Abstract
Purpose
To compare logistic regression to elastic net for identifying and ranking clinical risk factors for blunt cerebrovascular injury (BCVI).
Materials and methods
Consecutive trauma patients undergoing screening CTA at a level 1 trauma center over a 2-year period. Each internal carotid artery (ICA) and vertebral artery (VA) was independently graded by 2 neuroradiologists using the Denver grading scale. Unadjusted odds ratios were calculated by univariate and adjusted odds ratios by multiple logistic regression with FDR correction. We applied logistic regression with the elastic net penalty and tenfold cross-validation.
Results
Total of 467 patients; 73 patients with BCVI. Maxillofacial fracture, basilar skull fracture, and GCS had significant unadjusted odds ratios (OR) for ICA injury and C-spine fracture, spinal ligamentous injury, and age for VA injury. Only transverse foramen fracture had significant adjusted OR for VA injury, with none for ICA injury, after FDR correction. Using elastic net, ICA injury variables included maxillofacial fracture, basilar skull fracture, GCS, and carotid canal fracture. For VA injury, these included cervical spine transverse foramen fracture, ligamentous injury, C1–C3 fractures, posterior element fracture, and vertebral body fracture.
Conclusion
Elastic net statistical learning methods identified additional risk factors and outperformed multiple logistic regression for BCVI. Elastic net allows the study of a large number of variables, and is useful when covariates are correlated.
Similar content being viewed by others
References
Harrigan MR, Falola MI, Shannon CN et al (2014) Incidence and trends in the diagnosis of traumatic extracranial cerebrovascular injury in the nationwide inpatient sample database, 2003–2010. J Neurotrauma 31:1056–1062
Rutman AM, Vranic JE, Mossa-Basha M (2018) Imaging and Management of Blunt Cerebrovascular Injury. Radiographics 38:542–563
DiCocco JM, Fabian TC, Emmett KP et al (2013) Functional outcomes following blunt cerebrovascular injury. J Trauma Acute Care Surg 74:955–960
Geddes AE, Burlew CC, Wagenaar AE et al (2016) Expanded screening criteria for blunt cerebrovascular injury: a bigger impact than anticipated. Am J Surg 212:1167–1174
DiCocco JM, Fabian TC, Emmett KP et al (2011) Optimal outcomes for patients with blunt cerebrovascular injury (BCVI): tailoring treatment to the lesion. J Am Coll Surg 212:549–557 (discussion 557-549)
Burlew CC, Sumislawski JJ, Behnfield CD et al (2018) Time to stroke: A Western Trauma Association multicenter study of blunt cerebrovascular injuries. J Trauma Acute Care Surg 85:858–866
Crissey MM, Bernstein EF (1974) Delayed presentation of carotid intimal tear following blunt craniocervical trauma. Surgery 75:543–549
Nace SR, Gentry LR. Cerebrovascular trauma. Neuroimaging Clin N Am 2014;24:487–511, viii
Biffl WL, Ray CE Jr, Moore EE et al (2002) Treatment-related outcomes from blunt cerebrovascular injuries: importance of routine follow-up arteriography. Ann Surg 235:699–706 (discussion 706-697)
Cogbill TH, Moore EE, Meissner M et al (1994) The spectrum of blunt injury to the carotid artery: a multicenter perspective. J Trauma 37:473–479
Franz RW, Willette PA, Wood MJ et al (2012) A systematic review and meta-analysis of diagnostic screening criteria for blunt cerebrovascular injuries. J Am Coll Surg 214:313–327
Martin RF, Eldrup-Jorgensen J, Clark DE et al (1991) Blunt trauma to the carotid arteries. J Vascular Surg 14:789–793 (discussion 793-785)
Buch K, Nguyen T, Mahoney E, et al. (2015) Association between cervical spine and skull-base fractures and blunt cerebrovascular injury. Eur Radiol
Bruns BR, Tesoriero R, Kufera J et al (2014) Blunt cerebrovascular injury screening guidelines: what are we willing to miss? J Trauma Acute Care Surg 76:691–695
Malhotra A, Wu X, Seifert K (2018) Blunt Cerebrovascular Injuries: Advances in Screening, Imaging, and Management Trends. AJNR Am J Neuroradiol 39:E103
Flashburg E, Ong AW, Muller A et al (2019) Fall downs should not fall out: Blunt cerebrovascular injury in geriatric patients after low-energy trauma is common. J Trauma Acute Care Surg 86:1010–1014
Grigorian A, Kabutey NK, Schubl S et al (2018) Blunt cerebrovascular injury incidence, stroke-rate, and mortality with the expanded Denver criteria. Surgery 164:494–499
