Über uns

Felix Winter, Catrin Bludszuweit-Philipp, Olaf Wolkenhauer
Journal of Cerebral Blood Flow & Metabolism, First Published 1 Jan 2017 https://doi.org/10.1177/0271678X17693024

Blood oxygen level-dependent functional magnetic resonance imaging (BOLD-fMRI) is a standard clinical tool for the detection of brain activation. In Alzheimer’s disease (AD), task-related and resting state fMRI have been used to detect brain dysfunction. It has been shown that the shape of the BOLD response is affected in early AD. To correctly interpret these changes, the mechanisms responsible for the observed behaviour need to be known. The parameters of the canonical hemodynamic response function (HRF) commonly used in the analysis of fMRI data have no direct biological interpretation and cannot be used to answer this question. We here present a model that allows relating AD-specific changes in the BOLD shape to changes in the underlying energy metabolism. According to our findings, the classic view that differences in the BOLD shape are only attributed to changes in strength and duration of the stimulus does not hold. Instead, peak height, peak timing and full width at half maximum are sensitive to changes in the reaction rate of several metabolic reactions. Our systems-theoretic approach allows the use of patient-specific clinical data to predict dementia-driven changes in the HRF, which can be used to improve the results of fMRI analyses in AD patients.
 

Annette Geltmeier, Beate Rinner, Dennis Bade, Katharina Meditz, Reiner Witt, Uwe Bicker, Catrin Bludszuweit-Philipp, Patrick Maier
PLoS ONE 2015 10(8): http://dx.doi.org/10.1371/journal.pone.0134999

Treatment options specifically targeting tumour cells are urgently needed in order to reduce the side effects accompanied by chemo- or radiotherapy. Differences in subcellular structure between tumour and normal cells determine their specific elasticity. These structural differences can be utilised by low-frequency ultrasound in order to specifically induce cytotoxicity of tumour cells. For further evaluation, we combined in silico FEM (finite element method) analyses and in vitro assays to bolster the significance of low-frequency ultrasound for tumour treatment. FEM simulations were able to calculate the first resonance frequency of MCF7 breast tumour cells at 21 kHz in contrast to 34 kHz for the MCF10A normal breast cells, which was due to the higher elasticity and larger size of MCF7 cells. For experimental validation of the in silico-determined resonance frequencies, equipment for ultrasonic irradiation with distinct frequencies was constructed. Differences for both cell lines in their response to low-frequent ultrasonic treatment were corroborated in 2D and in 3D cell culture assays. Treatment with ~ 24.5 kHz induced the death of MCF7 cells and MDA-MB-231 metastases cells possessing a similar elasticity; frequencies of > 29 kHz resulted in cytotoxicity of MCF10A. Fractionated treatments by ultrasonic irradiation of suspension myeloid HL60 cells resulted in a significant decrease of viable cells, mostly significant after threefold irradiation in intervals of 3 h. Most importantly in regard to a clinical application, combined ultrasonic treatment and chemotherapy with paclitaxel showed a significantly increased killing of MCF7 cells compared to both monotherapies. In summary, we were able to determine for the first time for different tumour cell lines a specific frequency of low-intensity ultrasound for induction of cell ablation. The cytotoxic effect of ultrasonic irradiation could be increased by either fractionated treatment or in combination with chemotherapy. Thus, our results will open new perspectives in tumour treatment.

Olaf Wolkenhauer, Charles Auffray, Olivier Brass, Jean Clairambault, Andreas Deutsch, Dirk Drasdo, Francesco Gervasio, Luigi Preziosi, Philip Maini, Anna Marciniak-Czochra, Christina Kossow, Lars Kuepfer, Katja Rateitschak, Ignacio Ramis-Conde, Benjamin Ribba, Andreas Schuppert, Rod Smallwood, Georgios Stamatakos, Felix Winter and Helen Byrne
genomemedicine.com/content/6/3/21
Genome Medicine 2014, 6:21

From reactions in cells to organ physiology: Systems medicine is an interdisciplinary approach that integrates data from basic research and clinical practice to improve our understanding and treatment of diseases. Systems medicine can be seen as a further development of systems biology and bioinformatics towards applications of clinical relevance. The term ’systems‘ refers to systems approaches, emphasizing a close integration of data generation with mathematical modeling [1-3]. The(mal)functioning of the human body is a complex process, characterized by multiple interactions between systems that act across multiple levels of structural and functional organization – from molecular reactions to cell-cell interactions in tissues to the physiology of organs and organ systems. Over the past decade, we have gained detailed insights into the structure and function of molecular, cellular and organ-level systems, with technologies playing an important role in the generation of data at these different scales.

