Skip to main navigation Skip to search Skip to main content

Quantitative assessment of apparent diffusion coefficient for neurological outcome prediction in status epilepticus: a pilot study

  • Soo Hyun Park
  • , Byung Euk Joo
  • , Tae Jung Kim
  • , Sang Bae Ko
  • , Kyungbok Lee
  • , Dong Eog Kim
  • , Wi Sun Ryu
  • Soonchunhyang University
  • Seoul National University
  • JLK Inc.

Research output: Contribution to journalArticlepeer-review

Abstract

Status epilepticus (SE) is a neurological emergency with high morbidity and mortality. Early prognostication remains challenging, particularly in intensive care settings where clinical evaluations are limited. We investigated whether the voxel-wise proportion of preserved diffusion on diffusion-weighted imaging (DWI)—defined as the percentage of brain voxels with apparent diffusion coefficient (ADC) values between 600 and 1300 × 10⁻⁶ mm²/s—may represent a potential imaging marker associated with clinical outcomes that warrants further validation. We retrospectively analyzed 59 patients with SE who underwent DWI and electroencephalography within 72 h of seizure onset. ADC quantification was performed using a fully automated, segmentation-free pipeline. Patients were stratified by tertiles and dichotomized using a receiver operating characteristic (ROC)-derived threshold. The primary outcome was defined as a good outcome, corresponding to no change or improvement of at least one point on the modified Rankin Scale from premorbid baseline to hospital discharge. Patients in the highest ADC tertile had significantly better outcomes (odds ratio [OR] 5.67, p = 0.024). At the optimal threshold of 0.797, preserved ADC was associated with favorable outcomes after adjustment for clinical variables (OR 6.05, p = 0.045), along with younger age and lower EEG severity. A combined clinical–ADC model achieved an area under the ROC curve of 0.868. The preserved ADC was associated with improved outcome prediction and may provide exploratory imaging information for early risk assessment in patients with SE.

Original languageEnglish
Article number14220
JournalScientific Reports
Volume16
Issue number1
DOIs
StatePublished - Dec 2026

Keywords

  • Apparent diffusion coefficient
  • Artificial intelligence
  • Diffusion-weighted imaging
  • Neurocritical care
  • Outcome prediction
  • Status epilepticus

Fingerprint

Dive into the research topics of 'Quantitative assessment of apparent diffusion coefficient for neurological outcome prediction in status epilepticus: a pilot study'. Together they form a unique fingerprint.

Cite this