Evaluating the Use of FMRI in Cognitive Process Research

This report investigates the application of Functional Magnetic Resonance Imaging (FMRI) in cognitive process research. It explores how FMRI, a functional neuroimaging technique, is used to map brain activity and test cognitive theories. The report discusses the two main categories of functional neuroimaging: localizing cognitive subsystems and testing cognitive theories using neuroimaging data. It examines the concept of reverse inference and its limitations, along with the use of Bayesian analysis. The report also delves into the challenges of avoiding consistency fallacies when testing cognitive theories with neuroimaging data. It further highlights the importance of specificity and sensitivity in cognitive-brain relationships. Through a review of existing research, the report evaluates the contributions and restrictions of FMRI in understanding cognitive processes, considering factors such as localization and the application of cognitive theories. The report also emphasizes the importance of strong assumptions to derive useful insights from neuroimaging experiments. This detailed analysis provides a comprehensive overview of FMRI's role in cognitive process investigation.

Introduction:
Cognitive process is the performance of some composite cognitive activities that are the function
which has certain impact on person’s thinking and mental content. This can also be known as
“process of thinking” or “the process of remembering”. The Functional Magnetic Resonance
Imaging (FMRI) is a technique which used to measure and map noninvasive brain activity. This
advance research is conducted to evaluate the use of FMRI tool to investigate cognitive process.
In the evaluation of the statement determines two categories of functional neuroimaging of
cognition. The first category of functional neuroimaging based on localizing the particular
cognitive subsystems in brain regions of individual. So, this research first define the
neuroimaging data‘s cognitive subsystems independently. The second category of functional
neuroimaging of cognition uses neuroimaging data to test the cognitive theories. Both the
research help to evaluate the use of FMRI tool in the cognitive process. The second category is
useful because cognitive theories effectively define and help to determine the relationships
between cognition and brain activities.
Investigating previous studies:
Functional Magnetic Resonance Imaging is a technique based on functional neuroimaging which
provide a map of one’s brain functioning in response to their cognitive tasks performed during
the scanning of fMRI. The data performed or generated through this technique help the cognitive
neuroscientist to infer particular information about the functioning or role of specific brain
region during the cognitive activities. But, the neuroimaging data is continuously used to make
opposite inference which used to infer the engagement of particular cognitive functions
according to the activity particular part of brain. The reverse inference analysis helps to evaluate
the fMRI uses for investigating cognitive process.
Coltheart (2004) & Levy & Wagne (2013) concluded in their research that FMRI can be used to
produce an activation map that help to indicate that which brain part is involved in a particular
mental process. Numbers of studies has been conducted to determine the involvement of brain
imaging of individual while performing cognitive tasks (Coltheart, 1999), and it is a crucial task
to identify the studies that provided research and investigation on relationship of FMRI tool and
cognition process. The research can be focused on evaluating the neuroimaging contribution

towards our understanding of the cognition process and whether it has restriction to the scope of
evaluation. To underline the evaluation in manageable way, the researchers, (Wixted & Mickes,
2013; Rugg & Thompson-Schill, 2013) have restricted their submission in two ways:
1. The research is carried for only one brain imaging method and that was functional
magnetic resonance imaging (FMRI).
2. Authors also searched for the question that if fMRI actually work for individual’s
cognition process and can it go beyond this and also investigate for the behavioral
methods applied to normal subkest, and neurophysiological work in animals and
investigating fMRI impact on brain lesion patient.
According to Chiao and ImmordinoYang ( 2013) research on neuroimaging informed that the
fMRI technique did not tell us anything about the cognitive theory. The articles gathered on this
topic have addressed some explicit facts about the cognitive theory. The first researched article
reviewed here will help us to evaluate the actual meaning of “informs cognitive theory”. Cabeza,
R., & Moscovitch discussed about the cross-cultural fMRI studies which main focus on
providing judgements on selves or others. The conclusion made by this research indicated that
self-reported cultural values and cultural group can predict the neural responses during self-
judgement (Cabeza, R., & Moscovitch, 2013, p. 59).
The second reviews research article is the work performed by Reuter-Lorenz’s. These
researchers work determine the mechanisms responsible for declining memory and cognitive
abilities (Reuter-Lorenz, 2013, this issue, p. 68). It state that the cognitive theory has been
informed with the work if neuroimaging technique. As the technique is about mapping
noninvasive brain activity while cognitive theories based on different brain functions of people.
Different processes which concentrate on learning analyzes the mental process.
Testing Cognitive Theories With Functional Neuroimaging Data
Cognitive neuroimaging data can be used to test cognitive theories in two different ways. It
might include just a single model to verify or disprove the model or it might include competing
models with the aim of judging between them. A decisive issue here is “consistency fallacy“,
term given by Mole and Klein in (2010). According to them, the data which are constant with a
theory cannot be proposed as evidence in support of that theory, instead some more additional

