An automated ASPECTS method with atlas-based segmentation

Yu, Zechen; Chen, Zhongping; Yu, Yang; Zhu, Hongtao; Tong, Dan; Chen, Yang

doi:10.1016/j.cmpb.2021.106376

Cited by 5 publications

(2 citation statements)

References 21 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…While the ASPECTS remains a dependable and uncomplicated scoring system for appraising early ischemic changes within the distribution zones of the MCA, it is crucial to acknowledge the potential variability in interpretations due to divergent levels of expertise among non-contrast CT observers. 34 In contrast, the RAPID ASPECTS system harnesses validated machine-learning algorithms to autonomously identify cerebral regions and compute scores, thereby enabling swift and more harmonious assessment for thrombectomy eligibility (j = 0.9), surpassing the concordance of human readers (j = 0.57 and j = 0.56). 31 Prior investigative trajectories focusing on the predictive potential of CT Hu values for DCI and clinical outcomes predominantly centered on average Hu values within specific hemorrhage tiers.…”

Section: Discussionmentioning

confidence: 99%

Insular cortex Hounsfield units predict postoperative neurocardiogenic injury in patients with aneurysmal subarachnoid hemorrhage

Jia,

Lin,

et al. 2023

Ann Clin Transl Neurol

View full text Add to dashboard Cite

ObjectiveOur study aims to investigate the association between the Hounsfield unit (Hu) value of the insular cortex (IC) during emergency admission and the subsequent occurrence of post‐operative neurocardiogenic injury (NCI) among patients afflicted with aneurysmal subarachnoid hemorrhage (aSAH).MethodsPatients baseline characteristics were juxtaposed between those with and without NCI. The significant variables were incorporated into a multivariable stepwise logistic regression model. Receiver operating characteristic (ROC) curves were drafted for each significant variable, yielding cutoff values and the area under the curve (AUC). Subgroup and sensitivity analyses were performed to assess the predictive performance across various cohorts and ascertain result stability. Propensity score matching (PSM) was ultimately employed to redress any baseline characteristic disparities.ResultsPatients displaying a right IC Hu value surpassing 28.65 exhibited an escalated risk of postoperative NCI upon confounder adjustment (p < 0.001). The ROC curve eloquently manifested the predictive capacity of right IC Hu in relation to NCI (AUC = 0.650, 95%CI, 0.591–0.709, p < 0.001). Further subgroup analysis revealed significant interactions between right IC Hu and factors such as age, history of heart disease, and Graeb 5–12 score. Sensitivity analysis further upheld the results' significant (p = 0.002). The discrepancy in NCI incidence between the two groups, both prior (p < 0.002) and post (p = 0.039) PSM, exhibited statistical significance. After PSM implementation, the likelihood of NCI displayed an ascending trend with increasing right IC Hu values, from the Hu1 cohort onward, receding post the Hu4 cohort.ConclusionThis study definitively establishes an elevated right IC Hu value in the early stages of emergency admission as an autonomous predictor for ensuing NCI subsequent to aSAH.

show abstract

Section: Discussionmentioning

confidence: 99%

Insular cortex Hounsfield units predict postoperative neurocardiogenic injury in patients with aneurysmal subarachnoid hemorrhage

Jia,

Lin,

et al. 2023

Ann Clin Transl Neurol

View full text Add to dashboard Cite

show abstract

“…The four other variables are the patient age, baseline NIHSS, glucose levels and baseline ASPECTS score (also an imaging biomarker). In this work, we used expert annotations for this variable, but there are several published works [42,43] and commercial products [44,45] that predict the ASPECT score using machine learning. The particular selection of five variables in the LR 5vars SN model came from a preliminary analysis of the feature importance of the LR 8vars experiment (second row of Figure 5).…”

Section: Hybrid Approachmentioning

confidence: 99%

Deep Learning-Based Extraction of Biomarkers for the Prediction of the Functional Outcome of Ischemic Stroke Patients

