MRI-negative epilepsy describes focal epilepsy in which no clear epileptogenic lesion is identified on structural MRI. It does not mean there is no structural abnormality, and it does not automatically rule out epilepsy surgery.
Why the term is imperfect
A 2025 systematic review found major variation in how studies define “MRI-negative”.
Some apparently negative scans were performed without modern epilepsy protocols, expert neuroradiology review or post-processing.
Step one: make sure the MRI is truly negative
Before accepting the label, specialists may repeat MRI using a dedicated epilepsy protocol, review the images with the electroclinical hypothesis available, and apply morphometric post-processing.
The 2025 meta-analysis found a substantial increase in lesion detection when post-processing was used.
Why surgery is harder without a visible lesion
A lesion provides an anatomical target and often strengthens confidence that EEG and semiology refer to the same region.
Without that target, the team must reconstruct the epileptogenic network from several indirect sources of evidence.
Multimodal localisation
Assessment may combine:
- video-EEG;
- high-resolution MRI and expert re-review;
- MRI post-processing;
- FDG-PET;
- ictal SPECT;
- MEG or EEG source imaging;
- neuropsychology;
- functional MRI;
- SEEG.
SEEG is used more often
The 2025 systematic review found that MRI-negative patients underwent SEEG more often than MRI-positive patients.
This reflects the greater uncertainty that must be resolved before treatment.
Surgical outcome
Average seizure outcomes are less favourable in MRI-negative epilepsy than when a concordant lesion is visible.
The same 2025 meta-analysis found favourable outcomes in roughly 58% of MRI-negative surgical patients compared with about 75% of MRI-positive surgical patients in the pooled analysis.
These figures are group averages, not predictions for an individual.
Pathology may still be found after surgery
Some people labelled MRI-negative are later found to have focal cortical dysplasia, gliosis or another subtle abnormality when tissue is examined histologically.
Concordance matters
Surgery is most convincing when independent tests converge on the same cortical region or network.
Discordant data increase the risk of removing tissue that is not the true epileptogenic zone.
When surgery may not be appropriate
If seizures arise from multiple independent regions, localisation remains uncertain, or the necessary treatment would cause unacceptable functional loss, resection may not be offered.
Neurostimulation or palliative strategies may then be considered.
Normal MRI should not delay referral
People with drug-resistant focal epilepsy should not be excluded from comprehensive epilepsy-surgery assessment solely because routine MRI is normal.
The central point
MRI-negative epilepsy is a localisation challenge, not a surgical contraindication. The pathway relies on modern imaging, multimodal concordance and often SEEG to identify a safe and credible treatment target.
Related reading
MRI Post-Processing →
FDG-PET →
SEEG →
Presurgical Assessment →
Sources
Evidence checked: September 2026.