You are here

Article Text

Article menu

Download PDFPDF

SNIS 21st annual meeting electronic poster abstracts
E-064 Predicting symptomatic presentation of intracranial aneurysms using radiomics
  1. A Gudino1,
  2. A Saleem1,
  3. C Dier1,
  4. E Sagues1,
  5. M Torres1,
  6. D Ojeda1,
  7. E Garces2,
  8. A Vargas1,
  9. C Aamot1,
  10. E Samaniego3
  1. 1Neurology, University of Iowa Hospitals and Clinics, Iowa City, IA
  2. 2Neurology, Universidad San Francisco de Quito, Quito, Ecuador
  3. 3Neurology, Neurosurgery and Radiology, University of Iowa Hospitals and Clinics, Iowa City, IA

Abstract

Background Radiomics is a novel image analysis tool that offers a voxel by voxel quantification of medical images signal, shape and texture characteristics. Our aim is to determine if radiomics analysis of intracranial aneurysms (IAs) can help predict their symptomatic presentation.

Methods Patients with >2 mm saccular aneurysms were imaged using a high-resolution vessel wall imaging 3T MRI at the University of Iowa. Symptomatic aneurysms were defined as IA presenting with cranial nerve palsy (e.g. III), rupture or sentinel headache. After co-registration of T1 and T1+Gd sequences, 3D Slicer was utilized to segment the aneurysms wall and extract radiomic features (figure 1). Machine learning was used to analyze the data in conjunction with the morphology and clinical variables to create predicting nomograms (figure 2).

Results Ninety patients with 104 IAs were included and 29 were defined as symptomatic. We selected 87 radiomic features that were significantly different between symptomatic and asymptomatic IAs. The clinical nomogram combining clinical, morphological and radiomic features achieved an AUC=0.83, 89% sensitivity and 73% specificity in predicting IAs symptomatic presentation.

Conclusions Radiomics may improve the prediction of IAs symptomatic presentation and may aid in the stratification of their risk of rupture. This information is essential to guide therapeutical management of unruptured IAs.

Abstract E-064 Figure 1
Abstract E-064 Figure 1

3T HR-MRI image acquisition and segmentation in 3D slicer

Abstract E-064 Figure 2
Abstract E-064 Figure 2

Nomogram for symptomatic presentation prediction

Disclosures A. Gudino: None. A. Saleem: None. C. Dier: None. E. Sagues: None. M. Torres: None. D. Ojeda: None. E. Garces: None. A. Vargas: None. C. Aamot: None. E. Samaniego: None.

https://doi.org/10.1136/jnis-2024-SNIS.169

Statistics from Altmetric.com

    Request Permissions

    If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.