Abstract

Traumatic brain injury (TBI) affects how the brain functions in the short and long term. Resulting patient outcomes across physical, cognitive, and psychological domains are complex and often difficult to predict. Major challenges to developing personalized treatment for TBI include distilling large quantities of complex data and increasing the precision with which patient outcome prediction (prognoses) can be rendered. We developed and applied interpretable machine learning methods to TBI patient data. We show that complex data describing TBI patients' intake characteristics and outcome phenotypes can be distilled to smaller sets of clinically interpretable latent factors. We demonstrate that 19 clusters of TBI outcomes can be predicted from intake data, a ~ 6× improvement in precision over clinical standards. Finally, we show that 36% of the outcome variance across patients can be predicted. These results demonstrate the importance of interpretable machine learning applied to deeply characterized patients for data-driven distillation and precision prognosis.

Details

Title
Data-driven distillation and precision prognosis in traumatic brain injury with interpretable machine learning
Author
Tritt, Andrew 1 ; Yue, John K. 2 ; Ferguson, Adam R. 3 ; Torres Espin, Abel 2 ; Nelson, Lindsay D. 4 ; Yuh, Esther L. 2 ; Markowitz, Amy J. 2 ; Manley, Geoffrey T. 5 ; Bouchard, Kristofer E. 6 ; Keene, C. Dirk 7 ; Madden, Christopher 8 ; McCrea, Michael 9 ; Merchant, Randall 10 ; Mukherjee, Pratik 11 ; Ngwenya, Laura B. 12 ; Robertson, Claudia 13 ; Schnyer, David 14 ; Taylor, Sabrina R. 11 ; Zafonte, Ross 15 

 Lawrence Berkeley National Laboratory, Applied Math and Computational Research Division, Berkeley, USA (GRID:grid.184769.5) (ISNI:0000 0001 2231 4551) 
 Zuckerberg San Francisco General Hospital and Trauma Center, Brain and Spinal Injury Center, San Francisco, USA (GRID:grid.416732.5) (ISNI:0000 0001 2348 2960); University of California San Francisco, Department of Neurosurgery, Weill Institute for Neurosciences, San Francisco, USA (GRID:grid.266102.1) (ISNI:0000 0001 2297 6811) 
 Zuckerberg San Francisco General Hospital and Trauma Center, Brain and Spinal Injury Center, San Francisco, USA (GRID:grid.416732.5) (ISNI:0000 0001 2348 2960); University of California San Francisco, Department of Neurosurgery, Weill Institute for Neurosciences, San Francisco, USA (GRID:grid.266102.1) (ISNI:0000 0001 2297 6811); San Francisco Veterans Affairs Healthcare System, San Francisco, USA (GRID:grid.266102.1) 
 Medical College of Wisconsin, Departments of Neurosurgery and Neurology, Milwaukee, USA (GRID:grid.30760.32) (ISNI:0000 0001 2111 8460) 
 Zuckerberg San Francisco General Hospital and Trauma Center, Brain and Spinal Injury Center, San Francisco, USA (GRID:grid.416732.5) (ISNI:0000 0001 2348 2960); University of California San Francisco, Department of Neurosurgery, Weill Institute for Neurosciences, San Francisco, USA (GRID:grid.266102.1) (ISNI:0000 0001 2297 6811); University of California San Francisco, Weill Neurohub, San Francisco, USA (GRID:grid.266102.1) (ISNI:0000 0001 2297 6811); University of California Berkeley, Weill Neurohub, Berkeley, USA (GRID:grid.47840.3f) (ISNI:0000 0001 2181 7878) 
 University of California Berkeley, Weill Neurohub, Berkeley, USA (GRID:grid.47840.3f) (ISNI:0000 0001 2181 7878); Lawrence Berkeley National Laboratory, Scientific Data Division, Berkeley, USA (GRID:grid.184769.5) (ISNI:0000 0001 2231 4551); Lawrence Berkeley National Laboratory, Biological Systems and Engineering Division, Berkeley, USA (GRID:grid.184769.5) (ISNI:0000 0001 2231 4551); University of California Berkeley, Helen Wills Neuroscience Institute and Redwood Center for Theoretical Neuroscience, Berkeley, USA (GRID:grid.47840.3f) (ISNI:0000 0001 2181 7878) 
 University of Washington, Seattle, USA (GRID:grid.34477.33) (ISNI:0000 0001 2298 6657) 
 UT Southwestern, Dallas, USA (GRID:grid.267313.2) (ISNI:0000 0000 9482 7121) 
 Medical College of Wisconsin, Milwaukee, USA (GRID:grid.30760.32) (ISNI:0000 0001 2111 8460) 
10  Virginia Commonwealth University, Richmond, USA (GRID:grid.224260.0) (ISNI:0000 0004 0458 8737) 
11  University of California, San Francisco, San Francisco, USA (GRID:grid.468726.9) (ISNI:0000 0004 0486 2046) 
12  University of Cincinnati, Cincinnati, USA (GRID:grid.24827.3b) (ISNI:0000 0001 2179 9593) 
13  Baylor College of Medicine, Houston, USA (GRID:grid.39382.33) (ISNI:0000 0001 2160 926X) 
14  UT Austin, Austin, USA (GRID:grid.89336.37) (ISNI:0000 0004 1936 9924) 
15  Harvard Medical School, Boston, USA (GRID:grid.38142.3c) (ISNI:000000041936754X) 
Pages
21200
Publication year
2023
Publication date
2023
Publisher
Nature Publishing Group
e-ISSN
20452322
Source type
Scholarly Journal
Language of publication
English
ProQuest document ID
2896087308
Copyright
© The Author(s) 2023. This work is published under http://creativecommons.org/licenses/by/4.0/ (the “License”). Notwithstanding the ProQuest Terms and Conditions, you may use this content in accordance with the terms of the License.