When an MRI shows a smaller-than-expected hippocampus, how can a radiologist determine whether the finding is clinically significant? The answer involves more than measuring its size. A hippocampal volume Z-score compares the measured volume with a reference population, helping clinicians understand how unusual the result may be. However, age, head size, image quality and segmentation methods can influence this comparison.
The purpose of this guide is to help radiologists interpret hippocampal volume measurements with greater confidence, understand the limitations of normative MRI comparison, and use quantitative findings alongside clinical evidence.
We explain the MRI hippocampal volumetry workflow, how Z-scores differ from percentiles, and what these measurements can reveal about hippocampal atrophy. We also examine why a low Z-score alone cannot diagnose Alzheimer’s disease and how AI-assisted volumetry can support clinical assessment.
A hippocampal volume Z-score shows how far a measured volume differs from the reference mean, expressed in standard deviations.
Think of it as a way to put a measurement into context rather than viewing the number in isolation.
The calculation is:
Z-score = (Observed volume − Reference mean) ÷ Reference standard deviation
For example, a Z-score of −1.5 means the measured volume is 1.5 standard deviations below the reference mean.
A positive score is above the mean, while a negative score is below it. Neither result automatically indicates a problem. The key question is whether the reference data appropriately represents the patient, including relevant factors such as age and intracranial volume.
Two people can have different hippocampal volumes without those differences having the same clinical meaning. Age, head size and natural anatomical variation all play a role.
Normative MRI comparison addresses this problem by showing how a person's measurement compares with an appropriate reference group.
For example, a UK Biobank study used data from more than 19,700 participants to develop age-related hippocampal volume nomograms. These reference values help clinicians assess an individual's measurement in relation to others of a similar age.
A hippocampal volume percentile provides another way to express this comparison. However, the reference population and measurement method must be considered before applying the result clinically.
MRI hippocampal volumetry turns a brain scan into a measurable estimate of hippocampal size. The process involves several stages, and a problem at any stage can affect the final result.
The process commonly begins with a high-resolution, 3D T1-weighted MRI sequence. The images must show enough anatomical detail to identify the hippocampus and its boundaries.
Motion, distortion and poor contrast can obscure important structures. Reviewing image quality early helps identify scans that may not be suitable for reliable analysis.
Depending on the software, preprocessing may include image alignment, intensity correction and brain extraction. These operations prepare the scan for segmentation.
Manual, semi-automated or AI-assisted software identifies the left and right hippocampus. Accurate segmentation matters because nearby brain structures can be mistakenly included or excluded.
The segmented regions should be reviewed against the original MRI in multiple planes. Automated software can produce convincing-looking results that still contain errors, so quality control remains essential.
Read the research on hippocampal segmentation quality control
The software calculates the physical volume of each segmented region using voxel dimensions. Results may be reported separately for the left and right hippocampus and, where appropriate, as a combined volume.
The measured volume is compared with a suitable normative dataset. Depending on the model, adjustments may account for age, sex or intracranial volume.
The software expresses the measurement relative to the reference population. Different models can produce different standardized results from the same raw volume, so the underlying method matters.
The radiologist reviews the measurements, segmentation quality and conventional MRI findings before interpreting the results alongside the patient's symptoms and clinical history.
For follow-up scans, consistent acquisition and processing methods can make comparisons more meaningful, although they cannot eliminate every source of measurement variability.
A Z-score helps show how unusual a measurement is relative to its reference population. It does not establish a diagnosis.
| Z-score | Statistical Interpretation | Practical Meaning |
|---|---|---|
| +2.0 or higher | Substantially above the reference mean | Review the result in context; higher volume does not necessarily mean better brain health. |
| +1.0 to below +2.0 | Above the reference mean | Consider the reference model and other MRI findings. |
| −1.0 to +1.0 | Within one standard deviation of the mean | The measurement is relatively close to the reference mean, but disease is not ruled out. |
| Below −1.0 but above −2.0 | Below the reference mean | Review age adjustment, image quality and clinical history. |
| −2.0 or lower | Markedly below the reference mean | May warrant closer assessment but does not independently establish atrophy or Alzheimer’s disease. |
These ranges are illustrative statistical guideposts, not universal clinical thresholds. Assuming a standard normal distribution, a Z-score of −2.0 corresponds approximately to the 2.3rd percentile. Clinical interpretation must follow the validated reference model.
A low score may prompt closer examination, but it should not be treated as a diagnosis. Equally, a score near zero does not guarantee that the hippocampus is healthy.
These three measurements answer different questions.
For example, a measurement at the 10th percentile is higher than approximately 10% of measurements in that reference distribution and lower than approximately 90%. Its clinical significance depends on the population and method used.
Looking at absolute volume, Z-score and percentile together can provide a fuller picture than relying on any one measurement.
A hippocampal volume that is typical at one age may be less typical at another. This is why normal hippocampal volume by age should be assessed using suitable age-related reference data rather than a single cutoff for everyone.
Researchers are beginning to examine these patterns across much larger age ranges.
A 2026 preprint, Brain Charts for Neuroanatomical gender Differences Across the Human Lifespan, provides a broader look at how brain anatomy changes with age
. Researchers analyzed 108,371 MRI scans from people ranging from mid-gestation to 99 years of age, examining 241 structural brain features, including regional volume, cortical thickness and surface area.Head size also influences raw volume measurements. Some normative models adjust for intracranial volume to make comparisons more meaningful.
Research examining sex differences in hippocampal volume has highlighted the importance of accounting for overall brain size. The appropriate adjustments depend on the validated model and its intended use.
Read the structural MRI volume meta-analysis
A numerical result may look precise, but several factors can influence its reliability:
These factors are especially important when comparing scans processed using different software or scanners. Standardized methods improve consistency, but they do not eliminate uncertainty.
No. A low hippocampal volume Z-score alone cannot diagnose hippocampal atrophy or Alzheimer’s disease.
Hippocampal atrophy on MRI can be relevant when assessing neurodegenerative conditions, but reduced volume can have different explanations. The pattern of structural changes, cognitive symptoms, medical history and other clinical findings all contribute to interpretation.
People searching for brain atrophy MRI, atrophy in brain MRI or MRI hippocampal atrophy are often trying to understand what volume loss means. The important distinction is that an imaging measurement describes anatomy; it does not independently establish the cause of that finding.
MRI can support an Alzheimer’s assessment and help investigate alternative causes of cognitive symptoms. Quantitative measurements should complement clinical assessment, not replace it.
National Institute on Aging: How Alzheimer’s Disease Is Diagnosed
Manual image analysis can be time-consuming, particularly when consistent measurements are needed across many scans. AI-assisted tools can automate parts of the process, including hippocampal segmentation and volume calculation.
For example, Alzevita's product brochure describes a workflow in which users upload compatible MRI scans, generate automated segmentations and volumetric measurements, and export standardized reports.
These capabilities may support a more consistent workflow, but automation does not remove the need for image-quality checks, validated reference data or professional interpretation.
Read the Alzevita product brochure
Hippocampal volumetry provides a quantitative way to assess brain structure, while Z-scores and percentiles show how measurements compare with a reference population.
The value of these results depends on image quality, segmentation accuracy and the suitability of the normative model. A low score may identify a measurement that deserves closer review, but it does not explain the cause.
The practical takeaway is straightforward: quantitative MRI of the brain adds valuable context when interpreted alongside conventional imaging and the patient's clinical picture.