On an MRI scan, brain atrophy appears as a combination of widened sulci (the grooves running across the brain’s surface), enlarged ventricles (the fluid-filled chambers within the brain), and an overall reduction in brain tissue volume that causes the brain to visibly pull away from the inner surface of the skull. These changes show up because as brain cells die and tissue shrinks, cerebrospinal fluid rushes in to fill the vacated space — and on MRI, that fluid appears dark or black on T1-weighted sequences, making the loss of tissue starkly visible. A radiologist looking at an MRI showing moderate atrophy might describe hippocampi that look flattened and smaller than expected, ventricles that have expanded noticeably, and sulci that are far wider than they should be for a person of a given age.
What the scan actually shows — and what it means — depends heavily on the pattern, the location, and the degree of the changes. Atrophy concentrated in the medial temporal lobes (where the hippocampus sits) looks very different from atrophy that is spread across the entire cortex, or atrophy that preferentially affects the frontal and temporal lobes. This article covers how radiologists visually identify and grade brain atrophy on MRI, what regional patterns are associated with specific conditions like Alzheimer’s disease and frontotemporal dementia, what the standardized rating scales measure, and what it means when a report notes atrophy that is “age-appropriate” versus “greater than expected.”.
Table of Contents
- What Does Brain Atrophy Actually Look Like on an MRI Scan?
- How Radiologists Grade and Measure Brain Atrophy on MRI
- Regional Atrophy Patterns and What They Suggest
- What the Hippocampus Looks Like When Atrophy Is Present
- Normal Aging Versus Pathological Atrophy on MRI — A Critical Distinction
- How Atrophy Reports Are Standardized — and Why It Matters
- Where Brain Atrophy Research Is Heading
- Conclusion
- Frequently Asked Questions
What Does Brain Atrophy Actually Look Like on an MRI Scan?
The most immediate visual clue is the sulci — those deep folds and fissures that run across the surface of the brain. In a healthy young brain, sulci are narrow channels with thick, robust ridges (gyri) on either side. In a brain showing significant atrophy, those sulci look wide and open, almost cavernous, because the gyri have thinned and shrunk. The ridges that once packed tightly together now sit far apart. On a standard MRI, this looks like the brain surface has loosened and deflated slightly inside the skull cavity. Inside the brain, the ventricles tell a similar story. Healthy ventricles are modest, relatively compact spaces.
When surrounding brain tissue shrinks, the ventricles expand passively to fill the void — a process called ex vacuo dilatation. On MRI, this shows up as larger-than-expected dark regions in the center of the brain. In cases of severe atrophy, the ventricles can look dramatically ballooned, and the overall brain appears reduced in bulk. A useful way to think of it: imagine a grape slowly drying into a raisin inside a rigid container. The container (skull) stays the same size, but the contents shrink and pull away from the walls, with dark fluid filling the gaps. The increased cerebrospinal fluid space is also visible as dark areas on T1-weighted MRI. On FLAIR and T2 sequences, white matter hyperintensities — bright spots in the brain’s white matter — often appear alongside atrophy, particularly in cases with a vascular component. These bright spots are not the same as atrophy itself, but their co-occurrence with tissue volume loss is a common finding in cerebrovascular disease and adds to the overall picture of brain change the radiologist is assembling.
How Radiologists Grade and Measure Brain Atrophy on MRI
Because the visual appearance of atrophy can be subtle in early stages and difficult to describe consistently in plain language, radiologists rely on standardized semi-quantitative scales to communicate findings. The most widely used is the MTA scale — Medial Temporal Atrophy — which grades hippocampal and entorhinal cortex shrinkage on a 0 to 4 scale. A score of 0 means no atrophy; a score of 4 reflects severe shrinkage with very little hippocampal tissue remaining visible. The MTA scale specifically assesses the width of the choroid fissure, the width of the temporal horn of the lateral ventricle, and the height of the hippocampal formation. An MTA score of 2 or above in a person under 75 is generally considered abnormal and warrants clinical attention. The GCA scale (Global Cortical Atrophy) measures the overall degree of cortical thinning across the brain’s surface. It runs from 0 (no atrophy) to 3 (knife-blade atrophy, where the gyri appear paper-thin).
The Koedam scale addresses posterior cortical atrophy — shrinkage in the parietal and posterior cingulate regions — which can be an early sign of atypical Alzheimer’s presentations. The Fazekas scale separately grades the severity of white matter hyperintensities, which frequently accompany atrophy in vascular dementia. These four scales are often reported together, giving clinicians a structured profile rather than a narrative impression that can vary from one radiologist to another. However, these scales are assessments, not diagnoses. A single MTA or GCA score does not tell a clinician what is causing the atrophy, only how much tissue appears to have been lost and where. The clinical picture — age, cognitive test scores, rate of change over time, family history — has to be layered on top of the imaging findings before a meaningful interpretation is possible. A GCA score of 2 in a 55-year-old with cognitive symptoms carries different implications than the same score in a person who is 82 and cognitively intact.