S Hübner, S Kressirer, D Kralisch, C Bludszuweit-Philipp, K Lukow, I Jänich, A Schilling, H Hieronymus, C Liebner, K Jähnisch
ChemSusChem 2/2012

Short diffusion paths and high specific interfacial areas in microstructured devices can increase mass transfer rates and thus accelerate multiphase reactions. This effect can be intensified by the application of ultrasound. Herein, we report on the design and testing of a novel versatile setup for a continuous ultrasound-supported multiphase process in microstructured devices on a preparative scale. The ultrasonic energy is introduced indirectly into the microstructured device through pressurized water as transfer medium. First, we monitored the influence of ultrasound on the slug flow of a liquid/liquid two-phase system in a channel with a high-speed camera. To quantify the influence of ultrasound, the hydrolysis of p-nitrophenyl acetate was utilized as a model reaction. Microstructured devices with varying channel diameter, shape, and material were applied with and without ultrasonication at flow rates in the mL min(-1) range. The continuous procedures were then compared and evaluated by performing a simplified life cycle assessment.

Bludszuweit-Philipp C, Kuehne S, Lukow K.
ASD Advanced Simulation & Design GmbH, Rostock, Germany
XXXX ESAO Congress
Glasgow, Scotland, UK, September 11-14 2013

Objective: Atrial fibrillation (AF) is a major risk factor for stroke due to frequent embolisation of thrombi from the heart. In AF, irregular blood flow patterns are presumed to form thrombi mainly in the left atrial appendage (LAA). Current treatment of all patients with AF involves precautionary administration of anticoagulants (ACs) despite a serious increased bleeding risk. Especially for the low-risk patient groups, the benefit of ACs does not outweigh their harmful damages. While modern image-based methods allow for a very detailed analysis of the heart’s anatomy and physiology, they cannot predict blood flow and thrombosis risk. From this deficiency, a novel prediction tool was developed which simulates the patient-specific blood flow and coagulation processes in the heart with AF.

Methods & Materials: A detailed patient-specific computer model of the beating heart based on the individual ECG is generated which calculates the three-dimensional, transient blood flow, particularly within the LAA. A complex thrombosis model fully couples the haemodynamics with the coagulation chemistry and realistically simulates the activation, deposition and agglomeration of platelets and resulting fibrin-rich thrombus formation. Key actors like activated and resting platelets, ADP, TxA2, Prothrombin, Thrombin and ATIII are considered. With the coupled flow-thrombosis model, the formation, growth and detachment of a thrombus in the LAA is simulated. Induced thrombolysis is simulated when an AC is given.

Results: Simulations reveal that insufficient flushing in the LAA with an absent contraction can lead to thrombus formation. Recurring sinus-rhythm contractions largely increase the risk of thrombus detachment.

Conclusions: The novel simulation tool enables the prediction of patient-specific risk factors, like LAA anatomy, for thrombosis and embolisation during AF as well as an optimised treatment which prevents over-use of anticoagulants.

Sandy F. C. Stewart, Eric G. Paterson, Greg W. Burgreen, Prasanna Hariharan, Matthew Giarra, Varun Reddy, Steven W. Day, Keefe B. Manning, Steven Deutsch, Michael R. Berman, Matthew R. Myers & Richard A. Malinauskas
Cardiovasc Eng Tech 3, 139–160 (2012)  DOI: 10.1007/s13239-012-0087-5

While computational fluid dynamics (CFD) is
commonly used for medical device development, its usefulness for demonstrating device safety has not been proven. Reliable standardized methods for this specialized need are lacking and are inhibiting the use of computational methods in the regulatory review of medical devices. To meet this need, participants from academia, industry, and the U.S.
Food and Drug Administration recently completed a computational interlaboratory study to determine the suitability and methodology for simulating fluid flow in an idealized medical device.

