How to Read an MRI of the Spine

Reading a spine MRI starts with understanding what the different shades of gray actually mean and then working through the images in a consistent order so you don’t miss anything. A spine MRI typically includes at least two types of images, T1-weighted and T2-weighted, each highlighting different tissues in different ways. Radiologists follow a systematic approach, checking alignment, disc health, the spinal canal, nerve openings, and the bones themselves on every scan. Even without years of training, knowing the basics of what each sequence shows and what common findings look like can help you make sense of your own report.

What T1 and T2 Images Show

Every spine MRI includes at least two main “weightings,” and the difference between them is essentially what appears bright versus dark. On T1-weighted images, fat is bright and water is dark. On T2-weighted images, water is bright and fat is moderately bright. This matters because the spinal canal is filled with cerebrospinal fluid, which is mostly water. On T2 images, that fluid lights up white, making it easy to see anything encroaching on the canal. On T1 images, the fluid is dark, which can make the bones and soft tissues stand out more clearly against it.

An early comparison of the two weightings in lumbar spine imaging found no major difference between T1 and T2 for spotting disc protrusions or bone spurs. T2 was better at showing disc desiccation (the drying out of a disc), but the clinical significance of that finding alone was uncertain. Certain conditions like metastatic disease and arachnoiditis were actually better seen on T1 images, in part because the bright cerebrospinal fluid on T2 sometimes obscured lesions sitting within the spinal canal.¹ The practical takeaway is that neither weighting is universally “better.” You need both to get the full picture.

Some scans also include a STIR sequence, which is a fat-suppressed technique that makes fluid and inflammation very bright while turning fat signal dark. STIR can be more sensitive than contrast-enhanced T1 for detecting inflammatory lesions. In a study of ankylosing spondylitis, STIR picked up inflammation in about 31% of vertebral units compared with roughly 27% on contrast-enhanced T1 imaging.² If your MRI report mentions STIR or “fat-suppressed” images, those are the sequences being used to hunt for swelling, fluid, or active inflammation.

Orientation and the Two Main Views

Spine MRIs are displayed in at least two planes. Sagittal images slice the body from side to side, giving you a profile view of the spine, like looking at someone from the side. You can see the vertebral bodies stacked up, the discs in between, and the spinal cord running through the canal behind them. This is usually the first set of images a radiologist scrolls through because it gives the big picture: overall alignment, disc heights, and any obvious compression.

Axial images slice the body horizontally, as if looking down through the top of someone’s head. Each axial slice typically cuts through one disc level or one vertebral body. These are essential for seeing how much space the spinal cord or nerve roots actually have. You can see the disc’s outline, the facet joints in the back, the ligaments, and whether anything is pushing into the canal or the nerve exit tunnels (foramina). Some MRIs also include coronal images (slicing from front to back), though these are less standard for routine spine work.

What Normal Looks Like

Knowing what healthy structures look like on MRI makes it much easier to spot when something is off. On T2 images, a healthy intervertebral disc appears bright white in its center because the nucleus pulposus (the gel-like core) is rich in water. The outer ring of the disc, the annulus fibrosus, is darker. When discs lose water content with age or degeneration, they darken on T2, a finding often described as “disc desiccation” or “loss of T2 signal.” Studies of lumbar MRIs that were read as normal have shown that the disc spaces should be relatively uniform in height and signal, without changes in the adjacent bone marrow.³

The spinal cord itself appears as a smooth gray structure on T2 images, surrounded by bright cerebrospinal fluid. It tapers to an end at the conus medullaris, which typically sits at the T12-L1 vertebral level. Below that, the cauda equina, a bundle of 20 nerve roots, fans out within the fluid-filled thecal sac and continues down into the sacrum.⁴ If the conus sits lower than expected, that can raise suspicion for a tethered cord, a condition more commonly evaluated in children.

Vertebral bodies should have fairly uniform signal. On T1, the bone marrow in adults is a mix of fatty (bright) and hematopoietic (slightly darker) marrow. A vertebral body that looks dramatically darker than its neighbors on T1, or that lights up abnormally on STIR, is a red flag worth investigating.