Regional Atrophy Patterns and What They Suggest
Where atrophy concentrates on the brain matters enormously, and different diseases tend to leave different footprints on MRI. Alzheimer’s disease characteristically begins with medial temporal atrophy, particularly affecting the hippocampus and entorhinal cortex. On MRI, the hippocampi in early Alzheimer’s appear smaller than normal, sometimes described as flattened, with less distinct borders compared to healthy hippocampi, which are well-defined and prominent on coronal sequences. As the disease progresses, the atrophy spreads to parietal and eventually frontal regions, and global cortical atrophy becomes evident. Frontotemporal dementia (FTD) presents a strikingly different pattern.
Rather than beginning in the medial temporal region, FTD preferentially attacks the frontal and temporal lobes, and the atrophy is often asymmetric — much more pronounced on one side of the brain than the other. In semantic dementia, a variant of FTD, left temporal lobe degeneration can be so lopsided that on an MRI the left temporal lobe looks dramatically smaller than the right, even to an untrained eye. This knife-blade pattern in the anterior temporal lobes is one of the more visually striking findings in clinical neuroimaging. Vascular dementia or cerebrovascular disease tends to produce a different picture again: widespread white matter changes (hyperintensities), often combined with cortical and subcortical atrophy that doesn’t follow the clean lobar pattern of a degenerative disease. A person with vascular contributions might show significant global cortical atrophy alongside extensive Fazekas-grade white matter lesions, reflecting both tissue loss and ischemic damage to the brain’s connecting pathways. Recognizing these distinct spatial signatures is one reason why MRI, rather than CT, is the preferred imaging tool in dementia workup — the contrast resolution and multi-sequence capability of MRI lets radiologists characterize patterns that a CT scan would miss.
What the Hippocampus Looks Like When Atrophy Is Present
The hippocampus is the structure that gets the most attention in dementia-related MRI reporting, and for good reason: it is among the earliest sites of neurodegeneration in Alzheimer’s disease, and its appearance on MRI can shift measurably before a person’s symptoms have progressed beyond mild cognitive impairment. In a healthy brain, the hippocampus is a curved, finger-sized structure sitting in the medial temporal lobe. On coronal MRI slices, it appears as a clearly defined, well-rounded structure with distinct internal architecture. When hippocampal atrophy is developing, the changes are visible as a reduction in size, a flattening of the structure’s profile, and a blurring of its borders. The surrounding temporal horn of the lateral ventricle, which in a healthy brain is a narrow slot, begins to widen as the hippocampus shrinks.
On MRI, this produces a distinctive appearance: a small, flattened hippocampus floating in a relatively large pocket of cerebrospinal fluid. An MTA scale score of 3 or 4 reflects this picture — not much hippocampal tissue left, wide surrounding fluid, and a choroid fissure that has opened up significantly. Automated volumetric software can now measure hippocampal volume precisely and compare it against age- and sex-matched normative databases, giving a percentile score. This approach provides greater precision than a radiologist’s visual estimate alone, but the two methods are complementary rather than interchangeable. Visual assessment catches asymmetry, internal structural changes, and incidental findings that automated tools can miss; volumetric software provides objective numbers over serial scans that may be too subtle for the human eye to reliably track year to year. The tradeoff is time and cost — automated volumetry is still not universally available or reimbursed.
Normal Aging Versus Pathological Atrophy on MRI — A Critical Distinction
One of the most important things to understand about brain atrophy on MRI is that it is not inherently a disease finding. Everyone loses brain volume across adult life, with the rate of loss typically accelerating after age 40. Radiologists are acutely aware of this, and MRI reports on older patients will frequently note “mild cerebral atrophy, age-appropriate” — meaning the degree of tissue loss seen is within the expected range for a person of that age, not a sign of disease. This language can alarm patients who are not prepared for it, but it is routine in reports on adults over 60. The clinically meaningful question is whether the atrophy seen is greater than expected for a patient’s age, and whether it is progressing faster than normal aging would predict.
A report that notes atrophy “greater than expected for stated age” or “out of proportion to age” is flagging that something beyond typical aging may be occurring. This is where clinical context becomes indispensable: a 68-year-old with subjective memory complaints, an MTA score of 3 bilaterally, and rapid cognitive decline on neuropsychological testing presents very differently from a 75-year-old with incidentally found mild hippocampal atrophy and no cognitive symptoms. A warning worth emphasizing: a single MRI is a snapshot, not a trajectory. Seeing atrophy on one scan does not tell you how fast the brain is losing tissue or whether the rate is accelerating. This is why follow-up imaging over 12 to 18 months is often more informative than a single scan when monitoring a person with early cognitive concerns. A 2025 study published in Nature Communications addressed this limitation directly by introducing a method to estimate lifetime brain atrophy from a single scan — calculating the difference between total brain volume and intracranial volume (which reflects the brain’s peak size before shrinkage began) — which may eventually reduce the need for serial scans to establish trajectory.
