Surgical Endoscopy

, Volume 22, Issue 8, pp 1838–1844

Retraction by surface ferromagnetisation of target tissues: Preliminary studies on feasibility of magnetic retraction for endoscopic surgery


  • Zhigang Wang
    • Institute for Medical Science and Technology (IMSaT), Ninewells Hospital and Medical SchoolUniversity of Dundee
  • Lijun Wang
    • Institute for Medical Science and Technology (IMSaT), Ninewells Hospital and Medical SchoolUniversity of Dundee
  • Benjie Tang
    • Cuschieri Skills Centre, Ninewells Hospital and Medical SchoolUniversity of Dundee
  • Timothy Frank
    • Institute for Medical Science and Technology (IMSaT), Ninewells Hospital and Medical SchoolUniversity of Dundee
    • Institute for Medical Science and Technology (IMSaT), Ninewells Hospital and Medical SchoolUniversity of Dundee
    • Institute for Medical Science and Technology (IMSaT), Ninewells Hospitaland Medical SchoolUniversity of Dundee
  • Alfred Cuschieri
    • Institute for Medical Science and Technology (IMSaT), Ninewells Hospital and Medical SchoolUniversity of Dundee

DOI: 10.1007/s00464-007-9716-8

Cite this article as:
Wang, Z., Wang, L., Tang, B. et al. Surg Endosc (2008) 22: 1838. doi:10.1007/s00464-007-9716-8



Magnetic retraction has potential advantages over existing direct physical retraction means (e.g., forceps) in terms of providing complete atraumatic retraction, avoiding tumour cell exfoliation as well as offering the possibility of noncontact retraction.

This paper describes a pilot study of surface magnetic retraction of the gastric mucosa to facilitate resection.


Fifteen porcine stomach specimens were used in this pilot study. The uniaxial tensile properties of retracted mucosa were initially studied using a tensiometer. Magnetic media of ferromagnetic microparticles (stainless steel 410) dispersed in cyanoacrylate liquid were prepared at four different concentrations, and a neodymium permanent magnet was used to magnetically retract the media. The media was finally surface-glued to the target mucosa for performing a simulated surgical procedure.


The force measurement data show that the retraction forces increased as the concentrations of microparticles and magnetic media volumes increased. A magnetic media concentration of 2 g/mL was most suitable since it offered sufficient retraction force from a small volume of applied media, e.g., the observed magnetic forces exerted on 50 μL of media were 1.42 N by a 3-mm magnet and 3.75 N by a 6-mm magnet, respectively, both being more than sufficient for the mucosal retraction. The additional forces required for dissection with four alternative instruments, i.e., electrosurgery hook, snares, scalpel or scissors, were also measured, e.g., the total force required to retract up to 10 mm and resect the mucosa with snares was 0.36 ± 0.17 N. In a simulated surgical procedure (resection of gastric mucosa with glued magnetic medium) retraction by the magnet allowed resection of the tented mucosa by an electrosurgical snare.


Surface ferromagnetisation of target mucosal tissues could enable magnetic retraction for endoscopic surgery.


Magnetic retractionmagnetic mediaGastric mucosaMagnetic force measurementFinite element modelling

Magnetic media have been used for several medical and surgical applications [1] including as contrast agents for magnetic resonance imaging (MRI) [2, 3]. Meeker et al. [4] described a method of advancing and navigating catheters within the Central Nervous System (CNS) by means of external magnets interacting with a small magnet embedded in the catheter tips. Navigation using magnetic properties was also reported by Uchiyama et al. [5] in a gel model containing dispersed fine magnetic particles, and they proposed this technology for the detection and imaging of brain tumours. Paired magnets have been employed to create sutureless vascular anastomosis [6, 7] in experiments involving pigs and dogs, in clinical cases to create an internal enteric fistula [8] and anastomosis for bilioenteric anastomotic stricture after living-donor liver transplantation [9].

