CT angiographic radiomics signature for risk stratification in anterior large vessel occlusion stroke

Background and Purpose: As “time is brain” in acute stroke triage, the need for automated prognostication tools continues to increase, particularly in rapidly expanding tele-stroke settings. We aimed to create an automated prognostication tool for anterior circulation large vessel occlusion (LVO) st...

Full description

Bibliographic Details
Main Authors: Emily W. Avery, Jonas Behland, Adrian Mak, Stefan P. Haider, Tal Zeevi, Pina C. Sanelli, Christopher G. Filippi, Ajay Malhotra, Charles C. Matouk, Christoph J. Griessenauer, Ramin Zand, Philipp Hendrix, Vida Abedi, Guido J. Falcone, Nils Petersen, Lauren H. Sansing, Kevin N. Sheth, Seyedmehdi Payabvash
Format: Article
Language:English
Published: Elsevier 2022-01-01
Series:NeuroImage: Clinical
Subjects:
Online Access:http://www.sciencedirect.com/science/article/pii/S2213158222000997
_version_ 1818232597743927296
author Emily W. Avery
Jonas Behland
Adrian Mak
Stefan P. Haider
Tal Zeevi
Pina C. Sanelli
Christopher G. Filippi
Ajay Malhotra
Charles C. Matouk
Christoph J. Griessenauer
Ramin Zand
Philipp Hendrix
Vida Abedi
Guido J. Falcone
Nils Petersen
Lauren H. Sansing
Kevin N. Sheth
Seyedmehdi Payabvash
author_facet Emily W. Avery
Jonas Behland
Adrian Mak
Stefan P. Haider
Tal Zeevi
Pina C. Sanelli
Christopher G. Filippi
Ajay Malhotra
Charles C. Matouk
Christoph J. Griessenauer
Ramin Zand
Philipp Hendrix
Vida Abedi
Guido J. Falcone
Nils Petersen
Lauren H. Sansing
Kevin N. Sheth
Seyedmehdi Payabvash
author_sort Emily W. Avery
collection DOAJ
description Background and Purpose: As “time is brain” in acute stroke triage, the need for automated prognostication tools continues to increase, particularly in rapidly expanding tele-stroke settings. We aimed to create an automated prognostication tool for anterior circulation large vessel occlusion (LVO) stroke based on admission CTA radiomics. Methods: We automatically extracted 1116 radiomics features from the anterior circulation territory on admission CTAs of 829 acute LVO stroke patients who underwent mechanical thrombectomy in two academic centers. We trained, optimized, validated, and compared different machine-learning models to predict favorable outcome (modified Rankin Scale ≤ 2) at discharge and 3-month follow-up using four different input sets: “Radiomics”, “Radiomics + Treatment” (radiomics, post-thrombectomy reperfusion grade, and intravenous thrombolysis), “Clinical + Treatment” (baseline clinical variables and treatment), and “Combined” (radiomics, treatment, and baseline clinical variables). Results: For discharge outcome prediction, models were optimized/trained on n = 494 and tested on an independent cohort of n = 100 patients from Yale. Receiver operating characteristic analysis of the independent cohort showed no significant difference between best-performing Combined input models (area under the curve, AUC = 0.77) versus Radiomics + Treatment (AUC = 0.78, p = 0.78), Radiomics (AUC = 0.78, p = 0.55), or Clinical + Treatment (AUC = 0.77, p = 0.87) models. For 3-month outcome prediction, models were optimized/trained on n = 373 and tested on an independent cohort from Yale (n = 72), and an external cohort from Geisinger Medical Center (n = 232). In the independent cohort, there was no significant difference between Combined input models (AUC = 0.76) versus Radiomics + Treatment (AUC = 0.72, p = 0.39), Radiomics (AUC = 0.72, p = 0.39), or Clinical + Treatment (AUC = 76, p = 0.90) models; however, in the external cohort, the Combined model (AUC = 0.74) outperformed Radiomics + Treatment (AUC = 0.66, p < 0.001) and Radiomics (AUC = 0.68, p = 0.005) models for 3-month prediction. Conclusion: Machine-learning signatures of admission CTA radiomics can provide prognostic information in acute LVO stroke candidates for mechanical thrombectomy. Such objective and time-sensitive risk stratification can guide treatment decisions and facilitate tele-stroke assessment of patients. Particularly in the absence of reliable clinical information at the time of admission, models solely using radiomics features can provide a useful prognostication tool.
