Jump to ContentJump to Main Navigation
Show Summary Details
More options …

Current Directions in Biomedical Engineering

Joint Journal of the German Society for Biomedical Engineering in VDE and the Austrian and Swiss Societies for Biomedical Engineering

Editor-in-Chief: Dössel, Olaf

Editorial Board: Augat, Peter / Buzug, Thorsten M. / Haueisen, Jens / Jockenhoevel, Stefan / Knaup-Gregori, Petra / Kraft, Marc / Lenarz, Thomas / Leonhardt, Steffen / Malberg, Hagen / Penzel, Thomas / Plank, Gernot / Radermacher, Klaus M. / Schkommodau, Erik / Stieglitz, Thomas / Urban, Gerald A.


CiteScore 2018: 0.47

Source Normalized Impact per Paper (SNIP) 2018: 0.377

Open Access
Online
ISSN
2364-5504
See all formats and pricing
More options …

Modeling the pelvic region for non-invasive pelvic intraoperative neuromonitoring

Tomasz Moszkowski
  • Corresponding author
  • AGH-University of Science and Technology, al. Mickiewicza 30, 30-059 Kraków, Poland, inomed Medizintechnik GmbH, 79312 Emmendingen, Germany
  • Email
  • Other articles by this author:
  • De Gruyter OnlineGoogle Scholar
/ Thilo Krüger / Werner Kneist
  • Department of General, Visceral and Transplant Surgery, University Medicine of the Johannes Gutenberg-University Mainz, Langenbeckstraße 1,55131 Mainz, Germany
  • Email
  • Other articles by this author:
  • De Gruyter OnlineGoogle Scholar
/ Klaus-Peter Hoffmann
Published Online: 2016-09-30 | DOI: https://doi.org/10.1515/cdbme-2016-0042

Abstract

Finite element analysis (FEA) of electric current distribution in the pelvis minor may help to assess the usability of non-invasive surface stimulation for continuous pelvic intraoperative neuromonitoring. FEA requires generation of quality volumetric tetrahedral mesh geometry. This study proposes the generation of a suitable mesh based on MRI data. The resulting volumetric mesh models the autonomous nerve structures at risk during total mesorectal excision. The model also contains the bone, cartilage, fat, skin, muscle tissues of the pelvic region, and a set of electrodes for surface stimulation. The model is ready for finite element analysis of the discrete Maxwell’s equations.

Keywords: electric field modeling; finite element method; pelvic intraoperative neurophysiological monitoring

1 Introduction

Rectal cancer surgery may result in anorectal and urogenital functional deficits which drastically diminish the quality of life of patients [1]. These complications may result from damaging the pelvic autonomous nervous system [2].

Pelvic intraoperative neuromonitoring (pIONM) may help preserve the function of the pelvic autonomous nervous system during rectal cancer surgery. Two-dimensional pIONM could predict urinary and anorectal function after low anterior rectal resection [3] and intraoperative neuromonitoring was associated with preserving the urinary [4] and urogenital [5] function in patients undergoing open total mesorectal excision (TME).

pIONM relies on injecting electrical current in-situ into the autonomous nervous tissue which is at risk of damage during TME. In pIONM, electrical stimulation evokes a contraction of m. detrusor vesicae and a modulation of the internal anal sphincter (IAS) activity. Any aberration in these responses can warn the surgeon of possible nerve damage.

The surgeon needs to introduce the stimulation probe in situ in order to perform neuromonitoring. This procedure prolongs the surgery because pIONM can nowadays only be performed intermittently in the crucial phases of the surgical procedure. Unobserved nerve damage can occur in between the phases involving neuromapping [6]. Thus, continuous methods of stimulation are desirable.

Transdermal stimulation of the sacral roots of the aforementioned autonomous nervous system might offer several advantages compared to in-situ stimulation. Electrodes could be permanently placed and continuously stimulate throughout the surgery. The surgeon could perform the standard surgery in parallel to the monitoring. This could facilitate near real-time reaction to possible nerve damage and increased practicability and usage of pIONM.

