Technology innovation in medical devices 2

Background:

Benefit of implantation of an internal biliary stent (IBS) during liver transplantation to reduce biliary complications was recently demonstrated [1]. Silicone IBS was used in practice, which require an endoscopic ablation procedure, a potentially morbid intervention for the patient [2]. In order to avoid this, and to reduce biliary complications after liver transplantation we develop a resorbable IBS, made from a radiopaque and degradable polymer visualizable by X-ray. The objective of this study was to evaluate mechanical and radiological behavior of this polymer during in vivo degradation after rats implantation.

 

Methods:

PLA₅₀-PEG-PLA₅₀ triblock copolymer was used as polymer matrix, in which was incorporated the X-ray visible triiodobenzoate-poly(e-caprolactone) copolymer (PCL-TIB/PLA₅₀-PEG-PLA₅₀) component in order to make the compound visible by X-rays [3]. Polymer sample corresponding to a dogbone tensile specimen (36x6x0.6mm) were implanted in the abdomen of 17 rats. At each degradation time (1 week, 1 month, 2 months, 3 months and 6 months), a X-ray scanner was performed with contrast intensity and implant volume measurements. The specimens were then explanted to perform ex-vivo biomechanical tests. The mechanical tests consist in a load at two strain levels (5% and 10%) with a strain rate during loading of 1% per second followed by a tensile relaxation test. There were realised with a Mechanical Tensile Service^Ò machine with a load cell of 25N, in physiological conditions (physiological serum immersed tests at controlled temperature of 37°C). A histological study at 3 and 6 months was also performed.

 

Results:

About radiological characteristics during degradation, the contrast intensity and the implant volume were stable over time up to 6 months (between 1000 and 2000 Hounsfield Units for the contrast intensity and 75 mm³ of median volume). The mechanical properties were tested between zero and 2 months, the evolution of the viscoelasticity of the materials was studied. After this period, mechanical tests were impossible due to the extreme brittleness of the specimens. Histological analysis indicated good tissue integration without major inflammation.

 

Conclusion:

In this study, radiological and mechanical behaviour of PCL-TIB/PLA₅₀-PEG-PLA₅₀ biopolymer was describe during in vivo degradation, in order to allow development of an IBS to reduce biliary complications after liver transplantation.

 

References:

1. Tranchart H, Zalinski S, Sepulveda A, et al (2012) Removable intraductal stenting in duct-to-duct biliary reconstruction in liver transplantation. Transpl Int 25:19–24 .

2. Girard E Evaluation de l’implantation peropératoire d’un Drain Biliaire Interne cholédochocholédocien en Transplantation Hépatique Orthotopique: étude comparative rétrospective à propos de 294 patients. https://hal.archives-ouvertes.fr/

3. Samuel R, Girard E, Chagnon G, et al (2015) Radiopaque poly(ε-caprolactone) as additive for X-ray imaging of temporary implantable medical devices. RSC Adv 5:84125–84133 .

Introduction: We have previously reported on extracardiac soft robotic devices for augmenting ventricular function in the failing heart [1,2]. Fixation of such devices on the moving myocardium is a major challenge with using these technologies. Efficient mechanical coupling allows for increased transmission of systolic or compressive assistance, transmission of twisting motion, and ability to assist with the diastolic or filling phase [2]. Here we conducted a study in a murine myocardial infarct model to evaluate the device-tissue interface provided by two potential device materials. We attached a patch of either medical mesh or silicone to the epicardium, and characterized the in vivo response in terms of (i) conformability using μCT and (ii) bio-integration as measured by material/tissue separation shear force at acute and chronic time-points. This characterization will guide the design of extracardiac devices and future therapeutic strategies.

Methods: The study design is outlined in Figure A. Male C57 mice (22-31g, n=10) were anesthetized using isoflurane and mechanically ventilated. The heart was exposed through a thoracotomy, and the pericardium was removed. The left anterior descending coronary artery was permanently ligated. 3x3mm patches of silicone (Nusil, 300µm thickness) and medical mesh (Parietex, Covidien) were attached to the epicardium at the site of infarction with two 8-0 prolene sutures. The skin was then closed with 4-5 interrupted sutures (5-0 polysorb). For μCT, hearts were fixed, then stained with iodine vapour for 24-72 hours before imaging. Adhesion testing was conducted in shear on fresh tissue samples using custom-built grips with a mechanical tensile tester at 20mm/second.

Results: The medical mesh demonstrated increased conformability compared to silicone by μCT (Figure B). Additionally, mesh showed superior adhesion (Figure C) to the epicardial surface with a pull-off force of ~100mN compared to 40mN at 4 days, with the mesh pull-off force increasing up to 4N at 14 days.