Oliveira,

Fonseca,

Ferro

et al. 2023

Diagnostics

View full text Add to dashboard Cite

Accurately predicting functional outcomes in stroke patients remains challenging yet clinically relevant. While brain CTs provide prognostic information, their practical value for outcome prediction is unclear. We analyzed a multi-center cohort of 743 ischemic stroke patients (<72 h onset), including their admission brain NCCT and CTA scans as well as their clinical data. Our goal was to predict the patients’ future functional outcome, measured by the 3-month post-stroke modified Rankin Scale (mRS), dichotomized into good (mRS ≤ 2) and poor (mRS > 2). To this end, we developed deep learning models to predict the outcome from CT data only, and models that incorporate other patient variables. Three deep learning architectures were tested in the image-only prediction, achieving 0.779 ± 0.005 AUC. In addition, we created a model fusing imaging and tabular data by feeding the output of a deep learning model trained to detect occlusions on CT angiograms into our prediction framework, which achieved an AUC of 0.806 ± 0.082. These findings highlight how further refinement of prognostic models incorporating both image biomarkers and clinical data could enable more accurate outcome prediction for ischemic stroke patients.

show abstract

Automated ASPECTS Segmentation and Scoring Tool: a Method Tailored for a Colombian Telestroke Network

Ortiz,

Rivera,

Granja

et al. 2024

j Imaging. Inform. med.

View full text Add to dashboard Cite

To evaluate our two non-machine learning (non-ML)-based algorithmic approaches for detecting early ischemic infarcts on brain CT images of patients with acute ischemic stroke symptoms, tailored to our local population, to be incorporated in our telestroke software. One-hundred and thirteen acute stroke patients, excluding hemorrhagic, subacute, and chronic patients, with accessible brain CT images were divided into calibration and test sets. The gold standard was determined through consensus among three neuroradiologist. Four neuroradiologist independently reported Alberta Stroke Program Early CT Scores (ASPECTSs). ASPECTSs were also obtained using a commercial ML solution (CMLS), and our two methods, namely the Mean Hounsfield Unit (HU) relative difference (RELDIF) and the density distribution equivalence test (DDET), which used statistical analyze the of the HUs of each region and its contralateral side. Automated segmentation was perfect for cortical regions, while minimal adjustment was required for basal ganglia regions. For dichotomized-ASPECTSs (ASPECTS < 6) in the test set, the area under the receiver operating characteristic curve (AUC) was 0.85 for the DDET method, 0.84 for the RELDIF approach, 0.64 for the CMLS, and ranged from 0.71–0.89 for the neuroradiologist. The accuracy was 0.85 for the DDET method, 0.88 for the RELDIF approach, and was ranged from 0.83 − 0.96 for the neuroradiologist. Equivalence at a margin of 5% was documented among the DDET, RELDIF, and gold standard on mean ASPECTSs. Noninferiority tests of the AUC and accuracy of infarct detection revealed similarities between both DDET and RELDIF, and the CMLS, and with at least one neuroradiologist. The alignment of our methods with the evaluations of neuroradiologist and the CMLS indicates the potential of our methods to serve as supportive tools in clinical settings, facilitating prompt and accurate stroke diagnosis, especially in health care settings, such as Colombia, where neuroradiologist are limited.

show abstract

An automated ASPECTS method with atlas-based segmentation

Cited by 5 publications

References 21 publications

Insular cortex Hounsfield units predict postoperative neurocardiogenic injury in patients with aneurysmal subarachnoid hemorrhage

Insular cortex Hounsfield units predict postoperative neurocardiogenic injury in patients with aneurysmal subarachnoid hemorrhage

Deep Learning-Based Extraction of Biomarkers for the Prediction of the Functional Outcome of Ischemic Stroke Patients

Automated ASPECTS Segmentation and Scoring Tool: a Method Tailored for a Colombian Telestroke Network

Contact Info

Product

Resources

About