elements should be included. For a hypothesis, the data must be consistent with the hypothesis
and must also hold back on the contradictions of hypothesis. The articles involving cognitive
neuroimaging generally end up with the statements about equilibrium between cognitive theory
and neuroimaging data. For example, according to Levy and Wagner, (2013),”these two studies
support that there is possibility of memory reactivation during sleep induced by environmental
factors. To point out clearly about the patterns of neuroimaging data is the only way to avoid
consistency fallacies that are inconsistent with the theory obtained in the neuroimaging study.
When a neuroimaging study is used to evaluate a single cognitive theory, then consistency
fallacy is avoided by identifying at least one possible conclusion of the study, which is, that if it
had been obtained, then it would have been inconsistent with what the theory actually says.
Three competing cognitive level hypotheses about the nature of associations and dissociations
between performances on direct versus indirect memory tests was discussed by Cabeza and
Moscovitch. They argue in support of these hypotheses on the grounds that it is more consistent
when we compare it with the results of neuroimaging work, moreover only consistency with
neuroimaging cannot support any cognitive theory. An experiment could have a possible result
that would be inconsistent with component process theory, specifically, a result in which a
hippocampus is activated by relational memory tasks. This exhibits the potential for distortion of
theory in this way aid in dealing with the consistency fallacy. So, it would be a useful practice to
acquire for cognitive neuroimaging researchers.
Rugg and Thompson –Schill and Wixted and Mickes also accentuate that the assumptions must
be strong to get any cognitive theory to predict about the happenings in functional neuroimaging
experiments. The general idea is that if neuroimaging data has to be used like this for testing
cognitive theory has to be elaborated in this way that, if cognitive process C is executed only in
brain region X, then Brain region X helps cognitive process C only. Wager and Atlas referred it
as combination of sensitivity i.e. only one area of brain is sensitive to a particular cognitive
process, and specificity i.e. this area of brain gets activated only by a particular cognitive
process. Blumstein and Amso (2013) assert that when we talk about language , there can be
absence of specificity , as there are evidences that explain that a particular area of brain can be
used in a specified single way of language processing.

There are many articles which also use neuroimaging data for the work of judging between
cognitive theories. E.g. White and Poldrack on two cognitive theories of learning and parkand
Mc Donough on three cognitive theories of ageing. On the basis of these articles some questions
come up to ones mind:
1. What are the cognitive theories that are to be tested by neuroimaging data.
2. In what way each cognitive theory has to be elaborated neutrally.
3. Does this elaboration always appear as a disagreement that, if cognitive process C is executed
only in brain region X, then Brain region X helps cognitive process C only? If this is the case
then what are the particular Cs and Xs for each theory, and are these findings about cognition-
brain relationshio possible.
4. In which way neural elaboration is possible for each theory. Are neuroimaging data available
that are inconsistent with what the theory assumes.
5. If yes, then what ate the assumptions of the theory and what are the neuroimaging data
available that are inconsistent with what the theory assumes.
6. If No, i.e. all the neuroimaging data are consistent with the theory, then then what are the
possible neuroimaging data results which could have been obtained in any of the work. Can it be
proved that the consistency fallacy has been avoided , only if examples of such failures of
assumptions are given.
Wixted and Mickes say that not all the theories of cognition give any assumptions about brain
activity and in such a case cognitive theory cannot be informed by any neuroimaging data. The
theory that ‘cognition is grounded in perception’ predicts that brain areas are active when the
corresponding form of cognition is there. Episodic memory is a particular form of cognition.
This is in accordance with the findings of Wheeler, Peterson, and Buckner, which says that the
visual and auditory aspects prompts activity in the same areas of visual and auditory cortex that
are activated during the initial perception of those things.
Rugg and Thompson (2013) offer embodied- cognition theory which can be tested by
neuroimaging data and which is an example of a cognitive theory. They focus on the area of

