Principles of MRI


Ganesh Athappan MD Sumit Karia MD

The Common Vein Copyright 2010


Nuclear magnetic resonance (NMR) is a tool that is used to obtain information regarding the structure and its chemical components. Its principle lays upon the principle in which a nucleus of an atom absorbs electromagnetic radiation at a specific frequency distinct from other nucleus, in the presence of a strong magnetic field. This is valid for particles whose nucleus’ spin is not zero.

The process of obtaining an MRI involves

 Placing a patient in a magnetic field – M

Applying a radiofrequency pulse – flipping of protons by absorbtion of high energy photons (resonance) – R

Termination of the RF pulse to allow relaxation and image acquisition –

Once the RF transmitter is turned off three things begin to happen simultaneously.

1. The absorbed RF energy is retransmitted (at the resonance frequency).

2. The excited spins begin to return to the original orientation. (T1 recovery- longitudinal relaxation ).

3. Initially in phase, the excited protons begin to dephase (T2- transverse relaxation ).

The signal intensity/ contrast on the MR image is determined by four basic parameters: 1) proton density, 2) T1 relaxation time, 3) T2 relaxation time, and 4) flow.

Proton density is the concentration of protons in the tissue in the form of water and macromolecules (proteins, fat, etc).

The most common effect of flow is loss of signal from rapidly flowing arterial blood.

T1- T1 is a characteristic of tissue and is defined as the time that it takes the longitudinal magnetization to grow back to 63% of its final value.

Different tissues have different rates of T1 relaxation. If an image is obtained at a time when the relaxation curves are widely separated, T1-weighted contrast will be obtained.

For example: White matter has a very short T1 time and relaxes rapidly. Cerebrospinal fluid (CSF) has a long T1 and relaxes slowly. Gray matter has an intermediate T1 and relaxes at an intermediate rate

T1 weighted

If we create an image at a time when these curves were widely separated, we would produce an image that has high contrast between these tissues- is termed T1-weighted contrast. Thus, white matter contributes to the lighter pixels, CSF contributes to the darker pixels, and gray matter contributes to pixels with intermediate shades of gray.

T2 weighted

T2 is a characteristic of tissue and is defined as the time that it takes the transverse magnetization to decrease to 37% of its starting value.

Example : Different tissues have different values of T2 and dephase at different rates. White matter has a short T2 and dephases rapidly. CSF has a long T2 and dephases slowly. Gray matter has an intermediate T2 and dephases intermediately

If we create an image at a time when the transverse magnetization curves were widely separated, then we would have high contrast between the tissues in our image. We would see that CSF is associated with lighter pixels, white matter is associated with darker pixels, and gray matter is associated with intermediate gray-level pixels.

T2 weighted

 The spin of the nucleus of an atom or nuclear spin depends on the number of protons and neutrons. Atoms whose nuclei are composed of a number pair of protons and neutrons have spin zero, as, for example, carbon-12 (6 protons and 6 neutrons) and oxygen-16 (8 protons and 8 neutrons), which are widespread in the body. Hydrogen, for instance, has only 1 proton, and thus the nuclear magnetic moment of one isolated proton, that can be detected using NMR techniques.

 NMR consists in changing the nuclear magnetic moment, i.e., to get the nucleus of an energy level to another, applying magnetic fields to the sample we wish to study. A high-frequency alternating field in the radio frequency range is used to change the magnetization direction of the static field to deflect them (tilt). After switching off the high-frequency alternating field the spins go back to the static magnetic field (relaxation). This relaxation has a characteristic decay time for various elements and therefore for different tissue types. The relaxation involves the emission of an electromagnetic wave that can be detected by a sensor.

In the absence of external magnetic field, protons are randomly oriented, that is, when the sum of elementary vectors of magnetization (for each proton) is zero and there is no macroscopic magnetization vector (M = 0). When the external magnetic field is applied, the protons will moving in its direction, becoming parallel to the vector M.

M z corresponds to the transition between proton energy levels. M xy corresponds to the spin phasing and dephasing. The time elapsed during longitudinal relaxation (M z) is denominated T1. The time for disappearance of the transverse magnetizations is named T2. Hence, T1 and T2 are characteristics of the response of the proton to the external magnetic field. In practical terms, T1 is a measure of how fast a tissue becomes magnetized, whereas T2 measures how fast the tissue loses its magnetization.

The T1 for pure low-viscosity fluids, like water, is within few seconds; the same applies to gels and porous materials. Conversely, structured solids have long relaxation times, sometimes in the range of hours. Typical values for T1 in human tissues range between a few seconds for body fluids such as blood or cerebrospinal fluid (CSF) and 100 ms for body fat. The latter will thus appear brighter (i.e. more intense) in T1WI’s compared to the former.

