Exploring Deep Clustering Methods in Vibro-Acoustic Sensing for Enhancing Biological Tissue Characterization

Nonlinear dimensionality reduction techniques, often referred to as manifold learning, are increasingly valuable for data visualization and unsupervised clustering. In the context of surgery and medicine, these methods facilitate the analysis of complex datasets, enabling pattern recognition in surg...

Full description

Saved in:
Bibliographic Details
Main Authors: Robin Urrutia, Diego Espejo, Montserrat Guerra, Karin Vio, Thomas Suhn, Nazila Esmaeili, Axel Boese, Patricio Fuentealba, Alfredo Illanes, Christian Hansen, Victor Poblete
Format: Article
Language:English
Published: IEEE 2025-01-01
Series:IEEE Access
Subjects:
Online Access:https://ieeexplore.ieee.org/document/10981752/
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1850278761977610240
author Robin Urrutia
Diego Espejo
Montserrat Guerra
Karin Vio
Thomas Suhn
Nazila Esmaeili
Axel Boese
Patricio Fuentealba
Alfredo Illanes
Christian Hansen
Victor Poblete
author_facet Robin Urrutia
Diego Espejo
Montserrat Guerra
Karin Vio
Thomas Suhn
Nazila Esmaeili
Axel Boese
Patricio Fuentealba
Alfredo Illanes
Christian Hansen
Victor Poblete
author_sort Robin Urrutia
collection DOAJ
description Nonlinear dimensionality reduction techniques, often referred to as manifold learning, are increasingly valuable for data visualization and unsupervised clustering. In the context of surgery and medicine, these methods facilitate the analysis of complex datasets, enabling pattern recognition in surgical data. This study explores the characterization of six tissue types through manifold learning and unsupervised clustering, utilizing vibro-acoustic (VA) signals collected from manual palpation experiments. A wireless sensor mounted at the tip of a surgical instrument was used to acquire 1,680 VA signals, which were processed using Fourier transform and cepstral analysis for feature extraction. We assessed the performance of two dimensionality reduction techniques: uniform manifold approximation and projection (UMAP) and variational autoencoder (VAE). Results indicate that cepstral features combined with UMAP yield superior clustering performance compared to VAE, achieving higher classification accuracy (<inline-formula> <tex-math notation="LaTeX">$92~\%$ </tex-math></inline-formula> vs. <inline-formula> <tex-math notation="LaTeX">$87~\%$ </tex-math></inline-formula>) and better-defined clusters with greater compactness. The observed differences in performance are linked to the intrinsic properties of the tissues, particularly surface characteristics such as friction and moisture, which affect signal consistency. Additionally, we compared our approach with previous works, including a study utilizing the same dataset, where our methodology demonstrated improved accuracy. Future research will focus on refining the VAE model, increasing the diversity of tissue samples, and validating the proposed approach in real surgical settings to enhance its applicability in minimally invasive surgery.
format Article
id doaj-art-2a365fd8291946dfb7a6cd360e6aa738
institution OA Journals
issn 2169-3536
language English
publishDate 2025-01-01
publisher IEEE
record_format Article
series IEEE Access
spelling doaj-art-2a365fd8291946dfb7a6cd360e6aa7382025-08-20T01:49:22ZengIEEEIEEE Access2169-35362025-01-0113803958040610.1109/ACCESS.2025.356628010981752Exploring Deep Clustering Methods in Vibro-Acoustic Sensing for Enhancing Biological Tissue CharacterizationRobin Urrutia0https://orcid.org/0009-0003-9430-3474Diego Espejo1https://orcid.org/0000-0001-8083-538XMontserrat Guerra2Karin Vio3Thomas Suhn4https://orcid.org/0000-0001-5132-2884Nazila Esmaeili5https://orcid.org/0000-0001-9741-9788Axel Boese6https://orcid.org/0000-0002-5874-7145Patricio Fuentealba7https://orcid.org/0000-0002-7119-0580Alfredo Illanes8https://orcid.org/0000-0002-0118-0483Christian Hansen9https://orcid.org/0000-0002-5734-7529Victor Poblete10https://orcid.org/0000-0003-2894-0267Faculty of Computer Science and Research Campus STIMULATE, University of Magdeburg, Magdeburg, GermanyAudio Mining Laboratory (AuMiLab), Science Engineering Faculty, Austral University of Chile, Valdivia, ChileInstitute of Anatomy, Histology and Pathology, Faculty of Medicine, Austral University of Chile, Valdivia, ChileInstitute of Anatomy, Histology and Pathology, Faculty of Medicine, Austral University of Chile, Valdivia, ChileSURAG Medical GmbH, Leipzig, GermanyDepartment of General, Visceral and Pediatric Surgery, Center for Digital Surgery, University Medical Center G&#x00F6;ttingen, G&#x00F6;ttingen, GermanyINKA Innovation Laboratory for Image Guided Therapy, Otto von Guericke University Magdeburg, Magdeburg, GermanyInstitute of Electricity and Electronics, Science Engineering Faculty, Austral University of Chile, Valdivia, ChileFaculty of Computer