Bludszuweit-Philipp C., Kuehne S., Geltmeier A., Lukow K., Tsakanikas V., Fotiadis D.
40. Jahrestagung der Deutschen Gesellschaft für Thorax, Herz und Gefäßchirurgie
Stuttgart, Germany, February 13-16 2011

Objective: For a personalised intervention of diseased coronary arteries, a suitable planning tool is required to support the clinician in stent selection and positioning based on the patient’s vessel anatomy and plaque characterisation to ensure an optimum stent alignment and favourable haemodynamics.

Methods: A 3D reconstruction of the stenosed coronary arteries is based on imaging data obtained by IVUS (fused with X-ray angiographies) or CT imaging including a computer-aided characterization of plaque components. A virtual positioning and realistic deployment of the selected stent/balloon device inside the stenosed artery is performed by computer simulation. Resulting information on the stented artery is transferred to a flow module to calculate local and global coronary blood flow characteristics. Relevant intervention parameters such as stent type and length, artery elasticity, inflation pressure and haemodynamic conditions can be varied.

Results: The simulations predict and visualise the outcome of the stenting procedure, like the lumen dilatation, lesion coverage and stent alignment as well as predictors for artery traumatisation during percutaneous coronary intervention. The subsequent flow analysis reveals the effects of the stenting intervention both on the local haemodynamics, which greatly affects proliferation and restenosis, and on global flow parameters such as perfusion rates and pressure drop of the affected arteries.

Conclusions: A new simulation tool is presented which provides support to cardiologists for a personalised intervention planning of diseased coronary arteries, based on the patient’s specific anatomy, plaque information, haemodynamic conditions and device specification.

Bludszuweit-Philipp C., Kuehne S., Geltmeier A., Lukow K.,
ANSYS Conference & 27. CADFEM Users´ Meeting
Leipzig, Germany, November 18-11 2009

One of the main problems in today’s design of blood flow devices is the reduction or avoidance of thrombotic events. These are not only influenced by the haemocompatibility properties of the materials but also by the specific flow conditions.To enhance the design process and to prognosticate the outcome of device intervention, a thrombosis model is needed which allows the prediction of thrombus formation and growth in medical implants and devices already in the design phase. The model developed fully couples the three-dimensional haemodynamics with the coagulation chemistry and, hence, realistically simulates the activation, deposition and agglomeration of platelets and resulting thrombus formation. The effect of the growing thrombus on flow deviation and possible obstructions is considered. The thrombosis model was applied to the prediction of clotting in patient-specific aneurysms for both, stented and unstented blood vessels. It was successfully validated against clinical patient data. The simulation greatly enhances the prediction of the outcome of a clinical intervention in the cardiovascular field and, thus, contributes to a risk reduction for the patient. It also serves as design accelerator for cardiovascular device manufacturers.

Tahir H., Bludszuweit-Philipp C., Geltmeier A., Lukow K., Kuehne S.
ANSYS Conference & 27. CADFEM Users´ Meeting
Leipzig, Germany, November 18-11 2009

Aneurysmal Subarachnoid Haemorrhage (SAH) is a major cause of morbidity and mortality.  Recent advancements have shown that hemodynamic characteristics such as wall shear stress, vorticity, jetting, recirculation, pressure fluctuation and Oscillatory Shear index (OSI) are the underlying factors involved in the genesis, growth and rupture of the intracranial aneurysms.In modern treatment of aneurysms, there has been a growing interest of using stents as a flow diverter to protect the aneurysm by reducing the action of hemodynamic forces and thus facilitating aneurysm thrombus. Due to limitations in the existing imaging techniques, the direct assessment of these hemodynamical indices after stent deployment is haltered and currently resorting to numerical simulations only. However, simulating blood-flow in a stented patient-specific model poses a number of challenges. These include: techniques for vascular anatomoy reconstruction from medical image, virtual deployment of stent, meshing the vascular model and the deployed stent, construction of computational volumetric grids, developing appropriate boundary conditions, and handling of relatively large computational meshes.In this study, numerical analysis of blood flow in stented and unstented cerebral aneurysm has been conducted. Patient-specific geometries are created using @neuFuse software. ANSYS ICEM CFD and ANSYS CFX are used to generate volumetric grids and to solve the unsteady Navier Stokes equations respectively. Assessment of simulation results allows the prediction of the effectiveness and outcome of the clinical intervention.

