Predicting shock-induced cavitation using machine learning - implications for blast-injury models

Aug 2, 2026

Machine learning is evaluated as a practical alternative for studying blast-induced traumatic brain injury where in-vivo cavitation measurements remain limited. The work compares k-nearest neighbor and support vector machine models to predict shock-induced cavitation behavior using experimental results from a three-dimensional shock tube setup. Models are trained and validated on experimental data to estimate the number of cavitation bubbles at a given temperature with good accuracy. The findings support machine learning utility for blast-injury research and for bridging experiments and numerical simulations.

LanguageEnglish

CategoryResearch & Report

FormatPDF

Pages19

Reading Time48 min

AccessFree to download

Introduction

Blast-induced traumatic brain injury and cavitation background

Limits of experimental measurement and need for validation

Rationale for machine learning

Methods

Machine learning models (kNN and SVM)

Training and validation data from a 3D shock tube

Results and Discussion

Prediction accuracy for bubble number versus temperature

Implications for blast-injury modeling

TYPE Original Research PUBLISHED 05 February 2024 DOI 10.3389/fbioe.2024.1268314 OPEN ACCESS EDITED BY Ould El Moctar, University of Duisburg-Essen, Germany REVIEWED BY Yirui Sun, Fudan University, China Tijana Geroski, University of Kragujevac, Serbia *CORRESPONDENCE Jenny L. Marsh, [jenny@iastate.edu](jenny@iastate.edu) RECEIVED 27 July 2023 ACCEPTED 16 January 2024 PUBLISHED 05 February 2024 CITATION Marsh JL, Zinnel L and Bentil SA (2024), Predicting shock-induced cavitation using machine learning: implications for blastinjury models. Front. Bioeng. Biotechnol. 12:1268314 . doi: 10.3389/fbioe.2024.1268314 COPYRIGHT © 2024 Marsh, Zinnel and Bentil. This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY) . The use, distribution or reproduction in other forums is permitted, provided the original author(s) and the copyright owner(s) are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms. Predicting shock-induced cavitation using machine learning: implications for blast-injury models Jenny L. Marsh 1*, Laura Zinnel 1,2 and Sarah A. Bentil 1 1Department of Mechanical Engineering, The Bentil Group, Iowa State University, Ames, IA, United States, 2Department of Mathematics, Iowa State University, Ames, IA, United States While cavitation has been suspected as a mechanism of blast-induced traumatic brain injury (bTBI) for a number of years, this phenomenon remains difﬁcult to study due to the current inability to measure cavitation in vivo. Therefore, numerical simulations are often implemented to study cavitation in the brain and surrounding ﬂuids after blast exposure. However, these simulations need tobe validated with the results from cavitation experiments. Machine learning algorithms have not generally been applied to study blast injury or biological cavitation models. However, such algorithms have concrete measures for optimization using fewer parameters than those of ﬁnite element or ﬂuid dynamics models. Thus, machine learning algorithms are a viable option for predicting cavitation behavior from experiments and numerical simulations. This paper compares the ability of two machine learning algorithms, k-nearest neighbor (kNN) and support vector machine (SVM), to predict shock-induced cavitation behavior. The machine learning models were trained and validated with experimental data from a three-dimensional shock tube model, and it has been shown that the algorithms could predict the number of cavitation bubbles produced at a given temperature with good accuracy. This study demonstrates the potential utility of machine learning in studying shockinduced cavitation for applications in blast injury research. KEYWORDS machine learning, cavitation, support vector machines, k-nearest neighbors, traumatic brain injury, shock tube 1 Introduction Blast-induced traumatic brain injury (bTBI) represents over 66% of injuries sustained by deployed U.S. military service members (Regasa et al., 2019) . From 2000 to the third quarter of 2022, the Department of Defense reported 486,424 traumatic brain injuries, with 387,456 of those attributed to bTBI from active deployments (DOD Worldwide TBI Numbers, 2023) . bTBI is not limited to military service members but may also impact civilians in war zones or in industrial explosions. Symptoms of bTBI include visual dysfunction, headaches, balance, and impulse control impairment (Capó-Aponte et al., 2012; Bryden et al., 2019) . Blast injury is also associated with an increased probability and severity of post-traumatic stress disorder (PTSD) and increased chances of developing neurodegenerative disorders (Barker et al., 2023; Borinuoluwa and Ahmed, 2023) . Diagnostics, treatment, and prevention of bTBI are dependent on an understanding of t

FAQs

Why is cavitation difficult to study in blast-induced traumatic brain injury?

Cavitation remains difficult to study in vivo because current approaches cannot directly measure cavitation in living tissue after blast exposure.

Which machine learning algorithms are compared in the study?

The study compares k-nearest neighbor (kNN) and support vector machine (SVM) to predict shock-induced cavitation behavior.

How are the machine learning models trained and validated?

The models are trained and validated using experimental data from a three-dimensional shock tube model, then used to predict cavitation bubble counts at given temperatures with good accuracy.

Aug 2, 2026

Upload by Emma Wilson