first_indexed 2024-12-12T11:08:49Z
format Article
id doaj.art-bb4bf1af3c8544edbf1615d58e844010
institution Directory Open Access Journal
issn 2213-1582
language English
last_indexed 2024-12-12T11:08:49Z
publishDate 2022-01-01
publisher Elsevier
record_format Article
series NeuroImage: Clinical
spelling doaj.art-bb4bf1af3c8544edbf1615d58e8440102022-12-22T00:26:21ZengElsevierNeuroImage: Clinical2213-15822022-01-0134103034CT angiographic radiomics signature for risk stratification in anterior large vessel occlusion strokeEmily W. Avery0Jonas Behland1Adrian Mak2Stefan P. Haider3Tal Zeevi4Pina C. Sanelli5Christopher G. Filippi6Ajay Malhotra7Charles C. Matouk8Christoph J. Griessenauer9Ramin Zand10Philipp Hendrix11Vida Abedi12Guido J. Falcone13Nils Petersen14Lauren H. Sansing15Kevin N. Sheth16Seyedmehdi Payabvash17Section of Neuroradiology, Department of Radiology and Biomedical Imaging, Yale School of Medicine, New Haven, CT, United StatesSection of Neuroradiology, Department of Radiology and Biomedical Imaging, Yale School of Medicine, New Haven, CT, United States; CLAIM - Charité Lab for Artificial Intelligence in Medicine, Charité Universitätsmedizin Berlin, Berlin, GermanySection of Neuroradiology, Department of Radiology and Biomedical Imaging, Yale School of Medicine, New Haven, CT, United States; CLAIM - Charité Lab for Artificial Intelligence in Medicine, Charité Universitätsmedizin Berlin, Berlin, GermanySection of Neuroradiology, Department of Radiology and Biomedical Imaging, Yale School of Medicine, New Haven, CT, United States; Department of Otorhinolaryngology, University Hospital of Ludwig Maximilians Universität München, Munich, GermanySection of Neuroradiology, Department of Radiology and Biomedical Imaging, Yale School of Medicine, New Haven, CT, United StatesSection of Neuroradiology, Department of Radiology, Northwell Health, Manhasset, NY, United StatesSection of Neuroradiology, Department of Radiology, Tufts School of Medicine, Boston, MA, United StatesSection of Neuroradiology, Department of Radiology and Biomedical Imaging, Yale School of Medicine, New Haven, CT, United StatesDivision of Neurovascular Surgery, Department of Neurosurgery, Yale University School of Medicine, New Haven, CT, United StatesDepartment of Neurosurgery, Geisinger Medical Center, Danville, PA, United States; Research Institute of Neurointervention, Paracelsus Medical University, Salzburg, Austria; Department of Neurosurgery, Paracelsus Medical University, Salzburg, AustriaDepartment of Neurology, Geisinger, Danville, PA, United StatesDepartment of Neurosurgery, Geisinger Medical Center, Danville, PA, United States; Department of Neurosurgery, Saarland University Medical Center, Homburg, GermanyDepartment of Molecular and Functional Genomics, Geisinger, Danville, PA, United States; Biocomplexity Institute, Virginia Tech, Blacksburg, VA, USADivision of Neurocritical Care and Emergency Neurology, Department of Neurology, Yale University School of Medicine, New Haven, CT, United StatesDivision of Neurocritical Care and Emergency Neurology, Department of Neurology, Yale University School of Medicine, New Haven, CT, United StatesDivision of Stroke and Vascular Neurology, Department of Neurology, Yale University School of Medicine, New Haven, CT, United StatesDivision of Neurocritical Care and Emergency Neurology, Department of Neurology, Yale University School of Medicine, New Haven, CT, United StatesSection of Neuroradiology, Department of Radiology and Biomedical Imaging, Yale School of Medicine, New Haven, CT, United States; Corresponding author at: Department of Radiology and Biomedical Imaging, Yale School of Medicine, 789 Howard Ave, PO Box 208042, New Haven, CT 06519, United States.Background and Purpose: As “time is brain” in acute stroke triage, the need for automated prognostication tools continues to increase, particularly in rapidly expanding tele-stroke settings. We aimed to create an automated prognostication tool for anterior circulation large vessel occlusion (LVO) stroke based on admission CTA radiomics. Methods: We automatically extracted 1116 radiomics features from the anterior circulation territory on admission CTAs of 829 acute LVO stroke patients who underwent mechanical thrombectomy in two academic centers. We trained, optimized, validated, and compared different machine-learning models to predict favorable outcome (modified Rankin Scale ≤ 2) at discharge and 3-month follow-up using four different input sets: “Radiomics”, “Radiomics + Treatment” (radiomics, post-thrombectomy reperfusion grade, and intravenous thrombolysis), “Clinical + Treatment” (baseline clinical variables and treatment), and “Combined” (radiomics, treatment, and baseline clinical variables). Results: For discharge outcome prediction, models were optimized/trained on n = 494 and tested on an independent cohort of n = 100 patients