable to operate at the rapid time-scale of melody perception. A method that produces data with high temporal resolution is the event related potential approach (ERP) which measures brain response that is direct result of a specific cognitive event. This is achieved by extracting population-level neural activity from Electroencephalography (EEG), which is time-locked to the stimulus. EEG measures the electrical activity on the scalp of a person. It is sensitive to postsynaptic dendritic currents generated by a population of neurons that fire synchronously. ERPs provide time resolution on the order of 10s to 100s of milliseconds, which is thus suitable for examining the neural response to individual tones in melodies in fine temporal detail. For instance, to investigate the influence of musical training on the brain’s processing of tonality relations, Besson and Faita studied musicians or nonmusicians who listened to musical phrases in the experiment. ERP was used to record subjects’ neural responses to a particular tone in multiple repetitions of similar melodies, such as an out-of-key note at the end of a melody. It was revealed that musicians and nonmusicians ERP to the end notes were different in terms of latency and amplitude, and that musical expertise affects decisional aspects of musical processing more so than solely perceptual aspects. Unfortunately, the brain sources of ERPs are difficult to localize due to the spatial spreading of bioelectric currents by the skull and scalp. The poor spatial resolution renders ERPs unsuitable for localization, unlike fMRI. The above illustrates that different brain imaging techniques can target different research areas regarding one particular cognitive process – fMRI analysis examined the overall response of various brain regions to entire tone sequences, while using EEG helped examine the temporal details of neural responses to individual tones. By taking both the haemodynamic approach and the evoked potential approach, scientists can obtain valuable information about melody and the brain to better investigate the complex nature of musical melody perception.

Similar to the study of music perception, a combination of brain imaging techniques has been used in studying the cognitive process of reading. In particular, these techniques used in complementary fashion helped informed the theory of phonological processing, which involves analysing and manipulating sound structures of words. There has been a long-standing debate over the role of phonological processing in skilled reading with some theorists offering online-processing data suggesting that there is rapid activation of phonological representation during reading; whereas other theorists highlighted neuropsychological dissociations between phonological and lexical processing. (Frost, 1998) FMRI analysis can indicate whether there is differential activation of the brain between normal readers and individuals with dyslexia – a learning disability characterised by trouble reading despite a normal intelligence. Neuroimaging results have shown that when solving phonological decoding tasks such as reading pronounceable non-words, dyslexic subjects, in comparison to control subjects, show a hyperactivation of Broca’s area which has functions linked to speech production. This reflects dyslexics’ increasing effort concerning phonological coding which can explain their delay in reading. However, due to the poor temporal resolution of fMRI, using fMRI alone cannot determine when these phonological effects arise during word processing – whether they occur early in the visual recognition of words or are the result of later feedback effects. To answer this question, techniques which provide a greater temporal resolution at the cost of lower spatial resolution can be used. Research conducted by Georgiewa et al therefore merged the fMRI study with a separate ERP study using similar reading tasks on the same subjects to investigate the group differences between dyslexic and normal-reading subjects. When subjects read non-words which demands phonological encoding,, the ERP recording showed that scalp topography between the dyslexic and normal groups had a significant difference. The time window of 250–500 ms in word reading is related to phonologic and lexical encoding as proposed by Posner. In this time window at which the biggest differences between the two groups were found in the left brain region in the fMRI data, a positive ERP-component is more accentuated in the dyslexic group in comparison to the control group. By using techniques that characterize the time course of early visual processes in reading, these early processes are shown to be delayed in individuals with dyslexia. A combination of brain imaging techniques are thus able to show that phonological variables can influence regions in the ventral visual stream which represents the cognitive route from orthography to semantics. This has provided converging evidence for the early phonology theory which emphasized the phonological processing component of reading. At the same time, such research highlighted the limitations of neuroimaging and the need to rely on multiple techniques to better inform a research topic in brain functioning.

As shown above, in studying music perception, analysis from fMRI with high spatial resolution can point to specific brain regions associated with the maintenance of a tone in memory; while ERP with high temporal resolution

The above illustrates that to address the complexity of brain functioning, it is important to utilize different features of brain imaging techniques to investigate a multifaceted brain process, as well as use the converging evidence from these techniques to inform theories of the human mind.