Use of an artificial neural network to predict head injury outcome

Clinical article

View More View Less
  • 1 Division of Neurosurgery and
  • 2 Department of Computer Science, University of Vermont, Burlington, Vermont
Restricted access

Purchase Now

USD  $45.00

JNS + Pediatrics - 1 year subscription bundle (Individuals Only)

USD  $505.00

JNS + Pediatrics + Spine - 1 year subscription bundle (Individuals Only)

USD  $600.00
Print or Print + Online

Object

The authors describe the artificial neural network (ANN) as an innovative and powerful modeling tool that can be increasingly applied to develop predictive models in neurosurgery. They aimed to demonstrate the utility of an ANN in predicting survival following traumatic brain injury and compare its predictive ability with that of regression models and clinicians.

Methods

The authors designed an ANN to predict in-hospital survival following traumatic brain injury. The model was generated with 11 clinical inputs and a single output. Using a subset of the National Trauma Database, the authors “trained” the model to predict outcome by providing the model with patients for whom 11 clinical inputs were paired with known outcomes, which allowed the ANN to “learn” the relevant relationships that predict outcome. The model was tested against actual outcomes in a novel subset of 100 patients derived from the same database. For comparison with traditional forms of modeling, 2 regression models were developed using the same training set and were evaluated on the same testing set. Lastly, the authors used the same 100-patient testing set to evaluate 5 neurosurgery residents and 4 neurosurgery staff physicians on their ability to predict survival on the basis of the same 11 data points that were provided to the ANN. The ANN was compared with the clinicians and the regression models in terms of accuracy, sensitivity, specificity, and discrimination.

Results

Compared with regression models, the ANN was more accurate (p < 0.001), more sensitive (p < 0.001), as specific (p = 0.260), and more discriminating (p < 0.001). There was no difference between the neurosurgery residents and staff physicians, and all clinicians were pooled to compare with the 5 best neural networks. The ANNs were more accurate (p < 0.0001), more sensitive (p < 0.0001), as specific (p = 0.743), and more discriminating (p < 0.0001) than the clinicians.

Conclusions

When given the same limited clinical information, the ANN significantly outperformed regression models and clinicians on multiple performance measures. While this paradigm certainly does not adequately reflect a real clinical scenario, this form of modeling could ultimately serve as a useful clinical decision support tool. As the model evolves to include more complex clinical variables, the performance gap over clinicians and logistic regression models will persist or, ideally, further increase.

Abbreviations used in this paper: ANN = artificial neural network; AUROC = area under the receiver operating characteristic curve; ED = emergency department; GCS = Glasgow Coma Scale; GOS = Glasgow Outcome Scale; ICP = intracranial pressure; NTDB = National Trauma Data Bank; TBI = traumatic brain injury.

JNS + Pediatrics - 1 year subscription bundle (Individuals Only)

USD  $505.00

JNS + Pediatrics + Spine - 1 year subscription bundle (Individuals Only)

USD  $600.00

Contributor Notes

Address correspondence to: Anand I. Rughani, M.D., 111 Colchester Avenue, Fletcher 5, Burlington, Vermont 05401. email: anand.rughani@vtmednet.org.

Please include this information when citing this paper: published online December 18, 2009; DOI: 10.3171/2009.11.JNS09857.

  • 1

    Baxt WG: Application of artificial neural networks to clinical medicine. Lancet 346:11351138, 1995

  • 2

    Bongard J, & Lipson H: Nonlinear system identification using coevolution of models and tests. IEEE Trans Evol Comput 9:361384, 2005

  • 3

    Chesnut RM, , Marshall SB, , Piek J, , Blunt BA, , Klauber MR, & Marshall LF: Early and late systemic hypotension as a frequent and fundamental source of cerebral ischemia following severe brain injury in the Traumatic Coma Data Bank. Acta Neurochir Suppl (Wien) 59:121125, 1993

    • Search Google Scholar
    • Export Citation
  • 4

    Cross SS, , Harrison RF, & Kennedy RL: Introduction to neural networks. Lancet 346:10751079, 1995