The present pilot study was designed to investigate various approaches for ferromagnetisation of target tissues, in order to advance the applications of ferromagnetism in surgical practice, particularly in minimal access surgery (MAS) and interventional flexible endoscopy. Currently in MAS the internal anatomy is accessed by slender instruments through sealed ports, thus reducing the traumatic insult and hence the postoperative pain and hospital stay with acceleration of recovery to full activity. In addition, compared to open conventional surgery, the MAS approach significantly reduces the risk of postoperative infections and adhesion formation. However, in MAS surgical tasks are more ergonomically demanding and difficult to execute. The surfaces of organs are smooth and moist and hence exhibit very low friction. To provide effective traction the jaws of grasping instruments feature ridges or toothed patterns, and while these are described as atraumatic in reality they do inflict trauma, especially if the tissue slips or is grasped too tightly, potentially resulting in delayed healing, surgical adhesions or perforation. Slippage during traction also promotes tumour cell exfoliation and hence risk of recurrence. In surgery for radical and potentially curative cancer resections, minimal and atraumatic contact with the tissues is clearly needed and a noncontact method of interacting with the tissue would seem ideal.

Ferromagnetic tissue would offer a number of far-reaching advantages in the advancement of surgical practice, especially in MAS and interventional flexible endoscopy: the target tissue could be retracted during surgical interventions by the use of magnets. Noncontact tissue retraction is a further possibility, which would directly address the tumour exfoliation problem inherent to direct instrumental grasping, while contact magnetic traction would provide a uniform and controllable pressure on the grasped tissue. Low-power thermal ablation with minimal collateral damage is another possibility, using magnetic heating to the evenly distributed magnetic domains in the component cells of the target tissue [10].

Although other approaches for ferromagnetisation of target tissue, (e.g., direct injection of magnetic media, incorporation of magnetic nanoparticles into cells) are under investigation in our laboratory and will be published elsewhere, this paper reports our preliminary results using surface target tissue impregnation for magnetic retraction of porcine gastric mucosa. Specifically, a ferromagnetic microparticle suspension in cyanoacrylate fluid (glue) was used in the present investigation for retraction by means of a permanent magnet. With this model the retraction forces were measured using magnetic glue suspensions of different volumes and concentrations.

In addition, finite element modelling (FEM) was carried out to provide a theoretical estimation of the forces between a permanent magnet and the magnetic media. The simulation results can be helpful for the selection of both the ideal magnetic particles and permanent magnets for particular applications. An FEM application, i.e., COMSOL Multiphysics 3.3 software (COMSOL Ltd, London, UK), was used to model the magnetostatic problems. Previously called FEMLAB, COMSOL Multiphysics FEM simulations have been used by other researchers for electromagnetic FEM studies in medical technology, e.g., in magnetic drug targeting [11], and cancer therapy [12].

Materials and methods

Mucosal retraction force measurement by tensiometry

The uniaxial tensile properties of the retracted mucosa were measured using an Instron tensiometer (Model 5564, Instron Ltd, Buckinghamshire, UK). The load frame of the tensiometer (Fig. 1A) is designed to secure a test specimen between the rigid frame base and the moving crosshead. The system is controlled by a testing program (i.e., Bluehill® 2 software, Instron Ltd.) run by an online computer (Dimension 1100, Dell Computer Co Ltd.). The measurement accuracy over the range of forces employed in this experiment is ± 0.5% of indicated load.
Fig. 1

(A) The tensiometer: the schematic highlights its load frame, and the experiment setup for measurement of the retraction force. (B) Photograph illustrating mucosal retraction by forceps

Sixteen porcine gastric specimens were used for this study. For the determination of the tensile properties of the retracted mucosa, the specimen was stretched at a steady rate of 10 mm/min to a predefined distance (e.g., 10 mm) while the displacement and corresponding load were simultaneously recorded and displayed by the software program. Surgical forceps was used to connect the test specimen to the load cell (Fig. 1). The additional load required to retract the tissue during dissection was also measured. Four alternative dissecting devices were investigated, i.e., electrosurgical hook (Force FX electrosurgical generator, Valleylab, Boulder, Colorado, USA), snares (20mm Mini-Std Oval, MICROVASIVE, Boston Scientific, USA), scalpel, and scissors. The cutting and dissection procedures were normally completed within 5 min.

Magnetic materials and magnetic retraction

Neodymium iron boron (NdFeB) magnets are the strongest permanent magnets as is reflected by their high residual magnetic flux density Br. Depending on the manufacturing process, Br varies between 1.0 and 1.4 Tesla (T) for NdFeB magnets, compared with about 0.3 T for ferrite magnets. In our experiment we used neodymium disc magnets with a remanence of 1.10 T (grade 30H, Eclipse Magnetics Ltd, Sheffield, UK), and of two dimensions: 3-mm diameter (2 mm height) and 6-mm diameter (6 mm height).