Leraas HJ, Kuchibhatla M, Nag UP et al (2019) Cervical seatbelt sign is not associated with blunt cerebrovascular injury in children: A review of the national trauma databank. Am J Surg 218:100–105
H Z, Hastie T, (2005) Regularization and variable selection via the elastic net. J R Stat Soc Ser B Stat Methodol 67:301–320
Hastie T, Tibshirani R, Friedman J (2001) The elements of statistical learning: Data Mining, Inference, and Prediction. Springer, New York, NY
Biffl WL, Cothren CC, Moore EE et al (2009) Western Trauma Association critical decisions in trauma: screening for and treatment of blunt cerebrovascular injuries. J Trauma 67:1150–1153
Biffl WL, Moore EE, Offner PJ et al (1999) Blunt carotid arterial injuries: implications of a new grading scale. J Trauma 47:845–853
Bensch FV, Varjonen EA, Pyhalto TT et al (2019) Augmenting Denver criteria yields increased BCVI detection, with screening showing markedly increased risk for subsequent ischemic stroke. Emerg Radiol 26:365–372
Benjamini Y, Hochberg Y (1995) Controlling the false discovery rate: a practical and powerful approach to multiple testing. J Roy Stat Soc: Ser B (Methodol) 57:289–300
Fox J (2015) Applied regression analysis generalized linear models. SAGE Publications, Inc., Thousand Oaks, CA
Hurvich CM, Tsai CL (1990) The Impact of Model Selection on Inference in Linear Regression. Am Stat 44:214–217
Thompson B (1995) Stepwise Regression and Stepwise Discriminant Analysis Need Not Apply here: A Guidelines Editorial. Educ Psychol Measur 55:525–534
Smith G (2018) Step away from stepwise. J Big Data 5:32
Hastie T, Tibshirani R, Wainwright M (2015) Statistical learning with sparsity: the lasso and generalizations. Taylor & Francis Group, LLC, Boca Raton, FL
Friedman J, Hastie T, Tibshirani R (2010) Regularization Paths for Generalized Linear Models via Coordinate Descent. J Stat Softw 33:1–22
Burlew CC, Biffl WL, Moore EE et al (2012) Blunt cerebrovascular injuries: redefining screening criteria in the era of noninvasive diagnosis. J Trauma Acute Care Surg 72:330–335 (discussion 336-337 quiz 539)
Cothren CC, Moore EE, Ray CE Jr et al (2007) Cervical spine fracture patterns mandating screening to rule out blunt cerebrovascular injury. Surgery 141:76–82
Berne JD, Cook A, Rowe SA et al (2010) A multivariate logistic regression analysis of risk factors for blunt cerebrovascular injury. J Vasc Surg 51:57–64
Desai NK, Kang J, Chokshi FH (2014) Screening CT angiography for pediatric blunt cerebrovascular injury with emphasis on the cervical “seatbelt sign.” AJNR Am J Neuroradiol 35:1836–1840
Rozycki GS, Tremblay L, Feliciano DV et al (2002) A prospective study for the detection of vascular injury in adult and pediatric patients with cervicothoracic seat belt signs. J Trauma 52:618–623 (discussion 623-614)
Chen T, Guestrin C (2016) XGBoost: A Scalable Tree Boosting System. Proceedings of the 22nd ACM SIGKDD International Conference on Knowledge Discovery and Data Mining. San Francisco, California, USA: Association for Computing Machinery:785–794
Le DT, Barhorst KA, Castiglione J et al (2020) Blunt cerebrovascular injury in the geriatric population. Neurosurg Focus 49:E10
Author information
Authors and Affiliations
Contributions
Jason Allen, Arthur Fountain, and Amanda Corey conceived and designed the study and collected the data. Data analysis was performed by all authors. The first draft of the manuscript was written by Maxwell Cooper and all authors commented on previous versions of the manuscript. All authors read and approved the final manuscript.
Corresponding author
Ethics declarations
IRB approval
The protocol for this study was reviewed by the Institutional Review Board at our institution.
Conflict of interest
The authors declare that they have no conflict of interest.
Additional information
Publisher's note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Supplementary Information
Below is the link to the electronic supplementary material.
Rights and permissions
About this article
Cite this article
Cooper, M.E., Risk, B., Corey, A. et al. Statistical learning of blunt cerebrovascular injury risk factors using the elastic net. Emerg Radiol 28, 929–937 (2021). https://doi.org/10.1007/s10140-021-01949-8
Received:
Accepted:
Published:
Issue Date:
DOI: https://doi.org/10.1007/s10140-021-01949-8