  • 5

    Eftekhar B, , Mohammad K, , Ardebili HE, , Ghodsi M, & Ketabchi E: Comparison of artificial neural network and logistic regression models for prediction of mortality in head trauma based on initial clinical data. BMC Med Inform Decis Mak 5:3, 2005. (Abstract)

    • Search Google Scholar
    • Export Citation
  • 6

    Farace E, & Alves WM: Do women fare worse: a metaanalysis of gender differences in traumatic brain injury outcome. J Neurosurg 93:539545, 2000

    • Search Google Scholar
    • Export Citation
  • 7

    Genkin A, , Lewis D, & Madigan D: Large-scale Bayesian logistic regression for text categorization. Technometrics 49:291 304, 2007

  • 8

    Hagen MD: Test Characteristics. How good is that test?. Prim Care 22:213233, 1995

  • 9

    Hanley JA, & McNeil BJ: The meaning and use of the area under a receiver operating characteristic (ROC) curve. Radiology 143:2936, 1982

    • Search Google Scholar
    • Export Citation
  • 10

    Hsu MH, , Li YC, , Chiu WT, & Yen JC: Outcome prediction after moderate and severe head injury using an artificial neural network. Stud Health Technol Inform 116:241245, 2005

    • Search Google Scholar
    • Export Citation
  • 11

    Hukkelhoven CW, , Steyerberg EW, , Rampen AJ, , Farace E, , Habbema JD, & Marshall LF, : Patient age and outcome following severe traumatic brain injury: an analysis of 5600 patients. J Neurosurg 99:666673, 2003

    • Search Google Scholar
    • Export Citation
  • 12

    Landwehr N, , Hall M, & Frank E: Logistic Model Trees. Machine Learning 59:161205, 2005

  • 13

    Lang EW, , Pitts LH, , Damron SL, & Rutledge R: Outcome after severe head injury: an analysis of prediction based upon comparison of neural network versus logistic regression analysis. Neurol Res 19:274280, 1997

    • Search Google Scholar
    • Export Citation
  • 14

    Lu Z, , Rughani AI, , Tranmer BI, & Bongard J: Informative sampling for large unbalanced data sets. Proceedings of the 2008 GECCO conference companion on Genetic and evolutionary computation New York, ACM Press, 2008. 20472054

    • Search Google Scholar
    • Export Citation
  • 15

    Machado SG, , Murray GD, & Teasdale GM: Evaluation of designs for clinical trials of neuroprotective agents in head injury. J Neurotrauma 16:11311138, 1999

    • Search Google Scholar
    • Export Citation
  • 16

    Pang BC, , Kuralmani V, , Joshi R, , Hongli Y, , Lee KK, & Ang BT, : Hybrid outcome prediction model for severe traumatic brain injury. J Neurotrauma 24:136146, 2007

    • Search Google Scholar
    • Export Citation
  • 17

    Perel P, , Wasserberg J, , Ravi RR, , Shakur H, , Edwards P, & Roberts I: Prognosis following head injury: a survey of doctors from developing and developed countries. J Eval Clin Pract 13:464465, 2007

    • Search Google Scholar
    • Export Citation
  • 18

    Segal ME, , Goodman PH, , Goldstein R, , Hauck W, , Whyte J, & Graham JW, : The accuracy of artificial neural networks in predicting long-term outcome after traumatic brain injury. J Head Trauma Rehabil 21:298314, 2006

    • Search Google Scholar
    • Export Citation
  • 19

    Seung HS, , Opper M, & Sompolinsky H: Query by Committee. Proceedings of the Fifth Annual ACM Conference on Computational Learning Theory New York, ACM Press, 1992. 287294

    • Search Google Scholar
    • Export Citation

Metrics

All Time Past Year Past 30 Days
Abstract Views 664 373 29
Full Text Views 257 47 12
PDF Downloads 179 26 2
EPUB Downloads 0 0 0