Custom-made magnetic media were made of ferromagnetic microparticles (stainless steel – AISI 410 powder, Goodfellow Cambridge Ltd., Huntingdon, Cambridgeshire, UK) dispersed in ethyl cyanoacrylate liquid (Henkel Loctite Adhesives Ltd, Herts, UK). The densities were 7.73 g/cm3 for the ferromagnetic SS 410 powder and 1.1 g/cm3 for the cyanoacrylate liquid. Media with concentrations of 0.5 g/cm3, 1 g/cm3, 2 g/cm3 and 2.5 g/cm3 were made and used in the study. Concentrations higher than 2.5 g/cm3 were extremely difficult to make and practically impossible to apply because of their high viscosity.

The magnetic susceptibility χ of normal tissue is commonly considered to be the same as that of water [13], which is χwater = −9.04 × 10−6 [14]. Assuming a magnetic susceptibility of around −10−5 for tissue, water and cyanoacrylate liquid, their relative permeability μr can be approximated to be μr = 1+ χ ≈ 1. The relative permeability of the magnetic (SS 410) particles is around 1000 [11].

The magnetic force between the permanent disc magnet and the ferromagnetic media was measured using the tensiometer under a control setup: the media were applied to a plastic substrate instead of mucosal tissue. This setup enables the medium to be firmly bonded to the substrate, thus permitting measurement of the maximal magnetic force generated between the magnet and medium as the effect of glue adhesion (to tissue) is excluded. Initial flexible (viscoelastic) displacement was simulated by using a rubber interface between the retraction device and the base support. The maximum magnetic pull-in-contact force was obtained when the lifted media started to fall off the disc magnet.

Magnetostatic theory and FEM model

COMSOL Multiphysics 3.3 software, was used to model the magnetostatic parameters of the experimental setup. Details about the magnetostatic theoretical equations and modelling results for the magnetic flux density and magnetic force on target magnetic materials (media) will be published elsewhere.


The data obtained showed that the force required to retract and lift the mucosa by 10 mm was 0.135 N (SD = 0.065, N = 16), and that there was a linear relationship between the retraction force and distance with an estimated mocusa elastic stiffness of 0.013 N/mm (Fig. 2). As shown in Table 1, the additional force required to retract the mucosa during resection with snares was 0.189 N (SD = 0.089, N = 16). The additional forces exerted by other dissecting devices were 0.233 ± 0.100 N (N = 16) with electrosurgical hook, 0.195 ± 0.082 N (N = 16) with scissors, and 0.621 ± 0.218 N (N = 16) with scalpel. The estimated total retracting force during cutting with snares (most commonly used in interventional flexible endoscopy) was 0.324 ± 0.154 N (N = 16).
Fig. 2

Forces required to retract mucosal tissue up to 10 mm (mean ± standard deviation). The trend line shows a linear relationship between the retraction force and distance with an estimated mocusa elastic stiffness of 0.013 N/mm

Table 1

Dissection force with instruments

Dissecting tool


E. hook



Dissecting force ± SD (N)

0.189 ± 0.089

0.233 ± 0.100

0.621 ± 0.218

0.195 ± 0.082

Total force* (N)

0.324 ± 0.154

0.368 ± 0.165

0.756 ± 0.283

0.330 ± 0.147

Number of specimens





* Total force was the dissecting force plus the 10-mm retraction force of 0.135 ± 0.065 N (N = 16)

There were variations in the retraction tests between the gastric specimens studied (0.135 ± 0.065 N, N = 16) due to individual variations and to different retraction locations. The important effect of retraction location within any one gastric specimen was confirmed by higher forces being consistently recorded near the proximal (gastroesophageal) and distal (pyloric) ends of the stomach. Repeat tests at the same location irrespective of site yielded similar results (e.g., 0.203 ± 0.002 N, N = 16) confirming reproducibility.

