The use of fMRI to analyze brain plasticity

Published: Last Edited:

This essay has been submitted by a student. This is not an example of the work written by our professional essay writers.

Technology helps us to understand the underlying mechanisms of our body. There are a few non invasive techniques to quantify brain activity. In this essay I will discuss the use of fMRI to analyze brain plasticity in the case of anterior cruciate ligament (ACL) related injuries. Kapreli et al. performed a fMRI study to investigate whether chronic ACL injury causes plastic changes in brain activation patterns. Before we can discuss this subject we first need to know the role of the ACL in proprioception. Which structures are responsible for communicating with the central nervous system? Is the fMRI technique, used during the study, advanced enough to distinguish the different areas of activation? Are the performed tasks specific tasks involving the ACL? These questions and more will be discussed in this essay.

Used method and applications

The ACL group includes seventeen right leg-dominant male participants with chronic anterior cruciate ligament deficiency. The control group includes eighteen matched participants with no special sport or habitual physical activity. Selection criteria for the ACL group comprised a complete right unilateral anterior cruciate ligament rupture minimal six months before testing. The brain activation patterns where examined by using fMRI technique (1,5-T).

All ACL ruptures were confirmed by MRI and examination by an orthopaedic surgeon.

Each session included 8 functional time series, each consisting of 120 whole brain gradient-echo echoplanar images, which were acquired every 2.5 seconds, with anterior-posterior phase encoding direction, followed by an anatomical 3D-high-resolution T1 weighted image.

The following specifications were used for the functional time series scans:

Repetition time[TR]/Echo time[TE]= 2500 / 50 ms

Field of view : 200 mm

Matrix: 64 x 64

Slice thickness: 4.85 mm

30 sagittal slices

The following specifications were used for the high resolution T1 scan:

[TR]/[TE] = 22/4.6 ms

Field of view: 220 mm

Matrix: 256 x 256

Slice thickness: 1 mm

150 sagittal slices

During the experiment, participants performed an unilateral movement of extension-flexion of the right knee. The participants were placed supine inside the scanner with their legs in a custom-built cushion that limited the range of knee flexion tot 45o. Several precautions were taken to minimize the movement of the head during measurements. Each unilateral movement lasted 25 seconds, during which 10 whole brain images were acquired.

The anatomy of the anterior cruciate ligament(ACL)

The ACL has a prominent role in the knee joint. The ACL resists anterior tibial translation and rotational loads. At this moment surgical reconstruction is the standard treatment in the field of sports medicine. The reconstruction aims on restoring the kinematics and the stability of the injured knee. First I will provide some information about the macroanatomy, second I will provide some information about the micro-anatomical composition of the ACL and in the end I will discuss some important information for a good understanding of this essay; the innervation of the ACL.

Macroanatomy of the ACL

(Duthon et al., 2006)

The ACL is a band-like structure of dens connective tissues. The bony attachement on the femur, is located at the posterior part of the inner surface of the latral femoral condyle. From this fermoral attachment, the ACL runs anteriorly, medially and distally to the tibia. The length of the ACL ranges from 22 to 41 mm (mean, 32 mm) and its width ranges from 7 to 12 mm.

Analogue to the change of the angle of flexion, the shape of the ACL changes. The most proximate part of the ACL is attached with an cross-sectional area of 34 mm2 and the most distal part of the ACL is attached with an cross-sectional area of 42 mm2.

Functionally, Girgis et al. divided the ACL in two parts:

Anteromedial bundle (AMB)

Posterolateral bundle (PLB)

Relatively to the AMB, the PLB contains a larger number of fascicles.

Microanatomy of the ACL

(Duthon et al., 2006):

The ACL is composed of multiple fascicles consisting mostly of collagen.

Microscopically, there are three zones to distinguish within the ACL:

Proximal part

Less solid

Highly cellular, rich in round and ovoid cells

Fusiform fibroblasts

Collagen type II


Middle part

Low cellularity

Fusiform and elongated spindle shaped fibroblasts

High density of collagen fibers

Zone of cartilage and fibro cartilage

Elastic (recurrent maximal stress) and oxytalan (multidirectional stress) fibers

Distal part

Most solid

Highly cellular


Ovoid fibroblasts

Low density of collagen fibers

High level of cellular activity


(Duthon et al., 2006)

The nerve fibers from the posterior articular branches of the tibial nerve innervate the ACL. These nerve fibers penetrate the posterior joint capsule and run along with the synovial and periligamentous vessel surrounding the ligament to reach as far anterior as the infrapatellar fat pad.