colour in their embodied- cognition theory where they discuss the results of work on
neuroimaging of Hsu, Kraemer, Olivier, Schlichting, and Thompson- Schill (2011), who
compared color perception task to a color memory task.
The color- perception task was based on the adjudication of lightness or darkness of chromatic or
achromatic visual stimuli. Brain regions involved in color perception were identified by
subtracting brain activations in the achromatic condition from brain activation in chromatic
condition.
The color- memory task involved presentation of two printed words i.e. names of objects with
characteristic colors at the top of the screen and at the bottom of the computer, and the task given
was to find out that which two bottom words are most similar to the top word.
The cognitive theory states that accessing a conceptual feature occupies the same processes that
are active when the characteristic is directly expressed.
Localization
The interest of people is increasing for using the functional neuroimaging techniques in order to
gain knowledge about cognition process. One common method or practice commonly used by
the researchers in this particular context is ‘reverse inference’. With the help of this reverse
inference the engagement of specific cognitive process is inferred from the activation of
particular region of brain. One cannot ensure about the validity of such inferences but can still
trust for providing some information. Bayesian analysisof BrainMap Neuroimaging database can
be used to characterize the amount of additional evidence for determining cognitive process
relation with reverse inference. The usefulness of such activity is limited and also valid for
particular region.
Probabilistic graph is used to determine the relationship between cognitive process, manipulation
and observed variables, it is known as Bayesian network, provided in figure 1. The nodes in the
graph denoted as entities and the edges of the graph known as conditional probabilities. There
are significant features in reverse inference and features are represented in this graph. First thing
cleared in the graph is about the cognitive process and it shows that those processes are
conditioned on the particular task manipulation. For example, before utilization of cognitive

process the particular task is being already performed. Second highlighted part in this graph
states about the strength reveres inference dependence on the degree to which the edge interest is
stronger than all other edges which leads the activation. If the probabilities of other edges in the
process are zero then the mapping between cognitive process and fMRI activation is one-to-one.
Third, it reminds us that fMRI data are not alone in suffering from the reverse inference problem:
Reverse inference based on any observable data (e.g. behavioral data) is limited by the same
characteristics.
Figure 1: A probabilistic graph representing the relationships between experimental
manipulations, cognitive processes and observed variables
Evaluate the skepticism which investigated whether localizing any particular cognitive process
in the brain can inform psychologist about how actual work process (i.e. could it really inform or
tell you anything about the cognition process). For example, with particular technique if one can
find a cell that recognize about particular individual’s grandparents does not indicate much about
brain processing because anyone can recognize their grandparents. The most important part is
how a brain cell can does it. This discovery may inform what does it, but not how it can be done.
According to the argument in the Coltheart (2004) article, to localize the cognitive
system model, the important thing to know is about what cognitive system modules actually
states. So, it is important to start the work with investigating model and then localization of
research. Imagining is not considered as model here but it depends on the modeling. Wixted and