The magnetic pull force exerted on the magnetic media by an external permanent magnet was measured under the control setup and the results are shown in Fig. 3. The magnetic forces generated increased with higher concentrations of microparticles and larger media volumes. In general, a concentration of 2 g/mL (concentrations >2 g/mL proved difficult to make and apply) in a small volume of medium (e.g., 25–50 μL) could generate sufficient force to enable both retraction and resection of the mucosa by both the 3-mm- and 6-mm-diameter magnets. For example, the measured magnetic forces exerted on 50 μL of media (at 2 g/mL) were 1.42 N and 3.57 N by the 3-mm and 6-mm disc magnets, respectively.
Fig. 3

Observed results of maximal magnetic retraction force between the media and magnet probe (both the 3-mm and 6-mm magnets used). (A) Magnetic forces at various media concentration with a fixed volume of 10 μL. (B) Magnetic forces at various media volumes with a fixed concentration of volume of 2 g/mL. The required force (retraction to 10 mm and cutting with snares) of 0.324 ± 0.154 N is plotted as a reference

Simulated mucosal resection

In one simulated surgical procedure, the medium was applied to the mucosa as an 8-mm-diameter droplet (Fig. 4A). A 3-mm magnet was placed in contact with the media, and the mucosa was magnetically retracted by 10 mm (Fig. 4B). An in vitro surgical resection of the retracted mucosa was then performed with use of electrosurgical snares, the magnet being capable of maintaining retraction of the mucosa (via the medium) throughout the procedure (Fig. 4C).
Fig. 4

Photographs illustrating a simulated surgical procedures using magnetic mucosal retraction. (A) The magnetic medium was applied to the mucosa (harvested porcine stomach tissues). (B) The 3-mm permanent magnet was in contact with the medium (for this test, the magnet was magnetically attached to forceps) and the mucosa was retracted magnetically to a height of 10 mm. Snares were placed around the target tissue prior to the application of the permanent magnet to the media. (C) The target tissue was resected with the magnet successfully retracting/holding the media and tissue

Effect of the microparticle concentration and volume of magnetic media on magnetic retraction force

Figure 3 shows the magnetic force generated as a function of media concentration/volume and magnet size, together with the required force (retraction to 10 mm and cutting/dissection with snares, with reference to the minimum required force of 0.324 ± 0.154 N). With the 6-mm magnet, even a medium volume of 10 μL with concentration as low as 0.5 g/mL was sufficient for the required retraction (Fig. 3A), and more force could be generated with larger media volumes or higher concentrations. For example, at a concentration of 2 g/mL, the magnet pull-in-contact forces were 2.63 N for 25 μL media and 3.57 N for 50 μL media (Fig. 3B). However, we observed instances where the medium peeled off the mucosa (due to weak glue adhesion) before the maximum magnetic retraction force was reached.

In general, a concentration of 2 g/mL was preferred as with a small volume of medium (e.g., 25–50 μL), this generated sufficient force to retract/dissect/cut the mucosa with both magnets (well above the required force, as shown in Fig. 3). The magnetic force was also measured in the presence of an intervening air gap (noncontact force), as shown in Fig. 5A. The force decreased exponentially as the air gap increased, especially with the 3 mm disc magnet. The noncontact magnetic force can be increased by using larger volumes of media or higher concentrations.
Fig. 5

Magnetic force study: (A) Observed magnetic force on one medium (10 μL at 2 g/mL) using both the 3-mm and 6-mm magnets. The trend lines show that the magnetic forces decrease exponentially with the air-gap distance x. (B) Observed and FEM model simulated magnetic force results with reference stainless steel (SS) disc (9 mm diameter and 1.2 mm thick) using both the 3-mm and 6-mm magnets. (C) The magnetic force was related to the diameter ratio between the permanent magnet and the target media

A stainless-steel (SS) disc supplied by the magnet manufacturer was used in order to determine the full pull strength of the magnet and to obtain a reference benchmark for verifying our method for the magnetic force measurement on the custom-made media. The SS disc is 9 mm in diameter and 1.2 mm high with a relative magnetic permeability of μ= 1000. The measured magnetic forces are plotted in Fig. 5B, together with their FEM simulation results. The results indicated that the maximum pull-in-contact forces were 8.92 N (1.87 N pull through a 1-mm air gap) for the 6-mm magnet, and 2.36 N (0.46 N pull through a 1-mm air gap) for the 3-mm magnet. There was good agreement between the observed results and FEM simulations. This, in turn, verified our magnetic force measurement method used for the media test.

