Related Experiment Video
Updated: Oct 21, 2025
12:30
A Comprehensive Protocol for Manual Segmentation of the Medial Temporal Lobe Structures
Published on: July 2, 2014
20.5K
An automated ASPECTS method with atlas-based segmentation
Zechen Yu1, Zhongping Chen2, Yang Yu2
1Laboratory of Image Science and Technology, Southeast University, Nanjing 210096, China; Key Laboratory of Computer Network and Information Integration (Southeast University), Ministry of Education, Nanjing 210096, China.
Computer Methods and Programs in Biomedicine
|September 9, 2021
Summary
An automated method for evaluating early ischemic changes in acute ischemic stroke (AIS) patients using non-contrast computed tomography (NCCT) shows promise. This auto-ASPECTS approach improves diagnostic accuracy and aids in treatment decisions.
Area of Science:
- Neuroradiology
- Medical Imaging Analysis
- Stroke Diagnostics
Background:
- Alberta Stroke Program Early CT score (ASPECTS) is a standard for evaluating ischemic stroke but suffers from inter-reader variability.
- Accurate and timely assessment of early ischemic changes is crucial for treatment selection and prognosis in acute ischemic stroke (AIS).
Purpose of the Study:
- To develop and validate an automated semi-quantitative method (auto-ASPECTS) for diagnosing acute ischemic stroke using non-contrast computed tomography (NCCT).
- To provide an objective reference for doctors in the diagnosis and evaluation of early ischemic changes.
Main Methods:
- Non-contrast computed tomography (NCCT) data from 90 patients were used for training and testing the auto-ASPECTS system.
- Atlas-based segmentation was employed to define regions of interest for ASPECTS.
- Brain density shifts (BDS) of contralateral brain regions were utilized as a quantitative standard for analysis.
Main Results:
- The auto-ASPECTS method using Brain Density Shifts (BDS) achieved an accuracy of 0.80 in the test set.
- Using different BDS thresholds improved accuracy by 6.67% compared to a consistent threshold.
- The agreement between dichotomized auto-ASPECTS and consensus scores was 83.3%.
Conclusions:
- The proposed automated ASPECTS method for NCCT images offers valuable information for the early diagnosis and evaluation of acute ischemic stroke (AIS).
- This automated approach can assist clinicians in making more consistent and accurate assessments of ischemic stroke severity.