from Yale. Receiver operating characteristic analysis of the independent cohort showed no significant difference between best-performing Combined input models (area under the curve, AUC = 0.77) versus Radiomics + Treatment (AUC = 0.78, p = 0.78), Radiomics (AUC = 0.78, p = 0.55), or Clinical + Treatment (AUC = 0.77, p = 0.87) models. For 3-month outcome prediction, models were optimized/trained on n = 373 and tested on an independent cohort from Yale (n = 72), and an external cohort from Geisinger Medical Center (n = 232). In the independent cohort, there was no significant difference between Combined input models (AUC = 0.76) versus Radiomics + Treatment (AUC = 0.72, p = 0.39), Radiomics (AUC = 0.72, p = 0.39), or Clinical + Treatment (AUC = 76, p = 0.90) models; however, in the external cohort, the Combined model (AUC = 0.74) outperformed Radiomics + Treatment (AUC = 0.66, p < 0.001) and Radiomics (AUC = 0.68, p = 0.005) models for 3-month prediction. Conclusion: Machine-learning signatures of admission CTA radiomics can provide prognostic information in acute LVO stroke candidates for mechanical thrombectomy. Such objective and time-sensitive risk stratification can guide treatment decisions and facilitate tele-stroke assessment of patients. Particularly in the absence of reliable clinical information at the time of admission, models solely using radiomics features can provide a useful prognostication tool.http://www.sciencedirect.com/science/article/pii/S2213158222000997RadiomicsStrokeLarge vessel occlusionCTAQuantitative imagingMechanical thrombectomy
spellingShingle Emily W. Avery
Jonas Behland
Adrian Mak
Stefan P. Haider
Tal Zeevi
Pina C. Sanelli
Christopher G. Filippi
Ajay Malhotra
Charles C. Matouk
Christoph J. Griessenauer
Ramin Zand
Philipp Hendrix
Vida Abedi
Guido J. Falcone
Nils Petersen
Lauren H. Sansing
Kevin N. Sheth
Seyedmehdi Payabvash
CT angiographic radiomics signature for risk stratification in anterior large vessel occlusion stroke
NeuroImage: Clinical
Radiomics
Stroke
Large vessel occlusion
CTA
Quantitative imaging
Mechanical thrombectomy
title CT angiographic radiomics signature for risk stratification in anterior large vessel occlusion stroke
title_full CT angiographic radiomics signature for risk stratification in anterior large vessel occlusion stroke
title_fullStr CT angiographic radiomics signature for risk stratification in anterior large vessel occlusion stroke
title_full_unstemmed CT angiographic radiomics signature for risk stratification in anterior large vessel occlusion stroke
title_short CT angiographic radiomics signature for risk stratification in anterior large vessel occlusion stroke
title_sort ct angiographic radiomics signature for risk stratification in anterior large vessel occlusion stroke
topic Radiomics
Stroke
Large vessel occlusion
CTA
Quantitative imaging
Mechanical thrombectomy
url http://www.sciencedirect.com/science/article/pii/S2213158222000997
work_keys_str_mv AT emilywavery ctangiographicradiomicssignatureforriskstratificationinanteriorlargevesselocclusionstroke
AT jonasbehland ctangiographicradiomicssignatureforriskstratificationinanteriorlargevesselocclusionstroke
AT adrianmak ctangiographicradiomicssignatureforriskstratificationinanteriorlargevesselocclusionstroke
AT stefanphaider ctangiographicradiomicssignatureforriskstratificationinanteriorlargevesselocclusionstroke
AT talzeevi ctangiographicradiomicssignatureforriskstratificationinanteriorlargevesselocclusionstroke
AT pinacsanelli ctangiographicradiomicssignatureforriskstratificationinanteriorlargevesselocclusionstroke
AT christophergfilippi ctangiographicradiomicssignatureforriskstratificationinanteriorlargevesselocclusionstroke
AT ajaymalhotra ctangiographicradiomicssignatureforriskstratificationinanteriorlargevesselocclusionstroke
AT charlescmatouk ctangiographicradiomicssignatureforriskstratificationinanteriorlargevesselocclusionstroke
AT christophjgriessenauer ctangiographicradiomicssignatureforriskstratificationinanteriorlargevesselocclusionstroke
AT raminzand ctangiographicradiomicssignatureforriskstratificationinanteriorlargevesselocclusionstroke
AT philipphendrix ctangiographicradiomicssignatureforriskstratificationinanteriorlargevesselocclusionstroke
AT vidaabedi ctangiographicradiomicssignatureforriskstratificationinanteriorlargevesselocclusionstroke
AT guidojfalcone ctangiographicradiomicssignatureforriskstratificationinanteriorlargevesselocclusionstroke
AT nilspetersen ctangiographicradiomicssignatureforriskstratificationinanteriorlargevesselocclusionstroke
AT laurenhsansing ctangiographicradiomicssignatureforriskstratificationinanteriorlargevesselocclusionstroke
AT kevinnsheth ctangiographicradiomicssignatureforriskstratificationinanteriorlargevesselocclusionstroke
AT seyedmehdipayabvash ctangiographicradiomicssignatureforriskstratificationinanteriorlargevesselocclusionstroke