We also investigated the optimal geometry for dispersing the selected volume of medium on the surface of the target tissue, assuming that when the liquid media is applied it forms a circular pool before solidifying. The modelling shows that distributing the magnetic media in a pool with twice the diameter of the magnet maximises the retraction force for a given medium volume, as shown in Fig. 5C.


We attribute the variations in the mucosal retraction forces observed in the study to intrinsic variations between the specimens studied. The experiments confirmed that the precise location of the retraction site within the stomach determines the force generated. The consistent recorded of higher forces close to the proximal (gastroesophageal) and distal (pyloric) ends of the stomach is in keeping with anatomy as in these two regions the gastric mucosa is more firmly bound to the underlying muscular layers. The excellent reproducibility of the mucosal retraction force measurement was confirmed by repeated tests at the same location within one specimen. Within the limitations of the present experiments, we found that a magnetic microparticle concentration of 2 g/mL dispersed in a small volume of cyanoacrylate medium (e.g., 25–50 μL), generated sufficient force to retract/dissect/cut the mucosa with both the 3-mm and 6-mm magnets.

The experimental data also confirmed the feasibility of noncontact magnetic grasping (with the development of the necessary sensor and control technology). Undoubtedly, the magnetic force in the presence of an intervening air gap decreases sharply as the air gap increases, and hence more powerful magnets would be needed for noncontact retraction at a safe operating distance. However, the noncontact magnetic force can be increased by using larger volumes of media and or with higher concentrations of magnetic microparticles.

The comparative studies with the stainless-steel (SS) disc together with their FEM simulation studies indicated that the maximum pull-in-contact forces were 8.92 N (1.87 N pull through a 1-mm air gap) for the 6-mm magnet and 2.36 N (0.46 N pull through a 1-mm air gap) for the 3-mm magnet. There was good agreement between the observed results and FEM simulations. This, in turn, verified our magnetic force measurement method used for the media test, although the force generated by the magnet using the reference SS disc was more than five times greater.

With FEM modelling, the optimal geometry for dispersal of the selected volume of medium on the surface of the target tissue indicated that the pool of magnetic medium on the tissue should have a diameter twice that of the magnet for maximal retraction. The in vitro resection of the magnetically retracted gastric mucosa by electrosurgical snares confirmed the feasibility of the proposed technology for surface magnetic retraction of target tissue which is strong enough to permit resection.

However, there are a number of residual problems, which have to be overcome before the technology can be used clinically and its full potential realised. In the first instance, further studies are needed to investigate other parameters necessary for an optimal match between the external magnet (e.g., size and configuration) and implanted magnetic media (geometry and material property). Equally important is the search for more effective biocompatible tissue adhesives, which are active and bond firmly to moist tissue surfaces. Drying the mucosal tissue prior to the application of the magnetic media was used in the present experiments to improve the glue bonding strength between the media and mucosa. Even so there was a tendency for the magnetic medium to peel off in the presence of a strong magnetic force. Another problem is related to the surface flatness and surface area ratio of the medium (to the permanent magnet). Since the maximum magnetic pull strength is achieved when the medium is in contact with the magnet end surface, the larger the contact area the bigger the retraction force. Also because the magnet end surface is completely flat, this requires a smooth and flat medium surface to match the magnet for maximal strength. New magnet configuration designs are under investigation to try to solve this problem. Other options of ferromagnetic tissue retraction that obviate the need for the use of adhesives are being explored in our laboratory and include direct injection of the magnetic particles into the tissue, although this option carries its own limitations, and transfection of tissues by biocompatible ferromagnetic nanoparticles.


Our preliminary experimental data demonstrate that magnetic retraction can be used to retract tissue for surgical dissection, cutting, and resection. Our custom-made magnetic media using permanent magnets as small as 3 mm diameter generated more than sufficient retraction forces to permit resection using various instruments. Larger magnets can generate stronger force but also require larger media surface area as a result of the ratio factor identified by the present study. Fifty microlitres of media with a concentration of 2 g/mL, distributed in a disc of 8 mm diameter, was found to be optimal for retraction by 3-mm and 6-mm permanent magnets.

Copyright information

© Springer Science+Business Media, LLC 2007