Most of the fibers have a vasomotor function, which means that they are involved in the constriction and dilation of the blood vessels. Besides the nerve fibers mentioned above, there are two more kind of fibers present in the ACL:

Myelinated nerve fibers

Unmyelinated nerve fibers

These nerve fibers have been observed coursing independently of the vessels and consisting the following receptors:

Ruffini receptors

Sensitive to stretching

Vater-Pacini receptors

Sensitive to rapid movements

Golgi-like tension receptors

Sensitive to stretching

Located near the attachments of the ACL and at the ACL's surface.

Free-nerve endings

Function as nocireceptors.

The Ruffini, Pacini and Golgi-like receptors have a proprioceptive function and provide the afferent for signaling knee postural changes. The activation of afferent nerve fibers in the proximal part of the ACL influences motor activity in the muscles around the knee; "ACL reflex".

The ACL reflex is involved in the updating of the muscle programs, and is therefore an essential part of the normal knee function.


fMRI is an specialized type of MRI scanning and is used a lot in neuroimaging. The fMRI technique used in the article is based on the BOLD contrast.

fMRI principles

(Fillipi et al., 2009)

fMRI and BOLD is described by Fillipi et al.:

"The BOLD contrast is closely linked to two physical phenomena, called "diamagnetism" and "paramagnetism"."

… "If a diamagnetic substance is brought into an external magnetic field, it tends to decrease slightly this field, whereas a paramagnetic substance tends to increase it."

… "Tissue is mainly diamagnetic. In contrast, blood contains a certain level of deoxyhaemoglobin which is paramagnetic. The presence of blood in tissue means a close vicinity of substances with different magnetic properties, giving rise to microscopic field distortions. The resulting field gradients cause spin dephasing and lower the T2* value."

… "After neuronal activation, blood is locally hyperoxygenated, corresponding to a wash-out of deoxyhaemoglobin and an increased concentration of oxyhaemoglobin. In contrast to deoxyhaemoglobin, oxyhaemoglobin is diamagnetic, so it has similar magnetic properties as tissue, leading to a more homogenous magnetic field and an increased signal intensity in T2* weighted images."

… "Quantification of the T2* signal enhancement therefore allows for the detection of neuronal activation."

fMRI can be used to quantify an approximation of brain activity. By using fMRI a derivation of the brain activity is measured. Instead of measuring the real membrane potentials, activation is measured by determining the concentration of oxyhaemaglobin compared to different areas. A relatively higher signal would be interpreted as an increase in brain activity. Because oxygen is needed for brain activity.

fMRI has advantages, as stated earlier, but also has some disadvantages:

Interpretation differences; correlation does not imply causality.

Statistical methods must be used with care. Wrong use of statistical methods can lead to false positives.

As mentioned before, the BOLD signal is an indirect measure of neural activity. Changes could be influenced by non-neural changes.

fMRI lacks temporal resolution.

BOLD response can be affected by a variety of factors:



Brain pathology

Local differences in neurovascular coupling


Composition of the blood

Because of the indirect measurement it is very important to evaluate the data to be sure it is valid. Another problem is the temporal resolution of fMRI, because of its poor temporal resolution, a second T1 weighted image is needed to obtain an high resolution image in which the fMRI measurement can be projected. The translation from fMRI data to an T1 weighted image could give another artifact, which would lead to an incorrect image. Nowadays it is possible to use a more advanced fMRI scanning method like 3-T or even 7-T scanning techniques. The great advantage of these advanced techniques, as compared to 1,5-T scanning, is the improved signal to noise ratio. The more advanced techniques give the possibility to distinguish more accurately different parts of the brain(T1 weighted imaging) and the different brain activation patterns (fMRI).

Discussing the null hypothesis

Kapreli et al. use the following null hypothesis:

"… ACL patients do not present a different activation pattern of the brain compaired with healthy controls."

The null hypothesis is not focused at all. In this study the scientists aim on the activation patterns of the whole brain, instead of focusing on a specific part which would be expected to be involved particularly in ACL proprioception. It seems to be obvious that a ruptured ACL would lead to deficiency of proprioception. But it is uncertain whether the change of brain activation patterns is a direct consequence of the ACL rupture. Indeed, the ACL rupture could cause secondary changes in brain activation patterns. For example, in absence of an ACL reflex there would be a missing link in the updating mechanism of the musculature of the lower limb. Especially in patients with a ruptured ACL, the loss of feedback from mechanoreceptors in the ACL leads to weakness of the quadriceps femoris. (Duthon et al., 2006) Maybe, the change in the amount of muscle tissue and the adaptive tasks of the muscles surrounding the knee joint could cause brain plasticity. The authors of the article never mentioned whether they wanted to measure direct consequences or secondary changes within the brain activation patterns. Probably the data acquired during this study does not contain the change in activation patterns due to ACL deficiency but the acquired data probably does contains plastic changes which can not be or only partially be related to the ACL rupture.

