<abstract xmlns="http://www.w3.org/1999/xhtml">

<p>The incorporation of sensors onto the stapling platform has been investigated to overcome the disconnect in our understanding of tissue handling by surgical staplers. The goal of this study was to explore the feasibility of <italic>in vivo</italic> porcine tissue differentiation using bioimpedance data and machine learning methods. <italic>In vivo</italic> electrical impedance measurements were obtained in 7 young domestic pigs, using a logarithmic sweep of 50 points over a frequency range of 100 Hz to 1 MHz. Tissues studied included lung, liver, small bowel, colon, and stomach, which was further segmented into fundus, body, and antrum. The data was then parsed through MATLAB's classification learner to identify the best algorithm for tissue type differentiation. The most effective classification scheme was found to be cubic support vector machines with 86.96% accuracy. When fundus, body and antrum were aggregated together as stomach, the accuracy improved to 88.03%. The combination of stomach, small bowel, and colon together as GI tract improved accuracy to 99.79% using fine k nearest neighbors. The results suggest that bioimpedance data can be effectively used to differentiate tissue types <italic>in vivo</italic>. This study is one of the first that combines <italic>in vivo</italic> bioimpedance tissue data across multiple tissue types with machine learning methods.</p>
</abstract>

The incorporation of sensors onto the stapling platform has been investigated to overcome the disconnect in our understanding of tissue handling by surgical staplers. The goal of this study was to explore the feasibility of in vivo porcine tissue differentiation using bioimpedance data and machine learning methods. In vivo electrical impedance measurements were obtained in 7 young domestic pigs, using a logarithmic sweep of 50 points over a frequency range of 100 Hz to 1 MHz. Tissues studied included lung, liver, small bowel, colon, and stomach, which was further segmented into fundus, body, and antrum. The data was then parsed through MATLAB's classification learner to identify the best algorithm for tissue type differentiation. The most effective classification scheme was found to be cubic support vector machines with 86.96% accuracy. When fundus, body and antrum were aggregated together as stomach, the accuracy improved to 88.03%. The combination of stomach, small bowel, and colon together as GI tract improved accuracy to 99.79% using fine k nearest neighbors. The results suggest that bioimpedance data can be effectively used to differentiate tissue types in vivo. This study is one of the first that combines in vivo bioimpedance tissue data across multiple tissue types with machine learning methods.

Electrical impedance characterization of <italic>in vivo</italic> porcine tissue using machine learning

University at Buffalo, The State University of New York

Journal of Electrical Bioimpedance

This work is licensed under the Creative Commons Attribution 4.0 International License.

Electrical impedance characterization of  porcine tissue using machine learning

The incorporation of sensors onto the stapling platform has been investigated to overcome the disconnect in our understanding of tissue handling by surgical...

Electrical impedance characterization of in vivo porcine tissue using machine learning

Pubblicato online: 02 lug 2021

Pagine: 26 - 33

Ricevuto: 21 mar 2021

DOI: https://doi.org/10.2478/joeb-2021-0005

Parole chiave
Bioimpedance, surgical staplers, tissue characterization

© 2020 Stephen Chiang, Matthew Eschbach, Robert Knapp, Brian Holden, Andrew Miesse, Steven Schwaitzberg, and Albert Titus, published by Sciendo

This work is licensed under the Creative Commons Attribution 4.0 International License.

Electrical impedance characterization of in vivo porcine tissue using machine learning

Pubblicato online: 02 lug 2021

Pagine: 26 - 33

Ricevuto: 21 mar 2021

DOI: https://doi.org/10.2478/joeb-2021-0005

Parole chiaveBioimpedance, surgical staplers, tissue characterization

© 2020 Stephen Chiang, Matthew Eschbach, Robert Knapp, Brian Holden, Andrew Miesse, Steven Schwaitzberg, and Albert Titus, published by Sciendo

This work is licensed under the Creative Commons Attribution 4.0 International License.

Parole chiave
Bioimpedance, surgical staplers, tissue characterization