EUVIMEDEUVIMED entdecken
Crossref

Natural Language Processing and Machine Learning Classification Model for Injury Mechanism in Trauma

Wang Pong Chan, Sophia M Smith, Dane R Scantling, Crisanto M Torres, Megan G Janeway, Yorghos Tripodis, Jacob E Nudel, Sabrina E Sanchez

Journal of Surgical Research · 2026-04 · 320 · 70-76

DOI: 10.1016/j.jss.2026.01.032 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Journal of Surgical Research, 2026-04. Verfasst von Wang Pong Chan und 7 weiteren Autor:innen. Beteiligte Einrichtung: Boston University Chobanian & Avedisian School of Medicine, Boston, Massachusetts; Department of Biostatistics, Boston University School of Public He…. Hauptthemen laut MEDLINE: Natural Language Processing, Machine Learning, Wounds, Penetrating, Wounds, Nonpenetrating. Die Arbeit nennt 41 Literaturangaben.

AbstractSchlagwörterBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

INTRODUCTION Creating and maintaining research databases in trauma can be resource intensive. Natural language processing (NLP) may assist by extracting structured data from free text. We aimed to develop an NLP machine-learning algorithm for classifying traumatic injury mechanism using clinical notes. MATERIALS AND METHODS History and physical and discharge summary documentation from trauma patients admitted to a level I trauma center (January 1, 2018 - December 31, 2021) were collected. A 60-20-20 train-tune-test split was used. The primary outcome was injury mechanism (penetrating versus nonpenetrating), with ground truth established by trained abstractors. Bag-of-words and term frequency-inverse document frequency were used to create a weighted corpus. Dimensions were reduced using singular value decomposition. Random forest, support vector machine, and logistic regression were employed. Primary evaluation metrics were area under the receiver operating characteristic curve and accuracy. RESULTS 6058 clinical notes from 3029 trauma patients were included. Twenty percent (n = 613) had penetrating injuries, and 48% were due to gunshot wounds (n = 294). Models were constructed with 4944 unique terms, forming 30 topics. Area under the receiver operating characteristic curve in the test set was 0.99, 1.00, and 1.00 for the random forest, support vector machine, and logistic regression models. Accuracy was 0.98, 0.98, and 0.98 for these models, respectively. The run time for training and classification was 234 and 0.15 s, respectively. CONCLUSIONS A simple NLP machine-learning algorithm for classifying injury mechanism in trauma patients had near-perfect discrimination and accuracy. NLP may be an accurate and scalable tool to automate aspects of chart review for trauma database development as a supplement to registry workers.

Schlagwörter

MeSH-Schlagwörter (MEDLINE)

Natural Language ProcessingMachine LearningWounds, PenetratingWounds, NonpenetratingWounds and InjuriesHumansClassification AlgorithmsPredictive Learning ModelsRandom ForestFemaleTrauma CentersAdultMaleSupport Vector MachineMiddle AgedPrediction AlgorithmsLogistic Models

Stichwörter der Autor:innen

Artificial intelligenceDatabaseInjury mechanismNatural language processing

Autor:innen und Institutionen

  • Boston University Chobanian & Avedisian School of Medicine, Boston, Massachusetts; Department of Biostatistics, Boston University School of Public Health, Boston, Massachusetts.
  • Department of Surgery, Boston Medical Center, Boston, Massachusetts.
  • Boston University Chobanian & Avedisian School of Medicine, Boston, Massachusetts; Department of Surgery, Boston Medical Center, Boston, Massachusetts.
  • Department of Biostatistics, Boston University School of Public Health, Boston, Massachusetts.
  • Boston University Chobanian & Avedisian School of Medicine, Boston, Massachusetts; Department of Surgery, Boston Medical Center, Boston, Massachusetts. Electronic address: sabrina.sanchez@bmc.org.

Zitieren

APA

Chan, W. P., Smith, S. M., Scantling, D. R., Torres, C. M., Janeway, M. G., Tripodis, Y., Nudel, J. E., & Sanchez, S. E. (2026). Natural Language Processing and Machine Learning Classification Model for Injury Mechanism in Trauma. Journal of Surgical Research, 320, 70-76. https://doi.org/10.1016/j.jss.2026.01.032

Vancouver

Chan WP, Smith SM, Scantling DR, Torres CM, Janeway MG, Tripodis Y, et al. Natural Language Processing and Machine Learning Classification Model for Injury Mechanism in Trauma. Journal of Surgical Research. 2026;320:70-76. doi:10.1016/j.jss.2026.01.032

BibTeX
@article{Chan2026,
  author = {Chan, Wang Pong and Smith, Sophia M and Scantling, Dane R and Torres, Crisanto M and Janeway, Megan G and Tripodis, Yorghos and Nudel, Jacob E and Sanchez, Sabrina E},
  title = {Natural Language Processing and Machine Learning Classification Model for Injury Mechanism in Trauma},
  journal = {Journal of Surgical Research},
  year = {2026},
  volume = {320},
  pages = {70--76},
  doi = {10.1016/j.jss.2026.01.032}
}
Kostenlos starten →