Disc Problems

Disc herniations are among the most common findings on a spine MRI, and the report usually classifies them by how far the disc material has moved. A bulge is a broad-based outpouching that extends beyond the edges of the vertebral body but doesn’t form a focal point. A protrusion is a more focused extension where the base of the herniation against the parent disc is wider than the part sticking out. An extrusion means the herniated piece has a narrower base than its leading edge, essentially mushrooming outward. A sequestration is the most extreme form, where a fragment has broken free from the parent disc entirely.

These distinctions matter for treatment planning. On MRI, the signal intensity of a herniated fragment compared to the remaining disc can give clues about how old the herniation is. Research has found that for extrusions, the signal intensity of the herniated fragment relative to the nucleus pulposus correlates strongly with the time since symptom onset, with more acute herniations tending to look more similar in signal to the parent disc and older ones becoming darker.⁵ This can be clinically useful because many acute herniations shrink on their own over months, while chronic, darker fragments are less likely to reabsorb.

Modic Changes and Bone Marrow Signals

If your MRI report mentions “Modic changes,” it’s describing signal changes in the vertebral bone marrow right next to a disc. These are classified into three types based on what the marrow signal looks like. Type 1 shows marrow edema and inflammation: dark on T1, bright on T2. Type 2 shows fatty replacement: bright on T1, bright or intermediate on T2. Type 3 shows sclerosis (dense bone): dark on both T1 and T2. These changes are closely associated with degenerative disc disease and are thought to reflect a spectrum where inflammation (Type 1) can eventually convert to fatty change (Type 2) over time.⁶

The relationship between Modic changes and actual pain is a subject that researchers have gone back and forth on. Modic Type 1 changes have the strongest association with low back pain, but the sensitivity of detecting them varies depending on the MRI machine’s field strength, the pulse sequences used, and the experience of the reader. A consensus statement from the International Society for the Study of the Lumbar Spine noted that the high specificity but low and variable sensitivity of Modic classification contributes to inconsistency across studies.⁷ In other words, if Modic Type 1 is present at a level, there’s a reasonable chance that level is a pain source, but the absence of Modic changes doesn’t rule pain out.

Spinal Stenosis on MRI

Spinal stenosis means the space available for the spinal cord or nerve roots has narrowed. This is one of the most clinically relevant findings on a spine MRI, and radiologists grade it to communicate the severity. Several grading systems exist, but the general approach is similar across the lumbar and cervical spine.

For central canal stenosis in the cervical spine, one widely used classification grades the narrowing on T2 sagittal images: grade 0 means no stenosis, grade 1 means the cerebrospinal fluid space is more than half obliterated, grade 2 means the spinal cord itself is deformed by the compression, and grade 3 means the spinal cord shows signal change within it (a sign of potential damage to cord tissue).⁸ A refinement of that system subdivides grade 2 into cases where the cord is deformed only on one side versus both sides, since the latter is significantly more associated with motor dysfunction.⁹

Looking at the cervical spine on your MRI, the key thing to look for on T2 sagittal images is whether the bright cerebrospinal fluid around the cord is intact or squeezed out. If the cord itself appears flattened or deformed, that’s more significant than just fluid-space narrowing. And if there’s a bright signal within the cord substance on T2, that suggests the cord tissue itself may be injured, a finding called myelopathy signal.

Neural Foraminal Stenosis

The neural foramina are the bony tunnels on each side of the spine where nerve roots exit. These are best evaluated on sagittal T1 images in the lumbar spine, because the fat surrounding the nerve root appears bright on T1, and you can see when it’s been squeezed away. A practical grading system classifies foraminal stenosis into four grades: grade 0 (normal, with fat visible around the nerve), grade 1 (mild, with fat loss in two directions), grade 2 (moderate, with fat loss in all four directions but the nerve still looks normal in shape), and grade 3 (severe, where the nerve root is visibly compressed or collapsed).¹⁰

The causes of foraminal narrowing differ by spinal level. At L4/L5, loss of disc height is the dominant contributor. At L5/S1, spondylolisthesis (forward slippage of one vertebra on another) and facet joint enlargement play bigger roles. In one study, disc height loss made foraminal stenosis at L4/L5 about four times more likely, while at L5/S1, facet hypertrophy was associated with more than a sixfold increase in odds of foraminal stenosis.¹¹

The grade of foraminal stenosis matters for predicting symptoms. Research using high-resolution 7 Tesla MRI in the cervical spine has shown that even slight-to-moderate nerve compression was associated with over a 22-fold increase in the likelihood of sensory or motor symptoms at that level, and severe compression pushed that to roughly a 48-fold increase.¹² Grading the foramen alone was quite good at predicting symptoms, but combining the foraminal grade with direct assessment of whether the nerve root was visibly compressed improved accuracy further.