Figure: A) Study design, B) Mesh conforms better to the myocardium compared to silicone as shown by μCT. Yellow dashed line shows conforming mesh edge, yellow solid shows malposition of silicone and C) Representative pull-off testing showing greater mesh myocardial adhesion than silicone adhesion at 4 days. Mesh adhesion increased up to 4N at 14 days.

Discussion: A mesh or textile-based approach for extracardiac devices may improve conformability, bio-integration and strength of adhesion to the myocardium for improved mechanical coupling at the device/tissue interface. An example therapeutic strategy could involve implantation of a mesh-based robotic sleeve in the passive state for an initial bio-integration period followed by activation of the device. This would enable passive ventricular restraint and more efficient myocardial coupling for active devices.

References:
1. Roche, ET, Science translational medicine, 2017;9(373) 
2. Payne, CJ, Soft Robotics, 2017;4(3):241-250.

Trauma is one of the major causes of mortality and morbidity worldwide affecting mainly young people between the ages of 15 to 29. Uncontrolled bleeding contributes to 30% to 40% of trauma-related deaths and is the leading cause of potentially preventable early in-hospital deaths. There is no effective method available to first responders to provide temporary control of non-compressible intra-abdominal (IA) bleeding while civilian and military patients are transported to the hospital. Previous studies demonstrated that abdominal insufflation (AI) provides effective temporary bleeding control. The long-term goal is to develop a Portable Abdominal Insufflation Device (PAID) for rapid and controlled AI of the abdominal cavity. The goal of the present study was to build the early prototype and test it in bench top and animal models. PAID applies uniform pressure to the entire abdominal cavity, and generates an intra-abdominal pressure gradient that compresses blood vessels to the extra-luminal space thus decreasing bleeding rate. The device is equipped with a Veress needle, a CO₂ cartridge, skin grippers to pull the abdominal wall away from intra-abdominal organs, a Maxon motor and mechanisms to insert the needle, valve, pressure sensors and a Raspberry PI 3 processing unit to automatically stop insertion when the abdominal cavity and required pressure are reached. The bench top testing models are abdominal cavities built from silicone of different thickness to simulate abdominal wall thickness. We used swine abdominal wall tissue for a closer resemblance of the skin and abdominal wall mechanical properties to the human tissue. The prototype was built from commercially available and rapid prototyping components. The bench top testing was performed by lifting the silicone wall, inserting the Veress needle and insufflating the simulated abdominal cavity. The animal tissue was placed over a box which simulates the abdominal cavity. The total box and tissue weight was matched to add up to the normal weight of a human abdominal wall. The prototype passed our acceptance criteria for clinical specifications: grip strength to hold a typical abdominal wall, gradual advancement of the needle throughout the entire abdominal wall, needle block when it entered the abdominal cavity, insufflation to the target pressure of 18mmHg maintained for 30 min. We have passed the technology readiness level TRL4 by building the functional PAID prototype and successfully testing in an artificial model. The next steps are to optimize the device dimensions to make it as small as possible while maintaining function and testing it on an animal model of internal bleeding. The research leading to these results has received funding from UEFISCDI Romania, under the project “Innovative portable insufflation device to stop uncontrolled abdominal bleeding in military and civilian trauma”, contract no. 244PED/2017, PN-III-P2-2.1-PED-2016-1587.

Introduction

Within the CARDIS project, we are exploring the use of a multi-beam laser Doppler vibrometer (LDV) to detect asymptomatic carotid stenosis from measurement of skin vibrations on the neck of affected patients. This method is based on the hypothesis that flow instabilities induced by the stenosis will propagate as mechanical waves through the soft tissues of the neck to the skin. As a first proof-of-principle, we report measurements on hydraulic bench models to assess the sensitivity of LDV to detect vibrations arising from flow instabilities.  

Methods

A compliant model of the carotid bifurcation with internal carotid artery (ICA) 76% area-stenosis was created in silicone rubber. It was then mounted in an open-topped Perspex box and surrounded by an aqueous hydrogel to mimic the neck’s soft tissues. A skin-like layer was then applied over the gel surface. To ensure adequate reflection of the laser light (wavelength 1540 nm), small patches of retro-reflective tape were attached to the skin layer. LDV measurements were acquired 1diameter (0.8 cm) downstream from the stenosis. For reference, intra-arterial pressure measurements were performed at the same location (see Fig1A). We set the flow ratio between the branches to physiological values ^1,2, while the nature and the level of the inflow were adjusted in order to match multiple flow conditions such as constant flow with 1) volumetric flow rate VFR=130ml/min, 2) pulsatile flow with VFR=340ml/min and pulse pressure PP=20 mmHg, and 3) pulsatile flow with VFR=800ml/min and PP=40mmHg. Water was used for the experiments. Signals were sampled at 20KHz (Powerlab 16/35) and Fast Fourier Transform frequency spectra were computed in LabChart with 8K points, Hann windowing and 50% overlap.  The spectra of the LDV and pressure signals were then normalized against the no-flow condition and plotted in Fig1B and Fig1C, respectively.