Bludszuweit-Philipp C., Geltmeier A., Buske M.
XXXV Annual meeting European Society for Artificial Organs
Geneva, Switzerland, September 3-6 2008

Objectives: The treatment of specific types of endovascular diseases like intracranial aneurysm increasingly require the development of customised or new devices because existing devices can not fulfil the required function. Small vascular dimensions, access difficulties and forms, like bifurcations, not suitable for conventional stent designs present a very challenging task for developers. A simulation tool which integrates all essential design tasks and enables the virtual implantation of a device can thus allow pre-clinical performance evaluation and would minimise patient treatment risk.

Methods: A simulation method was developed which facilitates the investigation of mechanical, haemodynamic, chemical and biological responses to endovascular treatment. This methodology features integrated tools for the import of patient-specific vessel geometries, structure-mechanical and haemodynamic simulations. It models the physical deployment of a stent in a patient geometry and investigates the mechanical properties of the stent and its interaction with the vessel wall. A numerical blood flow simulation of the stented vessel including the prediction of thrombus formation and growth as well as possibly drug elution characteristics is integrated part of the analysis.

Results: The virtual stenting methodology was successfully applied to a number of patient cases with cerebral aneurysm within the EC-funded research project „@neurIST“. Existing stent designs were evaluated and new, customised designs investigated and virtually tested with respect to their performance and haemocompatibility.

Conclusions: The method developed enables the virtual testing of existing or new endovascular implants not only with respect to their optimal design and positioning within the patient’s blood vessel. Simultaneously, it allows to optimise their biological compatibility and interactions. Thus, it supports the further improvement of implant quality and safety of delivery and minimises the risk for patients before a clinical study.

Bludszuweit-Philipp C., Buske M., Kuehne S.
XXXV Annual meeting European Society for Artificial Organs
Geneva, Switzerland, September 3-6 2008

Objectives: One of the main problems in today’s design of blood flow devices is the reduction or avoidance of thrombotic events. These are not only influenced by the haemocompatibility properties of the materials but also by the specific flow conditions in the device. To enhance the design process and to prognosticate the outcome of device interventions, a thrombosis model is needed which allows the prediction of thrombus formation and growth in medical implants and devices already in the design phase.

Methods: The thrombosis model developed fully couples the three-dimensional haemodynamics with the coagulation chemistry and, hence, realistically simulates the activation, deposition and agglomeration of platelets and resulting thrombus formation. The model considers the key players in thrombus formation like activated and resting platelets, ADP, TxA2, prothrombin, thrombin, ATIII and anticoagulants.  The effect of the growing thrombus on flow deviation and possible obstructions is modelled through the consideration of  flow property changes in platelet aggregates or fibrin networks. With these combined features, the model describes the essential thrombosis processes which are characteristic for medical devices being in contact with blood.

Results: The thrombosis model was successfully validated against experimental data of different test cases. The capabilities of this approach for thrombosis prediction in medical devices is shown for different applications, such as stents, centrifugal blood pumps or oxygenators. Design-related causes for thrombotic events within the device are identified and appropriate design improvements can be recommended and verified.

Conclusions: The model developed enables a realistic prediction of thrombosis in blood flow devices. Using this virtual technique, thrombus growth within the device can be identified and minimised already in the early design phase.

Geltmeier A., Bludszuweit-Philipp C., Buske M.
1st European HyperWorks Technology Conference 2007, EHTC2007
Berlin, Germany, October 22-24 2007

Within the frame of the EC-funded research project „@neurIST“ cerebral aneurysm are investigated concerning their patient-specific prediction of rupture risks. Modern simulation methods are exploited to support the clinical decision management and the selection of a suitable aneurysm treatment.

A typical simulation procedure includes the analysis of blood flow in the diseased blood vessel, the virtual placement and deployment of medical devices, like stents, into the vessel and the simulation of the haemodynamic effects of the treatment. To enable an efficient simulation process, the necessary transfer between patient/device model generation and simulation software is performed using the import/export functionalities of HyperMesh and HyperView.