Red Flags on a Spine MRI

Most spine MRIs show degenerative findings, but certain patterns should raise alarm for infection, tumor, or other serious conditions.

Spinal infection (vertebral osteomyelitis and discitis) has a fairly recognizable MRI appearance. The classic pattern involves dark signal in two adjacent vertebral bodies on T1, bright signal in the disc between them on T2, and destruction of the endplates that normally form a clean, sharp border between bone and disc. A review of osteomyelitis cases found that about 95% showed decreased vertebral body signal on T1, 95% had loss of endplate definition, and 95% had increased disc signal on T2.¹³ With contrast, the disc and bone enhance avidly, and if there’s an abscess in the surrounding soft tissues, it tends to show ring-like enhancement rather than solid enhancement (which suggests phlegmon, an earlier stage of infection).

The tricky part is that early infection can mimic degenerative Modic Type 1 changes, since both involve marrow edema near the endplates. In a study of early vertebral osteomyelitis, four cases initially looked like routine degenerative endplate change on MRI, only to progress to obvious infection on follow-up.¹⁴ If the clinical picture includes fever, rising inflammatory markers, or recent bacteremia, even subtle endplate edema deserves close follow-up.

For tumors, the worry is usually metastatic disease. Most spinal metastases replace normal fatty marrow with tumor tissue, making the affected vertebral body dark on T1 (because the bright fat signal is gone) and variably bright on T2 and STIR. Hemangiomas, which are common benign vascular lesions in vertebral bodies, can sometimes look similar. Typical hemangiomas are easy to identify because they are very bright on both T1 and T2 (their fat and slow-flowing blood give them a distinctive signal). Atypical hemangiomas, however, can look darker on T1 and mimic metastases. Comparing T1 images with and without fat suppression can distinguish them with high accuracy, since hemangiomas lose more signal when fat is suppressed than metastases do.¹⁵ Diffusion-weighted imaging can also help, as malignant lesions tend to restrict diffusion (appearing bright on DWI with corresponding low values on ADC maps) more than hemangiomas do.¹⁶

The Post-Surgical Spine and Contrast Enhancement

If you’ve had prior spine surgery, reading the MRI gets harder. Scar tissue (epidural fibrosis) and recurrent disc herniation can both fill the same space near the surgical site, and both can compress a nerve root. On standard non-contrast images, they can look frustratingly similar.

This is the main reason contrast (gadolinium) is used in post-surgical spine MRIs. Scar tissue has blood vessels running through it and enhances brightly and relatively uniformly within about ten minutes of receiving contrast. A recurrent disc herniation, being avascular, does not enhance in that early window. On delayed images (20-30 minutes later), the disc fragment may start to show some peripheral enhancement as contrast seeps in from surrounding tissue, but the early-versus-late pattern is what distinguishes the two.¹⁷ This technique has been validated as highly accurate for separating scar from disc.¹⁸

Why Many Findings Don’t Match Symptoms

This is arguably the most important thing to understand about spine MRI: an abnormal-looking scan does not necessarily mean you have a problem. A systematic review of imaging findings in people with no back pain at all found that degenerative features, including disc bulges, disc degeneration, and facet arthropathy, were present in high proportions of asymptomatic individuals, and the prevalence increased steadily with age.¹⁹ By middle age, the majority of people walking around with no symptoms at all would have at least one “abnormal” finding on an MRI.

Even spinal cord compression can exist without symptoms. A meta-analysis pooling data from thousands of healthy individuals estimated that about 24% of asymptomatic people have some degree of spinal cord compression visible on MRI, with older individuals and those of American or European descent showing higher rates. Only about 2% of the healthy population had evidence of actual degenerative cervical myelopathy.²⁰ This underscores a point that radiologists and spine surgeons emphasize constantly: MRI findings must be interpreted alongside the patient’s clinical picture. A disc bulge at L4/L5 only matters if it’s compressing something in a way that explains your specific symptoms.