Science and Research Campus STIMULATE, University of Magdeburg, Magdeburg, GermanyFaculty of Computer Science and Research Campus STIMULATE, University of Magdeburg, Magdeburg, GermanyAudio Mining Laboratory (AuMiLab), Science Engineering Faculty, Austral University of Chile, Valdivia, ChileNonlinear dimensionality reduction techniques, often referred to as manifold learning, are increasingly valuable for data visualization and unsupervised clustering. In the context of surgery and medicine, these methods facilitate the analysis of complex datasets, enabling pattern recognition in surgical data. This study explores the characterization of six tissue types through manifold learning and unsupervised clustering, utilizing vibro-acoustic (VA) signals collected from manual palpation experiments. A wireless sensor mounted at the tip of a surgical instrument was used to acquire 1,680 VA signals, which were processed using Fourier transform and cepstral analysis for feature extraction. We assessed the performance of two dimensionality reduction techniques: uniform manifold approximation and projection (UMAP) and variational autoencoder (VAE). Results indicate that cepstral features combined with UMAP yield superior clustering performance compared to VAE, achieving higher classification accuracy (<inline-formula> <tex-math notation="LaTeX">$92~\%$ </tex-math></inline-formula> vs. <inline-formula> <tex-math notation="LaTeX">$87~\%$ </tex-math></inline-formula>) and better-defined clusters with greater compactness. The observed differences in performance are linked to the intrinsic properties of the tissues, particularly surface characteristics such as friction and moisture, which affect signal consistency. Additionally, we compared our approach with previous works, including a study utilizing the same dataset, where our methodology demonstrated improved accuracy. Future research will focus on refining the VAE model, increasing the diversity of tissue samples, and validating the proposed approach in real surgical settings to enhance its applicability in minimally invasive surgery.https://ieeexplore.ieee.org/document/10981752/Dimensionality reductionhaptic informationminimally invasive surgeryrobot-assisted surgerysurgery augmentationsurgical data science
spellingShingle Robin Urrutia
Diego Espejo
Montserrat Guerra
Karin Vio
Thomas Suhn
Nazila Esmaeili
Axel Boese
Patricio Fuentealba
Alfredo Illanes
Christian Hansen
Victor Poblete
Exploring Deep Clustering Methods in Vibro-Acoustic Sensing for Enhancing Biological Tissue Characterization
IEEE Access
Dimensionality reduction
haptic information
minimally invasive surgery
robot-assisted surgery
surgery augmentation
surgical data science
title Exploring Deep Clustering Methods in Vibro-Acoustic Sensing for Enhancing Biological Tissue Characterization
title_full Exploring Deep Clustering Methods in Vibro-Acoustic Sensing for Enhancing Biological Tissue Characterization
title_fullStr Exploring Deep Clustering Methods in Vibro-Acoustic Sensing for Enhancing Biological Tissue Characterization
title_full_unstemmed Exploring Deep Clustering Methods in Vibro-Acoustic Sensing for Enhancing Biological Tissue Characterization
title_short Exploring Deep Clustering Methods in Vibro-Acoustic Sensing for Enhancing Biological Tissue Characterization
title_sort exploring deep clustering methods in vibro acoustic sensing for enhancing biological tissue characterization
topic Dimensionality reduction
haptic information
minimally invasive surgery
robot-assisted surgery
surgery augmentation
surgical data science
url https://ieeexplore.ieee.org/document/10981752/
work_keys_str_mv AT robinurrutia exploringdeepclusteringmethodsinvibroacousticsensingforenhancingbiologicaltissuecharacterization
AT diegoespejo exploringdeepclusteringmethodsinvibroacousticsensingforenhancingbiologicaltissuecharacterization
AT montserratguerra exploringdeepclusteringmethodsinvibroacousticsensingforenhancingbiologicaltissuecharacterization
AT karinvio exploringdeepclusteringmethodsinvibroacousticsensingforenhancingbiologicaltissuecharacterization
AT thomassuhn exploringdeepclusteringmethodsinvibroacousticsensingforenhancingbiologicaltissuecharacterization
AT nazilaesmaeili exploringdeepclusteringmethodsinvibroacousticsensingforenhancingbiologicaltissuecharacterization
AT axelboese exploringdeepclusteringmethodsinvibroacousticsensingforenhancingbiologicaltissuecharacterization
AT patriciofuentealba exploringdeepclusteringmethodsinvibroacousticsensingforenhancingbiologicaltissuecharacterization
AT alfredoillanes exploringdeepclusteringmethodsinvibroacousticsensingforenhancingbiologicaltissuecharacterization
AT christianhansen exploringdeepclusteringmethodsinvibroacousticsensingforenhancingbiologicaltissuecharacterization
AT victorpoblete exploringdeepclusteringmethodsinvibroacousticsensingforenhancingbiologicaltissuecharacterization