During surface stimulation, the contact area of the electrodes and the distance to the target autonomous neural structures are bigger than during in-situ stimulation, which casts doubt on the selectivity of stimulation. Knowing the distribution of current due to surface stimulation could help assess its usability in pIONM. Because direct measurement of the distribution of current in tissue is problematic, the chosen approach is the use of numerical modeling of the electric field by means of the finite element method (FEM) based on MRI data [7].

The aim of this contribution is to describe the preparation of a quality three-dimensional volumetric mesh to gain insight into the usability of surface stimulation in continuous pIONM.

2 Material and methods

In order to process the surface meshes of tissues obtained from MRI-data (the Duke model, the Virtual Population V2.0 [8]) and generate a volumetric mesh suitable for FEM, we used Octave open-source software (GNU) and the problem solving environment SCIRun with BioMesh3D (CIBC, University of Utah, USA).

3 Results

We separately modelled the autonomous nerves controlling the function of the urinary bladder, the external and internal anal sphincters (IASs) because the Virtual Population 2.0 did not contain them. Based on [9], [10] and clinical experience, we created a 3D model of the autonomous innervation of the IAS and the bladder (Figure 1). To obtain a model of each nerve, we created a set of Bézier curves and extruded a 3-mm circle normal to each curve (Figures 1 and 2).

Each nerve was created using a three-dimensional Bézier curve (A). To obtain the nerve volume, a circle (3 mm in diameter) was extruded along each Bëzier curve (B). The resulting nerve model (C) comprised of the superior hypogastric plexus (PHS), two hypogastric nerves (HN), two inferior hypogastric plexi (PHI), two pudendal nerves (PN), pelvic splanchnic nerves (NSP), branches of bladder, internal (IAS) and external anal sphincter (EAS) innervation. The description of the nerve structures in (C) was done asymmetrically to maintain the image clarity.
Figure 1:

Each nerve was created using a three-dimensional Bézier curve (A). To obtain the nerve volume, a circle (3 mm in diameter) was extruded along each Bëzier curve (B). The resulting nerve model (C) comprised of the superior hypogastric plexus (PHS), two hypogastric nerves (HN), two inferior hypogastric plexi (PHI), two pudendal nerves (PN), pelvic splanchnic nerves (NSP), branches of bladder, internal (IAS) and external anal sphincter (EAS) innervation. The description of the nerve structures in (C) was done asymmetrically to maintain the image clarity.

The spatial relationship of the modelled autonomous nerve tissue (A) to the pelvis (B), pelvis and bladder (C), and pelvis, bladder and the small and lower intestines (D).
Figure 2:

The spatial relationship of the modelled autonomous nerve tissue (A) to the pelvis (B), pelvis and bladder (C), and pelvis, bladder and the small and lower intestines (D).

We trimmed the resulting surfaces of all tissues to the region of interest of TME surgery [11]. The model spanned from the sigmoid colon to the male reproductive organs.

We subsequently obtained the volume information inside the three-dimensional surface meshes using a ray-tracing algorithm and subdivided the entire volume into voxels (cuboid representations of the volume with 1 mm edge) according to the method described in [12] (Figure 3).

Voxelization” of surface meshes produced segmented matrices, where each segment corresponded to a different tissue group of the model. This example shows the frontal view of the “voxelized” bone tissue. The image is represented as an intensity plot, where each pixel corresponds to the average intensity along a line normal to the view plane.
Figure 3:

Voxelization” of surface meshes produced segmented matrices, where each segment corresponded to a different tissue group of the model. This example shows the frontal view of the “voxelized” bone tissue. The image is represented as an intensity plot, where each pixel corresponds to the average intensity along a line normal to the view plane.

We subdivided the resulting voxelized volume into tetrahedral meshes in BioMesh3D® using Delaunay triangulation (Figure 4) according to [13].