Mickes (2013) also pointed out the similar point, and stated that the research conducted to
localize cognitive processes in brain through functional neuroimaging should need to connect or
have connection with specific cognitive-level theory which needs to interpret the neuroimagining
data. Instead of highlighting the cognitive theories, the localization studies focused on cognitive
theories to interpret the data. The interpretation is valid only if the cognitive theory on which it
based is valid.
The research generated through reviewing the article generate the conflicting data and
does not find any evident which shows that any particular form of brain activation indicate that
particular form of cognitive operation is performed (Hsu, Kraemer, Oliver, Schlichting, 2011).
But, it may possible that the future invention can get evidence of the similar interaction. In that
situation localization could be used to investigate and test the cognitive theories but it surely not
possible in today’s condition.
In general, then we can state that the fMRI technique used for brain mapping and aimed
to localize cognitive functions in not relevant for special section, because such fMRI work does
not inform cognitive theory. In fact, fMRI work is informed by cognitive theory.
Conclusions
It is a critical research topic and at the same time vastly researched too. Numerous researches are
available on the topic and all are excited about investing the ability of functional neuroimaging
to discover the organization of cognitive processes. The evaluation data performed here
concludes that caution should be taken while using reverse inference, particularly in the case
when the engagement of a cognitive process and selectivity of activation in particular region is
low. Reverse inference must be considered as different tool through which one can advance
understanding between the mind and brain. This might be true as per the evaluation of data
presented in this research, but the true success and usefulness of the reverse inference strategy is
when it successfully understands and propose connectivity between mind and brain.

References
Blumstein, S. E., & Amso, D. (2013). Dynamic functional organization of language: Insights
from functional neuroimaging. Perspectives on Psychological Science, 8, 44–48.
Cabeza, R., & Moscovitch, M. (2013). Memory systems, processing modes, and components:
Functional neuroimaging evidence. Perspectives on Psychological Science, 8, 49–55.
Chiao, J. Y., & Immordino-Yang, M. H. (2013). Modularity and the cultural mind:
Contributions of cultural neuroscience to cognitive theory. Perspectives on Psychological
Science, 8, 56–61.
Coltheart, M. (2004). Brain imaging, connectionism, and cognitive neuropsychology. Cognitive
Neuropsychology, 21, 21–25.
Hsu, N. S., Kraemer, D. J. M., Oliver, R. T., Schlichting, M. L., & Thompson-Schill, S. L.
(2011). Color, context, and cognitive style: Variations in color knowledge retrieval as a function
of task and subject variables. Journal of Cognitive Neuroscience, 23, 2544–2557.
Levy, B. J., & Wagner, A. D. (2013). Measuring memory reactivation with functional MRI:
Implications for psychological theory. Perspectives on Psychological Science, 8, 72–79.

Mole, C., & Klein, C. (2010). Confirmation, refutation and the evidence of fMRI. In S. J. Hanson
& M. Bunzl (Eds.), Foundational issues of human brain mapping (pp. 99–112). Cambridge, MA:
MIT Press.
Petersen, S. E., Fox, P. T., Posner, M. I., Mintun, M. A., & Raichle, M. E. (1988). Positron
emission tomographic studies of the cortical anatomy of single word processing. Nature, 331,
585– 589.
Posner, M. I., Petersen, S. E., Fox, P. T., & Raichle, M. E. (1988). Localization of cognitive
operations in the human brain. Science, 240, 1627–1631.
Reuter-Lorenz, P. A. (2013). Aging and cognitive neuroimaging: A fertile union. Perspectives
on Psychological Science, 8, 68–71.
Rugg, M. D., & Thompson-Schill, S. L. (2013). Moving forward with fMRI data. Perspectives
on Psychological Science, 8, 84–87.
Wager, T. D., & Atlas, L. Y. (2013). How is pain influenced by cognition? Neuroimaging
weighs in. Perspectives on Psychological Science, 8, 91–97.
Wheeler, M. E., Petersen, S. E., & Buckner, R. L. (2000). Memory’s echo: Vivid remembering
activates sensory-specific cortex. Proceedings of the National Academy of Sciences, USA, 97,
11125– 11129.
Wixted, J. T., & Mickes, L. (2013). On the relationship between fMRI and theories of cognition:
The arrow points in both directions. Perspectives on Psychological Science, 8, 104–107.