The general approach of this demanding simulation task involving different software tools and their interfaces is demonstrated and illustrated with clinically relevant examples.

Geltmeier A., Bludszuweit-Philipp C., Lukow K.
12th workshop on „The Finite Element Method in biomedical engineering, biomechanics and related fields“
Ulm, Germany, July 20-22 2005

In this study a complex numerical bone model which includes the orientation-dependent elasticity and strength material properties as well as the remodelling behaviour of bone was developed and its practical application was shown.

The first part of the project was an extensive literature research and the arrangement of the numerous anisotropic elasticity and strength material properties of different human bones in a data base.

Within the second part, a mathematical algorithm of bone remodelling procedure was implemented into the numerical analysis, validated and its application was shown. From the great variation of simple to very complicated theories an algorithm was choosen which is suited for implementation in a commercial FE-software and industrial applications. The chosen theory describes the strain induced internal bone adaptation. The implementation of this algorithm into the numerical analysis was realised by user subroutines within the commercial FE-Software package Marc / Mentat 2003. The algorithm was validated with a proximal femoral bone model under different load cases. The calculated density distribution of a typical daily load case was compared to an x-ray of a healthy femoral bone. A very good correspondence could be found. The application of the implemented algorithm was shown for a femoral bone model with an implanted hip stem under different contact conditions.

Summarising this study a complex numerical bone model was developed and validated which allows the consideration of anisotropic material properties as well as strain-induced remodelling behaviour. It represents a further important numerical tool for more realistic numerical investigations of bone structures itself and the interaction of bone and implants.

Bludszuweit C., Kuehne S.
XXX Annual meeting European Society for Artifical Organs
Aachen, Germany, September 3-6 2003

Aim: Applying the methods of CFD to the analysis of artificial organs, qualitative and quantitative evaluations of the flow conditions are possible. To assess the hemolysis potential of such devices more exactly, additional models describing the mechanisms of mechanical and thermal blood damage are necessary. The aim of this investigation is the development of models describing the phenomena of mechanical and thermal hemolysis, implemented in CFD codes and, thus, using the advantages of numerical design tools.

Methods: Using the Lagrange method to predict the mechanical and thermal loads of representative particles along their tracks, both local and integral assessment of the hemolysis potential is possible. Taking into account the experimental data, a comprehensive thermal hemolysis model valid for a wide temperature range is proposed. Applying a linear reaction-kinetic approach, the thermal hemolysis is predicted considering the time-derivatives of temperature and the exposure time of the blood cells in areas of elevated temperature. To predict the mechanical hemolysis, an approach analogous to the prediction of the life time in engineering applications is used.

Results: The proposed models were implemented in a commercial CFD software and validated against experimental data. The method was applied to different examples (blood pumps etc.) A detailed prediction of hemolysis potential with a high spatial resolution within these devices was obtained.

Conclusion: The presented models enable very efficiently the prediction and localisation of hemolysis potential of artificial organs.

Bludszuweit C., Krueger K., Kuehne S., Mennenga H.
XXX Annual meeting European Society for Artifical Organs
Aachen, Germany, September 3-6 2003

Aim: Modern bioreactors are characterised by complex flow and cell growth conditions with significant spatial gradients of species and cell concentrations, temperature and flow properties. The process controlling of such reactor types is very demanding. It depends not only on the biological growth kinetics but on the convective and diffusive transport processes in the reactor too. A new method was developed which combines the advantages of a detailed CFD simulation and a simplified process analysis. Thus, the controlling parameters of the coupled biological, thermal and fluid flow processes of bioreactors can be investigated numerically.

Methods: The proposed method consists of two principle steps. Firstly, a detailed CFD analysis is performed with standard boundary conditions. Therein, the biological growth and reaction rates are implemented in the CFD code as additional conservation equations considering the dependence between cell concentration and flow characteristics. Based on this simulation, simplified process analyses are performed to obtain the controlling parameters of the bioreactor.

Results: The developed model was successfully applied to a fermentor. As results the cell growth was obtained as a function of the controlling parameters (feed solution, flow velocity and temperature).

Conclusion: The proposed method is an effective way to determine the process controlling behaviour of modern bioreactors without extensive experimental investigations. Especially for design and scale-up tasks of bioreactors this method presents a valuable tool.