The tissue boundaries are extracted from the voxelized representation of the tissues (A). For each tissue, a medial axis (B) and a distance field is computed (C). Seed points are randomly generated on material boundaries. Using the distance field, the seed points are then reallocated in such a way as to minimize the energy between each seed point (D). The seed points were subsequently used to generate a volumetric mesh.
Figure 4:

The tissue boundaries are extracted from the voxelized representation of the tissues (A). For each tissue, a medial axis (B) and a distance field is computed (C). Seed points are randomly generated on material boundaries. Using the distance field, the seed points are then reallocated in such a way as to minimize the energy between each seed point (D). The seed points were subsequently used to generate a volumetric mesh.

The resulting tetrahedral volumetric mesh consisted of autonomous nerve tissue, subcutaneous and visceral fat, skin, bone, cartilage, gastro-intestinal tract, muscle tissue, urinary bladder and cerebrospinal fluid (Table 1, Figure 5).

Table 1:

Conductivities assigned to the used tissues [14].

A point cloud representation of the volumetric mesh. Each color represents a different tissue group.
Figure 5:

A point cloud representation of the volumetric mesh. Each color represents a different tissue group.

We supplemented the resulting volumetric mesh with electrode pads for the simulation of the sources of the electric field. We created the electrode model prototypes (three rectangular pads with rounded corners and 10 circular electrodes forming a stimulation array) using Blender®. Subsequently, we orthogonally projected the flat electrodes onto the surface of skin of the model in SCIRun® (Figure 6).

(Right) The placement of electrode models onto the surface of the skin of the volumetric model. The bone (white) and skin (gray) surface meshes serve as spatial reference. Separate electrode pads are depicted using a different color.
Figure 6:

(Right) The placement of electrode models onto the surface of the skin of the volumetric model. The bone (white) and skin (gray) surface meshes serve as spatial reference. Separate electrode pads are depicted using a different color.

Next, we increased the density of the mesh near the contact between the tissue model and the modelled electrodes (Figure 7). We saved the resulting field in. fld format which can be directly read by the SCIRun software and is ready to perform the finite element method analysis.

We created a distance field (A) where each vertex of the volumetric mesh represented the distance to the nearest vertex belonging to the electrode pads. We locally refined the volumetric mesh (B) by setting a 1 mm threshold to the distance field.
Figure 7:

We created a distance field (A) where each vertex of the volumetric mesh represented the distance to the nearest vertex belonging to the electrode pads. We locally refined the volumetric mesh (B) by setting a 1 mm threshold to the distance field.

4 Discussion

We created a model of the pelvis minor whose dimensions correspond to the region in question during TME [11]. The model consisted of tissue compartments whose electrical properties can be freely assigned. We supplemented the model with autonomous neural tissue based on [9], [10] because the fine neural fibres were not included in the Virtual Population V2.0. We modelled the neural plexi as intersections of 3-mm tubular nerves. In reality, however, an intricate network of fine nerve fibres with small dimensions forms the autonomous nervous system [11]. Such high complexity would be impractical, hard to interpret, and the resulting change of accuracy would be questionable.

The results of numerical modeling of the nerve activation due to ex-situ electrical stimulation of the cauda equina compared with animal studies [15]. In the case of pIONM, numerical modeling could help assess and optimize the use of surface stimulation in non-invasive continuous monitoring. The proposed model will undergo FEM analysis of the electric field due to surface stimulation. It is crucial to assess the selectivity of ex-situ stimulation of the autonomous nervous system, its sensitivity and specificity to monitoring nerve damage.

A one-size-fits-all model may fail to predict universal guidelines for stimulation because of inter-individual differences of patients’ anatomy. The numerically predicted parameter ranges may need to be adjusted for individual patient. A system for automatic control of stimulation could help individually navigate the stimulation process.

Acknowledgement

We used SCIRun thanks to the National Institute of General Medical Sciences of the National Institutes of Health under grant number P41 GM103545-17.

Author’s Statement

Research funding: This contribution was funded under the autoPIN Project (BMBF, grant number 13GW0022C). Conflict of interest: Authors state no conflict of interest. Material and Methods: Informed consent: Informed consent is not applicable. Ethical approval: The conducted research is not related to either human or animal use.