Results

Peaks in the pressure spectra were absent for the lowest flow rate (130ml/min, steady flow), while their width and amplitude increased when the PP matched more physiological values. No peaks were found in the LDV spectrum at the lowest flow rate while a narrow 70-90Hz peak in the PP=20mmHg became a wider 100-150Hz when PP=40mmHg.

Discussion

The absence of peaks at the lowest flow rate was due by the absence of stenosis-induced-flow instabilities. Once the VFR was high enough to trigger disturbed flow, both pressure and displacement signals showed peaks in the low frequency range. The difference in frequency range could be due to wave propagation mechanisms which form the basis of a second set of experimental tests investigating shear wave propagation in soft tissues.

Acknowledgments

Funded by H2020 ICT-26-2014 CARDIS.

References

1. Likittanasombut P, J Neuroimaging. 2006;16(1):34-38.

2. Groen HC, J Biomech 2010;43(12):2332-2338.

Introduction

Flow diverting stents are widely used for treating brain aneurysms with large necks to effectively occlude the aneurysmal sac [1]. Computational Fluid Dynamics (CFD) is often used to investigate the effects of such devices on aneurysmal hemodynamics. However, the rather complex configuration of such devices leads to a tedious meshing procedure and lengthy CFD simulations. To reduce these complexities, a homogeneous porous medium approach is often used to model the effects of the device on aneurysmal hemodynamics. Unfortunately, the homogeneity assumption leads to inaccurate results due to inhomogeneity of the configuration. [2]. In this study, we propose a heterogeneous porous medium approach to investigate the aneurysmal hemodynamics after device deployment, which simplifies the application of CFD while not sacrificing accuracy.

Methods

Devices were first virtually deployed into an anatomical aneurysm geometry using a FE (Finite Element) method and then broken into a lattice of hexahedral elements. For each hexahedral element, porosity and permeability, two essential quantities for porous medium studies, were calculated. A 3D map of these quantities was then generated to be used as an input to the Navier-Stokes equations. Darcy’s law was then incorporated to consider the effect of the porous medium on hemodynamics. Aneurysmal VRMS was used to quantify resulting hemodynamics in the saccular region.

Results

In the following table, the effects of different porosity/permeability map resolutions on the aneurysmal VRMS are shown. With map refinement, the aneurysmal VRMS predicted by the porous medium method approached the value obtained when considering the explicit geometry of the flow diverter used to treat the aneurysm. There is a small deviation with the most refined map resolution that will be addressed as we continue to advance the study.

Map Resolution	Aneurysmal VRMS	% Difference
Porous Medium	Explicit Geometry
16	0.12157	0.05913	364.14%
32	0.09248	56.40%
64	0.06589	11.43%

 

Discussion

We showed that representing a flow diverter as a heterogeneous porous medium can recover aneurysmal hemodynamics when map resolution is refined enough. The proposed method reduced the complexities of the CFD study by a factor of 10. Specifically, since the explicit geometry of the device was not used, the meshing procedure was simplified, and extremely small boundary layer mesh elements of the device were removed, leading to a lower computational cost.

[1]         B. N. Roszelle, P. Nair, L. F. Gonzalez, M. Haithem Babiker, J. Ryan, and D. Frakes, “Comparison Among Different High Porosity Stent Configurations: Hemodynamic Effects of Treatment in a Large Cerebral Aneurysm,” J. Biomech. Eng., vol. 136, no. 2, pp. 021013-021013-9, Feb. 2014.

[2]         H. G. Morales and O. Bonnefous, “Modeling hemodynamics after flow diverter with a porous medium,” in 2014 IEEE 11th International Symposium on Biomedical Imaging (ISBI), 2014, pp. 1324–1327.

IntroductionMitral regurgitation is the most common heart valve disease. Restrictive annuloplasty using a rigid or semi-rigid ring is a central technique for its repair. Although annuloplasty is generally effective, post-operative dehiscence of the sutures that secure the ring to the valve annulus is an increasingly recognized and often catastrophic problem.^1-2 Suture dehiscence is easily dismissed as a byproduct of surgical error. However, our group is undertaking a multi-part study to identify ring design parameters that can affect dehiscence likelihood systematically, in pursuit of safer designs.