Metal Hardware and Image Artifacts

If you have spinal hardware from a prior fusion, screws and rods can create dark voids and bright flares on the MRI known as susceptibility artifacts. These obscure the tissue immediately around the hardware, which is often exactly the area your doctor needs to see. The severity of these artifacts depends on the type of metal (titanium produces less artifact than stainless steel), the MRI sequences used, and the magnetic field strength of the scanner.

Lower-field MRI machines produce less metal artifact. A comparison of 0.55 Tesla and 1.5 Tesla MRI in patients with posterior spinal fusion hardware found that the lower-field scanner produced milder susceptibility artifacts while still providing equally good assessment of the spinal canal and neural foramina. The tradeoff was slightly lower overall image resolution and contrast at the lower field strength.²¹ Metal artifact reduction sequences (sometimes abbreviated MARS or MAVRIC) are another option radiologists can deploy when hardware is present.

Advanced Sequences Worth Knowing About

Beyond the standard T1, T2, and STIR, some spine MRIs include specialized sequences that answer specific clinical questions.

Diffusion-weighted imaging (DWI) measures how freely water molecules move through tissue. When cells are packed tightly together, as in a tumor or in tissue damaged by a sudden loss of blood supply, water movement is restricted. DWI has shown promise for detecting spinal cord infarction (stroke of the spinal cord), which can be very difficult to see on conventional MRI in the early hours. Studies have found that DWI can pick up spinal infarction as a bright signal with corresponding low values on ADC maps. Characteristic patterns include an “owl’s eyes” sign (bilateral ventral horn involvement) and a “pencil-like” hyperintensity on DWI, both of which help distinguish infarction from inflammatory conditions like neuromyelitis optica spectrum disorder, where DWI signal behaves differently on ADC maps.²² While the total body of evidence is still based on small patient numbers, DWI is increasingly recognized as a useful and feasible technique for suspected spinal infarction.²³

MR spectroscopy is another emerging tool, particularly for evaluating vertebral lesions. It can measure the chemical composition of tissue within a vertebral body, and the ratio of lipid to water content helps distinguish benign hemangiomas from metastatic deposits.²⁴ This technique remains mainly a research and problem-solving tool rather than part of routine imaging.

Pediatric Spine MRI and Tethered Cord

In children, spine MRI serves some different purposes than in adults. The most common reason for pediatric spine MRI is evaluating for tethered spinal cord, a condition where the end of the spinal cord is abnormally fixed in position, often associated with occult spinal dysraphism (hidden spinal defects under intact skin). MRI has moderate strength of evidence for diagnosing tethered cord, with medium to high sensitivity and specificity.²⁵

The key finding is a low-lying conus medullaris. As noted earlier, the spinal cord normally ends around T12-L1. In tethered cord, it may extend below L2 or even into L3. One case report documented a borderline low-lying cord at the mid-L3 level identified on both ultrasound and MRI.²⁶ Alongside the low conus, radiologists look for a thickened filum terminale (the thin strand that anchors the cord), fatty infiltration of the filum, and associated anomalies like lipomas or split cord malformations. Since the availability of MRI has increased, more cases of tethered cord are being identified, though variability in evaluation and management remains an ongoing concern.²⁷

A Practical Reading Order

If you’re trying to make sense of your own spine MRI images rather than just reading the radiologist’s report, having a consistent approach helps. Radiology literature recommends a systematic method so that nothing gets overlooked.²⁸ Here’s a practical order that applies whether you’re looking at cervical, thoracic, or lumbar images:

Alignment: Start with the T2 sagittal. Are the vertebral bodies lined up, or is one slipped forward (spondylolisthesis) or the curve abnormal?
Vertebral bodies: Scan each body for uniform signal. Any dark body on T1 or abnormally bright body on STIR stands out and needs explanation.
Discs: On T2, check each disc’s brightness. Dark discs are degenerated. Look at the disc contour on both sagittal and axial images for herniations.
Spinal canal: On T2 axial images, check whether the bright fluid around the cord or cauda equina is present or obliterated. On sagittal images, look for areas where the bright fluid column is pinched.
Foramina: On T1 sagittal images, check the bright fat around each exiting nerve root. If the fat is gone, the nerve may be compressed.
Posterior elements: Check the facet joints for hypertrophy and the ligamentum flavum for thickening, both visible on axial images.
Soft tissues: Glance at the paraspinal muscles and any pre-vertebral soft tissue. Abscesses, masses, or unusual fluid collections can hide here.