How Atrophy Reports Are Standardized — and Why It Matters
Until recently, radiological reporting of brain atrophy was inconsistent. Different radiologists used different language, different reference points, and different levels of detail, making it difficult for clinicians to compare reports across institutions or track changes over time. A 2025 study published in PMC compared three approaches to reporting: narrative text reports, semi-quantitative scale-based reporting (MTA, GCA, Koedam, Fazekas), and automated volumetric analysis.
The study found significant variability in narrative reports and made the case for wider adoption of the standardized scales as a baseline expectation. For families and patients navigating a dementia diagnosis or a cognitive concern, this matters practically. If a report from one hospital uses the MTA and GCA scales and a follow-up scan at a different center produces only a narrative description, comparison becomes difficult. Asking a referring physician or neurologist which scales were used, and whether a follow-up scan will use the same methodology, is a reasonable and useful question.
Where Brain Atrophy Research Is Heading
The tools for measuring and interpreting brain atrophy on MRI are improving faster than at any point in the field’s history. Automated volumetric software is becoming more accessible and more precise, machine learning models are being trained to detect subtle atrophy patterns earlier than visual inspection can, and methods like the single-scan lifetime atrophy estimation published in Nature Communications in 2025 may eventually allow earlier detection without requiring years of serial imaging.
Plasma and CSF biomarkers for Alzheimer’s pathology are now often interpreted alongside MRI atrophy data, giving clinicians a more complete picture of what is happening biologically, not just anatomically. For people whose family members are undergoing MRI assessment for cognitive symptoms, the trajectory of the field is toward earlier, more precise, and more standardized measurement. Understanding what the scan is actually showing — widened sulci, enlarged ventricles, flattened hippocampi — is the first step toward being able to have a meaningful conversation with the clinical team about what those findings indicate and what comes next.
Conclusion
Brain atrophy on MRI has a characteristic visual signature: widened sulci, enlarged ventricles, reduced gyral volume, and increased cerebrospinal fluid spaces visible as dark regions on T1-weighted sequences. The specific pattern of atrophy — where it is concentrated, how severe it is, and whether it is symmetric — provides important diagnostic information. Hippocampal atrophy assessed by the MTA scale is the key finding in Alzheimer’s workup; frontotemporal patterns are characteristic of FTD; global cortical atrophy and white matter hyperintensities together point toward vascular contributions.
No single finding is definitively diagnostic on its own, but together these imaging features give neurologists and geriatricians a crucial piece of the clinical picture. For families and patients, the most important takeaway is that atrophy on a report does not automatically mean disease. The clinical question is whether the degree of change is appropriate for the person’s age, and whether it is progressing at an abnormal rate. Asking the reporting radiologist or treating neurologist whether the findings were scored using standardized scales like the MTA, GCA, and Fazekas, and whether follow-up imaging is warranted to assess progression, will give you far more actionable information than the report language alone.
Frequently Asked Questions
Does brain atrophy on an MRI always mean dementia?
No. Some degree of brain atrophy is a normal part of aging, and radiologists routinely note “age-appropriate” atrophy in patients over 60 without any cognitive impairment. The concern arises when atrophy is greater than expected for age, concentrated in regions associated with dementia (particularly the hippocampus), or shown to be progressing faster than normal aging would predict.
What does the hippocampus look like on an MRI when Alzheimer’s is developing?
In early Alzheimer’s disease, the hippocampi typically appear smaller than normal, flattened in profile, and less well-defined compared to healthy hippocampi. The temporal horns of the lateral ventricles, which sit adjacent to the hippocampus, appear wider as the hippocampus shrinks. Radiologists grade this using the MTA scale from 0 (normal) to 4 (severe atrophy).
What do “widened sulci” and “enlarged ventricles” mean on a brain MRI report?
Sulci are the grooves on the brain’s surface; when they appear wider than normal, it reflects shrinkage of the ridges (gyri) between them. Enlarged ventricles mean the fluid-filled chambers inside the brain have expanded, typically because surrounding brain tissue has reduced in volume. Both findings together are the classic visual markers of cerebral atrophy.
What is the MTA scale and what scores are considered abnormal?
The MTA (Medial Temporal Atrophy) scale is a 0-to-4 rating system radiologists use to grade hippocampal atrophy. A score of 0 indicates no atrophy; 4 indicates severe atrophy with very little hippocampal tissue remaining. An MTA score of 2 or above is generally considered abnormal in a person under 75, though clinical context always applies.
Can one MRI scan tell you how fast the brain is shrinking?
Not reliably on its own. A single scan is a snapshot of current volume, not a rate of change. Serial scans over 12 to 18 months are typically needed to assess whether atrophy is progressing faster than normal aging. A 2025 study in Nature Communications proposed a method for estimating lifetime brain atrophy from a single scan by comparing total brain volume against intracranial volume, which may eventually reduce the need for multiple scans.
What is the difference between brain atrophy on CT versus MRI?
CT can show gross atrophy — very widened sulci, clearly enlarged ventricles — but lacks the soft-tissue contrast resolution to detect subtle early changes. MRI is significantly more sensitive to early hippocampal atrophy, white matter changes, and regional patterns that are critical for differentiating between causes. For dementia workup, MRI is the preferred imaging modality when available.