We recently reported that, despite lower cyclic force (F_C) in the beating heart on posterior annular sutures, the weaker posterior mitral annular tissue creates higher likelihood of dehiscence in that region.³ This corroborated the clinical predominance of dehiscence from the posterior annulus, and established the need to specifically reduce F_C in that region. We subsequently detected differences in F_C dynamics according to degree of annular downsizing, and/or the specific commercial ring selected.⁴ However, none of these variables had the desired effect on posterior sutures. Presently, F_C dynamics were investigated as a function of ring stiffness.

MethodsCustom suture force transducers were attached to annuloplasty rings of interest and implanted in normal ovine subjects (N=22) via standard annuloplasty procedure. In the beating heart, transducers detected F_C on each individual suture. The semi-rigid Physio^TM ring was compared against fully-rigid and fully-flexible ring prototypes of equivalent geometry. Based on findings from this cohort, a novel saddle-shaped hybrid ring, possessing rigid anterior and posterior segments separated by flexible commissural regions, was also developed and tested.

Results/DiscussionWhile no F_C differences across stiffnesses were detected for anterior sutures, posterior F_C was significantly reduced following use of the flexible ring (rigid: 1.95±0.96N, semi-rigid: 1.76±1.19N, flexible: 1.04±0.63N; p<0.001). Increasing flexibility also led to greater asymmetry between anterior and posterior F_C (p<0.001). These findings suggested a mechanism, whereby a more rigid ring enables more rapid/complete force equilibration around the suture network. The resultant transfer of higher anterior forces to the weaker posterior tissue increases overall dehiscence likelihood.

Compared to all other rings, hybrid rings exhibited no difference in anterior F_C. Yet, their posterior F_C (1.17±0.81N) was lower, and their anterior-posterior F_C ratio was higher, than all but the fully-flexible variety (each quantity p<0.001 vs. rigid, p>0.9 vs. flexible). The hybrid design effectively disrupted the problematic anterior-to-posterior force transfer, validating the proposed suture force distribution mechanism. These findings establish significant potential for a new ring design that can optimally balance suture dehiscence likelihood and repair efficacy.

References1. Dumont et al, Ann.Thorac.Surg, 2007. 2. Tsang et al, J.Am.Coll.Cardiol, 2009. 3. Pierce et al, Ann.Thorac.Surg, 2016. 4. Pierce et al, J.Thorac.Cardiovasc.Surg, 2017.

Introduction

Single ventricle patients need to undergo 3-stage surgical intervention to survive.In result, the blood return from the inferior vena cava and superior vena cava bypassing the right ventricle is diverted directly to the pulmonary artery by use of an extra cardiac Gore-Tex graft.At later ages, these children might suffer from life-threatening problems, likely from elevated venous pressure and ventricular dysfunction.We propose a novel Fontan right-side assist device that incorporates the aortic balloon pump concept into the extra cardiac Fontan graft.This design concept requires no additional thoracotomy for device implementation and has minimal risk for the patient.This research will determine the feasibility of this concept and identify the balloon pump modifications needed to produce clinically significant hemodynamic benefits.

Methods

We developed a model of the Fontan circulation using clinical data across a full range of resting and exercise states.Then, using a realistic model of the balloon pump coupled with the Full-body physiology model of the Fontan circulation(Fig 1A), we investigated a range of the critical device characteristics and quantified their impact on the Fontan circulation.To verify the model of the balloon pump mounted on a Gore-Tex graft, the model is implemented into a computational lumped-parameter network physiology model(Fig 1B) and validated against the equivalent benchtop experiment(Fig 1C).After identifying the optimal balloon and driver design, we constructed a benchtop experiment to conduct physical device testing under physiologic conditions, and performed a set of experiments each using a different graft to investigate device operation.

Results

We identified device criteria for achieving good hemodynamic outcomes.The criteria include a device heart rate up to 170bpm, stroke volume approximately equaling the volume of the Fontan graft, systolic-diastolic ratio of 30%, and -5mmHg of device diastolic pressure.The physical experiment under the physiologic condition shows that for the Fontan balloon pump application, the graft material with stiffness in the range of durometer~70A could improve flow augmentation.

Discussion

In this research, we confirmed that implementation of a balloon pump will improve the hemodynamics of a Fontan patient.We observed significant improvement in physiologic outputs by increasing the balloon pump heart rate.However, high balloon heart rates could limit the improvement in physiologic outcomes due to incomplete graft filling.We identified two critical device criteria to alleviate this.1)Decreasing the systolic fraction and consequently prolonging the duration of graft filling.2)Applying negative cuff pressure during diastole.An elastic graft could create negative pressure during device diastole.Conversely, the Gore-Tex graft cannot withstand mechanical wear of the balloon pump.Hence, we recommend stiffer material than the Gore-Tex for the Fontan graft that could “pop back” to its reference geometry after compression, thereby creating a negative pressure during device diastole.