Geltmeier A., Bludszuweit C., Krüger K.
XXX Annual meeting European Society for Artifical Organs
Aachen, Germany, September 3-6 2003

Background: The use of the Finite Element (FE) Method in the early phase of development of medical implants gains more and more significance for medical device manufacturers. Not only the calculation of the static strength but also the life time prediction of the implants are of high interest. These need to endure a defined number of load cycles stated for mechanical tests. With examples of an orthopaedic and a dental implant, the procedure of theoretical prediction of life time on the basis of a Finite Element Analysis will be shown.

Methods: A proximal femoral nail and a dental implant are chosen for FE – Analysis. The non-linear 3D–modelling of the complex geometry of each implant, which consists of different parts, considers the fact that the components can contact and influence each other. Using the calculated principal stresses of at least two static load cases of the dynamic loading and considering notch factors, surface roughness and fatigue data of the material, the life time of the implant can be predicted.

Results: The results for implant life time obtained with this procedure were validated using experimental data of biomechanical testing of the proximal femoral nail. The predicted life time was compared with reached load cycles during the test under a defined load value and a good correspondence was found.

Conclusion: The described procedure enables the life time prediction of medical implants on the basis of an FE-Analysis, surface conditions and material data. The procedure was validated using experimental data of biomechanical testing of a proximal femoral nail.

Geltmeier A., Bludszuweit C., Krueger K., Raschke S.
9th workshop on „The Finite Element Method in biomedical engineering, biomechanics and related fields“
Ulm, Germany, July 18-19 2002

In this study, a finite element model of femoral bone and hip prosthesis was developed which includes the patient specific gait analysis data as a more „dynamic“ load case than one leg stance and stair climbing. This model together with a data base enables the optimisation of biomechanical parameters of the prostheses stem regarding minimisation of relative motion and bone strains.

The FE-model of femoral bone with and without prosthesis was developed using CT-data of cadaveric bones, which were used in strain gauge experiments. The material data for spongious and cortical bone were taken from the literatur and defined in terms of grey steps of the CT-slices.
Three load cases were investigated: one leg stance, stair climbing and one gait cycle (40 steps quasistatic). The loading during gait cycle was taken from gait analyses of different persons carried out in a gait analysis laboratory in Canada (BCIT). For all load cases, the influence of the biomechanical parameter femoral offset on relative motion of the prosthesis stem and the principal strains on the bone surface was investigated.

It could be shown that biomechanical parameters like the femoral offset have a considerable influence on long term standing of hip prostheses. This FE-model with the complex load cases enables the variation and optimisation of different design parameters and the prediction of the expected biomechanical durability. Together with the available patient specific data (age, weight, sizes, gait analysis data …), an extensive data base was built, which allows a better preparation of the implantation and higher long-term success.

Lawford P.V., Hose D.R., Bludszuweit C., Keggen L.A.
Sixth annual hilton head workshop „Prosthetic heart valves: past, present and future“
Hilton Head Island, South Carolina, USA, March 6-10 2002

The application of computational tools to heart valve design has been explored for two decades. In principle, gross fluid performance indices, such as pressure drop or effective orifice area, are readily computable for candidate designs using standard computational fluid dynamics tools. The solution of transient flow systems is now routine, and physical flow characteristics are typically determined from direct solution of the transient Navier Stokes equations. The recent development of fluid-solid interaction software has enabled the opening and closing characteristics of valves, and the influence on the flow fields, to be investigated. In parallel with these developments there has been significant progress in the use of transient particle tracking to monitor shear history data for particles in the flow. Although many challenges remain in the physics domain, the software is now sufficiently functional and mature to address meaningful problems in heart valve design.

This paper will present a solid-fluid interaction environment based on the external coupling of CFX to ANSYS, and will demonstrate applications in the cardiovascular sphere, including a study of the opening and stability characteristics of a mechanical valve under a prescribed pulsatile wave form.