Spin-echo pulse sequences yield T1 weighted images, T2 weighted images, and proton density weighted images. T1WIs provide incredible anatomic detail, in which fat, blood and proteinaceous fluids are identifiable. T2WIs are very sensitive to detect edema (water) and presence of potentially pathologic lesions. PDWIs, which convey information about the density of hydrogen in the tissue, are most useful in brain imaging. Spin-echo pulses are adjusted to yield these different images, by selecting the time between radiofrequency pulses (time of repetition), and time allowed for generation and detection of photons secondary to relaxation (time of echo). T1WI’s have both low TR and low TE, whereas T2WI’s have long TR and TE. A long TR with low TE yields PDWI’s.

T1 relaxation times can be emphasized, using a technique called inversion recovery. An extra setting, selectable, is added to the pulse sequences, called time of inversion. This is the time that occurs between the radiofrequency pulse that hits the proton at 90° and the pulse that hits it at 180°. When protons transit from one energy level to another, they do not all do it at the same time; there is a difference, in which some are faster than others, also element specific. By using radiofrequency pulses at these angles separated by a given time, it is possible to interact with the nucleus and show the difference between the slow and fast moving protons. Tissues with short T1 times yield a brighter signal. Short T1 inversion recovery (STIR) sequences thus increase lesion detail. Fat has a short T1. Hence by choosing a short TI of 140 milliseconds, the fat signal can be suppressed. STIR should not be used to suppress fat signal when using gadolinium contrast, because gadolinium–enhanced tissues short T1 and may be erased.

Fluid attenuated inversion-recovery (FLAIR) sequences are used to suppress liquid signals by adjusting the TI. Water has a long T1, and its signal can be erased at TI of 200 milliseconds.

Fast spin echo sequences reduce image (slice) acquisition time. This is done by using the time after the first echo to apply 180° pulses to obtain a different echo, called spin echo train. Basically, more echoes are received using the same time of repetition.

In a somewhat similar process, in multi-echo spin echo sequences, after the spin echo is received, a 180° pulse is  applied. The difference is that the echo train received thereafter has the same phase encoding, and is used to build a second image. Since the echo time differs between the 2 images, the second image will be more T2 weighted than the first. This will also reduce the number of repetitions (TR), thus reducing acquisition time. Hence, PD- and T2-weighted images can be obtained simultaneously.

Fast repetition of 180° pulses changes T2 relaxation times however, by perturbing a phenomenon called J coupling, which refers to a coupling of nuclei occurring with the spins, causing T2 lengthening. Thus, fat has a higher T2 signal in fast spin echo than in standard spin echo.

The risk of artifacts is increased.

 Again using the spin echo concept, ultrafast spin echo sequence is an even faster technique. There is no repetition of pulses; single 90° pulses are applied, throughout the sample desired to image (as much needed to obtain enough information to obtain a slide). That pulse will generate image for a segment of the image and the 180° pulse is applied to obtain T2 and PD information. Time is greatly reduced, but increases noise and blurring. Time will be dependent on the number of image encodings required. They provide better results in non-moving structures, appearing as a T2-weighted hypersignal. They are commonly used for cholangio-MRI and uro-MRI.

 Gradient echo sequences are those obtained by stimulating the object of study with a radiofrequency pulse used a flip angle below 90°, used in spin echo sequences. This will decrease the amount of magnetization in the transverse plane, and therefore a faster recovery in the z plane, which allows shorter TR’s and TE’s to be used, therefore decreasing scan time. Because of their speed, they are useful to minimize motion artifact that occurs during breathing, cardiac cycle, pulsations. Moving particles will have a bright signal. However, T2-weighted images will be more prone to artifacts. This occurs because the signal obtained is actually more T2*-weighted rather than T2-weighted, in other words, T2 is affected by static field inhomogeneities (and becomes called T2*). T2* is much shorter than T2.

Echo planar

Many categories of conditions can be made; the following represent circumstances in which such studies are deemed necessary for an accurate and final diagnosis:

 Head trauma, there being:

Change in mental status

Signs of increased intracranial pressure.

Loss of consciousness

New neurological deficits.


Worsening headaches

Skull fracture.

Suspected intracranial hemorrhage

CT or MRI are in these cases medically necessary. The former is preferred when there is suspicion for hemorrhage, acute subdural/epidural hematoma and fracture. MRI is preferred for suspected shearing lesions and diffuse axonal injury in closed head trauma, as well as for evaluation of subacute and chronic sequelae of head injuries.