A systematic approach prevents the common trap of anchoring on one dramatic finding and missing a second, potentially more important one.²⁹ Even if you’re just trying to understand the radiology report you’ve been handed, knowing this sequence helps you follow the logic of how findings were identified and why certain things are mentioned before others.

1
Elsevier. Comparison of T1 and T2 weighted images of the lumbar spine
2
Annals of the Rheumatic Diseases. Inflammation in ankylosing spondylitis: a systematic description of the extent and frequency of acute spinal changes using magnetic resonance imaging
3
PubMed Central. Measurement of the Normal Lumbar Intervertebral Disc Space Using Magnetic Resonance Imaging
4
Contemporary Diagnostic Radiology. MRI of Conus Medullaris, Cauda Equina, and Filum Terminale Lesions
5
PubMed Central. Signal Intensity of Lumbar Disc Herniations: Correlation With Age of Herniation for Extrusion, Protrusion, and Sequestration
6
PubMed Central. Modic changes – An evidence-based, narrative review on its patho-physiology, clinical significance and role in chronic low back pain
7
PubMed Central. Measuring and reporting of vertebral endplate bone marrow lesions as seen on MRI (Modic changes): recommendations from the ISSLS Degenerative Spinal Phenotypes Group
8
PubMed. New MRI grading system for the cervical canal stenosis
9
Egyptian Journal of Radiology and Nuclear Medicine. Reliability and clinical validity of the Kang MRI grading system for cervical central spinal stenosis
10
PubMed. A practical MRI grading system for lumbar foraminal stenosis
11
PubMed Central. The Relationship between Neural Foraminal Stenosis and Imaging Features of Lumbar Spine MRI in Patients Older Than 60 Years with Lumbar Radiculopathy
12
Investigative Radiology. 7 T MRI of the Cervical Neuroforamen: Assessment of Nerve Root Compression and Dorsal Root Ganglia in Patients With Radiculopathy
13
PubMed. MR imaging of vertebral osteomyelitis revisited
14
PubMed. The MRI appearances of early vertebral osteomyelitis and discitis
15
PubMed Central. Differential diagnosis of hemangiomas from spinal osteolytic metastases using 3.0 T MRI: comparison of T1-weighted imaging, chemical-shift imaging, diffusion-weighted and contrast-enhanced imaging
16
Polish Journal of Radiology. Accuracy of diffusion-weighted imaging in discriminating atypical vertebral haemangiomas from malignant masses in patients with vertebral lesions: a cross-sectional study
17
PubMed. Lumbar spine: postoperative MR imaging with Gd-DTPA
18
PubMed. Gadolinium-enhancement characteristics of magnetic resonance imaging in distinguishing herniated intervertebral disc versus scar in dogs
19
PubMed Central. Systematic literature review of imaging features of spinal degeneration in asymptomatic populations
20
PubMed Central. The Prevalence of Asymptomatic and Symptomatic Spinal Cord Compression on Magnetic Resonance Imaging: A Systematic Review and Meta-analysis
21
Academic Radiology. Comparison of Image Quality and Metal Artifact Severity at 0.55 T and 1.5 T MRI in Patients with Spinal Implants Following Posterior Fusion Surgery
22
PubMed Central. The utility of diffusion-weighted imaging in patients with spinal cord infarction: difference from the findings of neuromyelitis optica spectrum disorder
23
PubMed. Diffusion-weighted MR imaging (DWI) in spinal cord ischemia
24
PubMed. MR spectroscopy in differentiating spinal atypical hemangioma from metastasis: preliminary results
25
Pediatrics. Diagnosis and Treatment of Tethered Spinal Cord: A Systematic Review
26
PubMed Central. Tethered Cord Syndrome: Role of Imaging Findings in Surgical Decision-Making
27
PubMed. Evaluation and management of tethered cord syndrome in occult spinal dysraphism: Recommendations from the international children’s continence society
28
PubMed. Systematic approach to interpretation of the lumbar spine MR imaging examination
29
PubMed. A practical approach to spine imaging