Skip to main page content
U.S. flag

An official website of the United States government

Dot gov

The .gov means it’s official.
Federal government websites often end in .gov or .mil. Before sharing sensitive information, make sure you’re on a federal government site.

Https

The site is secure.
The https:// ensures that you are connecting to the official website and that any information you provide is encrypted and transmitted securely.

Access keys NCBI Homepage MyNCBI Homepage Main Content Main Navigation
. 2022 Sep 22:3:1005168.
doi: 10.3389/fresc.2022.1005168. eCollection 2022.

Machine learning predicts improvement of functional outcomes in traumatic brain injury patients after inpatient rehabilitation

Irene Say  1 , Yiling Elaine Chen  2 , Matthew Z Sun  1 , Jingyi Jessica Li  2 , Daniel C Lu  1   3   4
Affiliations

Machine learning predicts improvement of functional outcomes in traumatic brain injury patients after inpatient rehabilitation

Irene Say et al. Front Rehabil Sci. .

Abstract

Survivors of traumatic brain injury (TBI) have an unpredictable clinical course. This unpredictability makes clinical resource allocation for clinicians and anticipatory guidance for patients difficult. Historically, experienced clinicians and traditional statistical models have insufficiently considered all available clinical information to predict functional outcomes for a TBI patient. Here, we harness artificial intelligence and apply machine learning and statistical models to predict the Functional Independence Measure (FIM) scores after rehabilitation for traumatic brain injury (TBI) patients. Tree-based algorithmic analysis of 629 TBI patients admitted to a large acute rehabilitation facility showed statistically significant improvement in motor and cognitive FIM scores at discharge.

Keywords: artificial intelligence; functional independence measure; machine learning; prediction model; rehabilitation; traumatic brain injury.

PubMed Disclaimer

Conflict of interest statement

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

Figures

Figure 1
Figure 1
Graphical illustration of tree-based algorithms such as random forests and XGBoost. Both algorithms build multiple trees for prediction. Random forests build independent trees whereas XGBoost builds trees based on the performance of previous trees.
Figure 2
Figure 2
Graphical illustration of ordinal regression. Suppose the class labels range from 1 to 7. The ordinal regression trains 6 logistic models to predict the probability of Y ≤ j, j = 1,…,6. The parallel ordinal regression model assumes that these 6 sub-models share the same set of coefficients, whereas the semi-parallel model forces the (K-1) sets of coefficients to be similar and close to 0.
Figure 3
Figure 3
Bar graph shows p-values comparing FIM scores at discharge with FIM scores at admission. P-values are adjusted for multiple correction. All p-values are smaller than 0.05, showing improvement in all 18 Functional Independence Measurement (FIM) items upon discharge in TBI patients compared to admission to acute rehabilitation.
Figure 4
Figure 4
Heatmap showing the error rates of machine learning prediction algorithms across the 18 Functional Independence Measurement (FIM) scores in TBI patients. Statistical algorithms included parallel and semi-parallel ordinal regression with lasso, ridge, or elastic net penalty, random forest, and XGBoost.
Figure 5
Figure 5
(A) Bar and whisker plot showing superior accuracy of tree-based algorithms, in random forest (rf) and XGBoost with controlled algorithm L1 loss compared to parallel and semi-parallel ordinal regression. (B) Bar graphs for individual Functional Independence Measure (FIM) items illustrating tree-based algorithms random forest (rf) and XGBoost with the lowest error.
See this image and copyright information in PMC

Similar articles

See all similar articles

Cited by

References

    1. Centers for Disease Control and Prevention. Report to Congress on Mild Traumatic Brain Injury in the United States: Steps to Prevent a Serious Public Health Problem (2014).
    1. Center for Disease Control and Prevention. Report to Congress on Traumatic Brain Injury in the United States: Epidemiology and Rehabilitation. National Center for Injury Prevention and Control. (2015).
    1. Malec J, Hammond F, Dams-O’Connor K. Chronic disease management for brain injury. In: Silver J, McAllister T, Arciniegas D, editors. Textbook of traumatic brain injury. Washington, DC USA: American Psychiatric Association; (2018). p. 733–45. 10.1176/appi.books.9781615372645.js40 - DOI
    1. Wilson L, Stewart W, Dams-O’Connor K, Diaz-Arrastia R, Horton L, Menon DK, et al. The chronic and evolving neurological consequences of traumatic brain injury. Lancet Neurol. (2017) 16(10):813–25. 10.1016/S1474-4422(17)30279-X - DOI - PMC - PubMed
    1. Brooks JC, Shavelle RM, Strauss DJ, Hammond FM, Harrison-Felix CL. Life expectancy of 1-year survivors of traumatic brain injury, 1988–2019: updated results from the TBI model systems. Arch Phys Med Rehabil. (2022) 103(1):176–9. 10.1016/j.apmr.2021.05.015 - DOI - PubMed