A major challenge in the application of these software tools is the building of the bridge from computable physical system characteristics to physiological processes such as haemolysis and thrombosis. The haemolytic potential of a candidate valve design might be assessed by integration of damage parameters over particle paths in the flow. Example computations based on cfd and mri data will be presented. The latter part of the blood coagulation cascade can be represented by the clotting of enzyme-activated milk. Results will be presented for in vitro tests studying the clotting of milk on test objects and on valve prostheses. The paper will illustrate the application of a computational model of thrombosis based on the changing of viscous fluid properties as a function of the local flow field. A simulation of the clotting process on a test object will be demonstrated and compared with in vitro data.

Bludszuweit C., Rosenow S.-E.
CFX-Update, No. 15, Spring 1998

Artificial organ engineers are generally faced with complex and demanding design challenges. Devices must both fulfil organ-specific functions such as physical/chemical transport and exchange processes and still be bio-compatible. Blood in direct contact with a mechanical system cannot be altered beyond a physiologically repairable degree. ASD Advanced Simulation & Design GmbH has developed a powerful method for designing and optimising such devices with its „Virtual Artificial Organ“ concept. In this technique, prototypes are developed using modern simulation tools such as CFD. Functionality and haemocompatibility can be predicted in advance, thus reducing costs as well as time invested in intensive animal and in-vitro testing. As such, development time is shortened considerably. ASD has applied this concept using CFX-TASCflow at organ-specific level with a centrifugal blood pump, an oxygenator and a haemodialyser, as these are the most significant organ assist and replacement devices for heart, lung and kidney.

Centrifugal blood pumps have a particularly high potential for causing mechanical blood damage. The numerical simulation performed therefore sought to maximise pump performance while minimising blood trauma. A flow analysis of the entire stator/rotor configuration in a centrifugal blood pump yielded 3D distributions of velocity, static pressure, turbulence parameters and molecular as well as Reynolds stresses. Particle streak lines with representative stress values and transit times were calculated to quantify the mechanical loading of blood elements on their passage through the pump and the most recent mechanical blood damage models were incorporated to help damage assessment. It was thus possible to study geometrical performance parameters and damage characteristics effectively.
Oxygenators and dialysers carry out mass exchange by using large artificial surface areas (hollow fibres) which are a potential source for bio-incompatibility. High mass transfer and transport in future generations must, hence, be achieved with a much reduced fibre area. ASD has numerically predicted exchange-unit 3D flow processes and optimised the results. Analysis of global flow characteristics such as blood velocity distribution within the entire fibre bundle has been combined with a break-down of highly complex local external fibre flow to enable optimisation of the design and operation of these devices. A deeper understanding of physical/chemical processes and their complex interactions in artificial organs can be expected as the virtual organ concept is further developed.

Bludszuweit C.
1997 ASME Fluids Engineering Division Summer Meeting FEDSM’97, June 22-26 1997

Artificial organs are required to satisfy high functionality and bio-compatibility demands for successful use. Efforts in the development of a modern generation of artificial organs are supported by the application of advanced Computational Fluid Dynamics (CFD) methods with the study presented.

Based on a detailed specification of design criteria and the incorporation of a novel model for mechanical blood damage prediction, a numerical design evaluation was performed for three characteristic artificial organs.
Flow conditions within a centrifugal blood pump, an oxygenator and a haemodialyzer were analysed for performance characteristics and blood trauma potential. Parameter variations were simulated and conclusions for latent design improvements drawn.

The results revealed that powerful simulation tools can be used to accompany the artificial-organ design optimisation process effectively.

Bludszuweit C., Gaylor J.D.S.
Artificial Organs, 19(7):590-596, Blackwell Science Inc., Boston
© 1995 International Society for Artificial Organs, https://doi.org/10.1111/j.1525-1594.1995.tb02386.x

The successful use of centrifugal pumps as temporary cardiac assistance devices strongly depends on their degree of blood trauma. The mechanical stress loading experienced by cellular components on their passage through the pump is a major cause of blood trauma. Prediction of the mechanical stresses will assist optimization of pump design to minimize haemolyses and platelet activation. As a theoretical approach to this task, the determination of the complete three-dimensional (3D) flow field including all regions of high shear stresses is therefore required. A computational fluid dynamics (CFD) software package, Tascflow, was used to model flow within a commercially available pump, the Aries Medical Isoflow Pump. This pump was selected in order to demonstrate the ability of the of the CFD software to handle complex impeller geometries. A turbulence model was included, and the Newtonian as well as the Reynolds stress tensor calculated for each nodal point.