Discussing the used methods

When conducting scientific research, the validity of an experiment is vital. In other words, do I measure what I want to measure? Kapreli et al. state that the overall brain activity of the control group, compared to the ACL group would be higher.

The ACL is a structure which is involved mostly in dynamic actions. As mentioned in its anatomy, the ACL is mostly involved in anterior tibial translation and rotational loads. The method used in this study contains a controlled motion without external loads and without any resistance of the motion(only a stop at 45o). Risberg et al. describe the importance of neuromuscular training versus strength training, in the first 6 months after reconstruction, in a randomized clinical trial. They conclude that neuromuscular training in combination with strength training has a positive effect on the motor control. According to their conclusion it seems possible that a major part of the motor control deficiency of the injured lower limb could be adopted by the surrounding structures(mostly muscles). During the neuromuscular training the patient needs to fulfill tasks with increasing neuromuscular effort which is part of a learning process. It could be possible that during these tasks, which are, in particular, related to the natural motions involving the ACL, brain plasticity would occur. The task which is included in the discussed method is probably not a ACL specific task. The performed task is a simple movement with low proprioception needed, while the ACL is highly involved in more sophisticated and advanced movements. Just a thought from my side, could it be that simple tasks do not need that much intervention of mechanoreceptors of the ACL instead of advanced tasks, which are not performed in this study?

This brings us to a second point of discussion in case of the used method. It is quite impossible to measure highly dynamic motions with fMRI. As mentioned in the principles, MRI and fMRI are highly sensitive to head motion. Besides, the measurement is performed in a horizontal position, that makes it hard to exert translational and rotational loads to conduct a specific ACL task.

The last point of discussion is based on the resolution issues of fMRI. fMRI has a low spatial resolution compared to normal MRI. The fMRI technique used in the discussed study is not very advanced. The 1.5-T imaging technique has a poor spatial resolution, which causes an T2* weighted image, which makes it difficult to distinguish the different specific areas of the brain. On the other hand 3-T imaging has a bettered resolution of 30-60% (theoretically 100%). With a 3-T imaging technique (or more advanced) it would be easier and more valid to conclude whether the measured brain activity is due to activity in the somatosensory cortices or due to activity in other non related parts of the brain.


The claim that plastic changes would occur after an ACL rupture seems to be correct. Hypothetically it seems possible to state that, if a structure with contribution to proprioception and brain activity is removed or ruptured, there would be a lower input or no input at all in the related cortical areas. This claim is very generalized and not very helpful to scientific research.

I mentioned four points of discussion, concerning the following aspects:

The null hypothesis

The null hypothesis states that ACL patients do not present a different activation pattern of the brain compared with healthy controls.

The hypothesis lacks on focus and does not determine what kind of brain activation pattern is expected. Moreover, the hypothesis does not determine whether the scientists aim at activation patterns in ACL proprioception related areas of the brain or areas of the brain which are related to somatosensory tasks.

The performed task

The task performed seems to be non-specific to ACL involved motion, because the ACL is involved in anterio tibial translational and rotational loads. The task performed during the experiment lacks translational and rotational loads. In addition, the task is quite simple. In case of a more advanced task, the measurement could reveal a more realistic image of the adaption pattern or plasticity within the brain.

The shortcoming of fMRI to measure ACL specific tasks

fMRI is very sensitive to head motion. Simple more specific tasks involving the ACL cannot be measured by fMRI, because of the range of motion needed.

The spatial resolution of fMRI

The spatial resolution depends on the strength of the magnetic field. The 1.5-T imaging technique used in this study is not advanced enough. The lack of spatial resolution could lead to wrong measurements or misinterpretations. A 3-T imaging technique leads to an 30-60% improvement of the spatial resolution, that makes is possible to distinguish smaller regional differences.

In general, the study could be improved by performing more specific ACL tasks. fMRI could be a problematic technique to cope with more dynamic specific ACL tasks.

Since the study about ACL brain plasticity measurements is still in its starting fase, the study could be seen as a first introduction to the field of brain plasticity caused by lower limb deficiency.

Using a higher magnetic field strength would be recommend for a better spatial resolution and to distinguish smaller area activations.