Cerebrovascular accident, where there can be changes in motor, sensory, visual, behavioral and cognitive functions.

CT or MRI are necessary.

When symptoms develop in less than 90 minutes, there is a possibility to intervene with thrombolysis or IR guided thrombectomy. In this situation, there is necessity to exclude hemorrhage or mass, which is better evaluated with CT is indicated (hyperacute setting).

MRI is indicated to detect subtle mass effects that may be associated with acute hemorrhagic infarction. It can also give information about the timing of the hemorrhage, that is, if it is hyperacute, acute, or chronic. It is very useful in the evaluation of infarcts in the brain stem and deep white matter infarcts / lacunar infarcts, since CT cannot give as much detail in these regions, yielding many times false negatives.

Inflammation and infection, for instance, abscess formation, encephalitis.

CT or MRI are the modalities indicated, MRI being preferred for evaluating bacterial, fungal and parasitic abscesses (i.e., cysticercosis), ependymitis and subdural empyemas.


The differential diagnosis includes hemorrhage, mass lesions. These are of more suspicion when headaches are accompanied by symptoms of increased intracranial pressure or suggest intracranial bleeding.

Again, CT or MRI are indicated if this headache has given characteristics:

Sudden very severe headache

With suspicion of intracranial lesion

Exacerbation of chronic headache with neurologic signs, such as motor or mental status changes

Chronic headache with increase in frequency or severity

If associated with fever and meningeal symptoms

If there is history of cancer or AIDS

If they awake patients from sleep

If they are exertional

Malignant and benign lesions

Both CT and MRI are necessary for the evaluation before and after therapy of malignant lesions treated with surgery, chemotherapy, or radiation.

MRI is preferred for evaluating the posterior fossa and primary intraaxial or extraaxial tumors (gliobastoma, astrocytoma, lymphoma, acoustic neuromas, menigiomas, cholesteotomas, etc.) and pituitary adenomas.

Demyelinating diseases

When there is a suspicion for multiple sclerosis or acute disseminated encephalomyelitis MRI is considered necessary for its diagnosis. Central nervous system demyelination will be observed and this is seen on MRI. Leukodystrophies, central pontine myelinolysis, progressive multifocal leukoencephalopathy are examples of other demyelinating diseases, less frequent, that can be diagnosed by observation of CNS demyelination in MRI, in conjunction with clinical findings.

Structural abnormalities

Due to the incredible detail MRI provides, it is the exam of choice in these evaluations:

Chiari I & II malformations. This set of malformations are part of a group of congenital abnormalities that involve components of the cranio-cervical junction (hindbrain abnormalities).

Congenital lesions such as Dandy-Walker malformations, craniosynostosis, macrocephaly, microcephaly)

Hydrocephalus – for the initial evaluation hydrocephalus.  CT is chosen when goal is to follow degree of hydrocephalus in shunted patient

Vascular abnormalities (e.g., aneurysm, arteriovenous malformations, venous capillary and cavernous angiomas, venous thrombosis) – MRI, more specifically MRA

Non-Traumatic and Non-CVA related Hemorrhage

In suspicion of hypertensive hemorrhage.

Spontaneous hemorrhage in patients at risk for bleeding, such as those receiving anticoagulation therapy.

Other CNS indications (not absolute):

Suspicion of TIA / stroke

New onset of seizure activity


Visual field loss, diplopia or other visual changes that remain unexplained after ophthalmologic evaluation


Focal neurologic deficit unexplained by clinical history

Change in mental status

Suspicion of degenerative neurologic process (e.g. ataxia, dyskinesia)

Suspicion of hypothalamic / pituitary lesion.

Parkinson’s disease

Persistent tinnitus

Further evaluation of changes detected in other radiologic exams, if clinically warranted (e.g. for therapeutic decision)

Symptoms of increased intracranial pressure, for decision of safety of performance of lumbar puncture (risk of herniation).

For the following, MRI is the recommended study and is considered necessary, in the presence of progressive asymmetrical hearing loss associated with:

Abnormal neurological evaluation

Vertigo, syncope and dizziness

Abnormal electronystagmography, audiometry or auditory brainstem response

When symptoms persist and evaluation other etiologies have been ruled out

Neurosurgical procedures:

Pre-operative evaluation prior to neurosurgery involving craniotomy, craniofacial surgery, cranial nerve procedures or a biopsy of the central nervous system.

Evaluation of hydrocephalus.

Evaluation of shunt for hydrocephalus or its revision.

Control Center