A novel aspect was the assignment of scalar stress values streak lines representing particle paths through the pump. Scalar stress values were obtained by formulating a theory that enabled the comparison of a three-dimensional state of stress with a uniaxial stress as applied in all mechanical blood damage tests. Stress loading time functions for fluid particles passing inlet, impeller and outlet domains were obtained. These showed that particles undergo complex, irregularly fluctuating stress loading. Future blood damage theories would have to consider an unsteady stress loading regime that realistically reflects the flow conditions occurring within the pump. Validation of the pump modelling was demonstrated with pressure head discrepancies predicted to within 15% of measured values.

Bludszuweit C., Gaylor J.D.S.
University of Strathclyde, Bioengineering Unit, Glasgow, Scotland, U.K.
Artificial Organs, 19(7):583-589, Blackwell Science Inc., Boston
© 1995 International Society for Artificial Organs,  https://doi.org/10.1111/j.1525-1594.1995.tb02385.x

Knowledge of the correlation between mechanical loading of formed blood elements and the amount of their destruction is important for the prediction of blood trauma in artificial circulatory devices as well as in natural circulation. A haemodynamic assessment and optimization of artificial organs to minimize trauma could be undertaken in the design phase given a comprehensive mechanical blood damage model.

A theory to determine blood trauma theoretically as a combination of a mechanical loading analysis and a phenomenological blood damage resistance hypothesis was presented. The arbitrary stress-time function of blood particles predicted by flow analysis were reduced to a set of simple time function for which the damage behaviour may, in principle, be obtained from mechanical blood damage tests.

A classification of those stress functions into damaging and non-damaging parts was followed by an overall trauma prediction taking account of cumulative effects by means of a damage accumulation hypothesis. Theoretical determination of blood destruction caused by mechanical stresses in a centrifugal pump is one possible application of the proposed theory.

The strategy of haemolysis prediction was demonstrated for the Aries Medical Isoflow Pump. Irregular stress-time loading functions of particles passing the pump domain obtained by three-dimensional numerical flow simulations were reduced and classified into harmonic components. To relate these functions to their haemolytic response can only be done in a qualitative manner since blood damage behaviour under transient stress loading has not been sufficiently investigated. Accurate prediction of blood trauma using the proposed theory will require detailed study of the influence of frequency and amplitude of harmonic stress loading on blood elements formed.

Bludszuweit C.

University of Strathclyde, Bioengineering Unit, Glasgow, Scotland, U.K.
Ph.D. Thesis, 1994, https://doi.org/10.1111/j.1525-1594.1995.tb02386.x

The successful use of centrifugal pumps as temporary cardiac assistance devices strongly depends on the extent to which they damage blood. The development of a theoretical pump evaluation model was performed in this study to facilitate effective pump optimisation. The optimisation process sought to maximise flow performance and minimise blood trauma which is primarily caused by hydrodynamic stresses. A general mechanical blood damage theory was developed comprising a combination of information about mechanical blood loading with the knowledge of its resistance properties. In this theory arbitrary loading-time functions were reduced to simple loading functions for which the damage behaviour is known. A linear damage accumulation theory contributed towards determining partial damage and the correlation in the overall damage process.

The application of this novel blood damage prediction theory was demonstrated for haemolysis prediction in a centrifugal blood pump. Particle loading-time functions were determined via a 3-dimensional numerical flow analysis of the entire pump domain by means of assigning scalar stress values to particle streak lines. Scalar stress values were obtained by a theory comparing a six-component stress tensor with uniaxial stresses as applied in blood damage tests. It was shown that particles undergo a complex, irregularly fluctuating stress loading and that turbulent stresses and flow conditions in the outlet domain are the most critical factors. Haemolysis tests using a oscillating capillary tube arrangement were performed to investigate blood damage resistance properties under cyclic stress loading in hitherto unexplored amplitude and frequency ranges. A non-linear damage curve for the stress amplitude-cycle number was derived which indicated the existence of a red blood-cell endurance strength.

Detailed information about the mechanical loading of blood within a centrifugal pump was obtained for the first time and linked to its traumatic effect. It offers the possibility of effective, multi-parameter optimisation of blood pumps in the design phase.