References

  • [1]

    Lange MM, van de Velde CJ. Urinary and sexual dysfunction after rectal cancer treatment. Nat Rev Urol. 2011;8:51–7. Google Scholar

  • [2]

    Wallner C, Lange MM, Bonsing BA, Maas CP, Wallace CN, Dabhoiwala NF, et al. Causes of fecal and urinary incontinence after total mesorectal excision for rectal cancer based on cadaveric surgery: a study from the Cooperative Clinical Investigators of the Dutch total mesorectal trial. Clin Oncol. 2008;27:4466–72. Google Scholar

  • [3]

    Kauff DW, Koch KP, Somerlik KH, Hoffmann KP, Lang H, Kneist W. Evaluation of two-dimensional intraoperative neuromonitoring for predicting urinary and anorectal function after rectal cancer surgery. Int J Colorectal Dis. 2013;5:659–64. Google Scholar

  • [4]

    Kneist W, Kauff DW, Juhre V, Hoffmann KP, Lang H, et al. Is intraoperative neuromonitoring associated with better functional outcome in patients undergoing open TME?: results of a case-control study. Eur J Surg Oncol. 2013;39:994–9. Google Scholar

  • [5]

    Kauff D, et al. Comparison of urogenital function following TME with and without pelvic intraoperative neuromonitoring, 133. Jahreskongress der DGCH, Berlin, 2016 Google Scholar

  • [6]

    Kauff D, Kempski O, Koch KP, Huppert S, Hoffmann KP, Lang H, et al. Continuous intraoperative monitoring of autonomic nerves during low anterior rectal resection: an innovative approach for observation of functional nerve integrity in pelvic surgery. Langebecks Arch Surg. 2012;397:787–92. Google Scholar

  • [7]

    Zaidi H, Tsui B. Review of computational anthropomorphic anatomical and physiological models. Proc IEEE. 2009;12:1938–53. Google Scholar

  • [8]

    Christ A, Kainz W, Hahn EG, Honegger K, Zefferer M, Neufeld E, et al. The virtual family – development of surface-based anatomical models of two adults and two children for dosimetric simulations. Phys Med Biol. 2009;2:N23–38. Google Scholar

  • [9]

    Sokołowska-Pituchowa J. Anatomia Człowieka. PZWL 2006. Google Scholar

  • [10]

    Aigner F. Clinical anatomy of the pelvic floor. Acta Chir Iugosl. 2006;2:11–3. Google Scholar

  • [11]

    Kneist W, Heintz A, Wolf HK, Junginger T. Darstellung Autonomer Beckennerven bei Partieller und Totaler Mesorektumexzision. Chirurg. 2004;3:276–83. Google Scholar

  • [12]

    Patil S, Ravi B, Voxel-based representation, display and thickness analysis of intricate shapes. CAD/CG. 2005:415–22. Google Scholar

  • [13]

    Available at: http://www.sci.utah.edu/devbuilds/scirun_docs/BioMesh3DGDGui.pdf

  • [14]

    Hasgall PA, et al. IT’IS Database for thermal and electro-magnetic parameters of biological tissues, V3.0, 2015, DOI: . CrossrefGoogle Scholar

  • [15]

    Laakso I, Matsumoto H, Hirata A, Terao Y, Hanajima R, Ugawa Y, et al. Multi-scale simulations predict responses to non-invasive nerve root stimulation. J Neural Eng 2014;11:056013. Google Scholar

About the article

Published Online: 2016-09-30

Published in Print: 2016-09-01


Citation Information: Current Directions in Biomedical Engineering, Volume 2, Issue 1, Pages 185–188, ISSN (Online) 2364-5504, DOI: https://doi.org/10.1515/cdbme-2016-0042.

Export Citation

©2016 Tomasz Moszkowski et al., licensee De Gruyter.. This work is licensed under the Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 License. BY-NC-ND 4.0

Comments (0)

Please log in or register to comment.
Log in