Design of a Hydrodynamic Performance Bench for Ventricular Assist Devices

Bruno Santos^1*, Tarcisio Leao², Eduardo Bock³, Evandro Drigo⁴, Jeison Fonseca² and Aron Andrade^2
*Correspondence: Bruno Santos, Department of Electric, Federal Institute of Sao Paulo, Rua Pedro Vicente, Sao Paulo, Brazil, Email:

Keywords

Test and evaluation bench • Ventricular Assist Devices (VADs) • Control and automation • Bioengineering

Introduction

In Brazil, cardiovascular diseases kill about 350,000 Brazilians every year, being the largest cause of death in the country [1]. The incidence of cardiovascular diseases in humans has been increasing over the years, the most serious disease being chronic heart failure (CHF) [2-4]. Heart transplantation is considered the standard treatment in patients with advanced and refractory CHF, however, it is a procedure limited by the number of available donors and possible contraindications that make it unfeasible [5]. Given the difficulty in obtaining organs for heart transplants and the current need of patients, the development of devices for pumping blood in mechanical circulatory assistance has highlighted importance in the area of bioengineering, among them the study of developments of Ventricular Assist Devices (VAD) [6,7], whose function is to replace the mechanical work of the left and/or right side of the heart, ensuring the cardiac performance necessary for the patient's life while providing favorable conditions for recovery (bridge to recovery), for transplantation (bridge to transplantation) or even the complete replacement of the heart by the device (destination therapy) [8].

To build efficient and non-harmful VADs for the implanted patient, it is necessary that VAD designs and prototypes undergo a series of tests and validations on test benches [9-11]. A VAD test bench aims to produce unambiguous results that can serve for [12]:

(i) Comparison with the state of the art

(ii) Validation with current standards of manufacture and use

(iii) System response to disturbance and noise

(iv) Finding any problems and confirming their resolution

The main VAD “in vitro” assays discussed in the literature as a reliable process to predict VAD success are [13-15] the hydrodynamic performance assay and the haemolysis assay.

In the test of hydrodynamic performance it is mainly evaluated the capacity of the VAD to support vital conditions for the patient by maintaining 5 l/min of flow with approximately 110 mmHg of differential pressure between the inlet and outlet cannulas. In addition, in this test a map of values of the rotation speed and electric current of the pump motor is generated with the respective flow and differential pressure generated, which serve as a catalog of the safe operation of the device [15,16]. In this test the pumped fluid can be blood or an analogous fluid with similar viscosity (usually a solution of water and a 1/3 of glycerin) [13].

Another application of the value map obtained for the VAD in the hydrodynamic performance test is to use it as a reference for estimating physiological values without the use of sensors by means of look-up table (LUT) 2D technique. The LUT is determined by a hydrodynamics test for operating range. There is a correspondence between rotation and current with flow and differential pressure and this correlation is mapped by means of a three-dimensional surface. This technique was patented in Brazil by Leao and Andrade under number BR1020160068363.

In the haemolysis assay, the possibility that pumping by the VAD causes trauma to the red cells is assessed, releasing haemoglobin into the plasma as a by-product of the breakdown of red blood cells, the higher the rate of haemolysis the more traumatic is the flow of assisted pumping into the blood [17]. The measurement of this trauma is done using the plasma free hemoglobin (PFH). The normalized index of hemolysis (NIH) should remain between 0.004 g/100 l and 0.02 g/100 l for VADs (Nose 1998) [9]. The normalized hemolysis test, according to the standards (ASTM F1841) and (ASTM F1830), should last 6 hours, in a closed circuit, at a flow rate of 5 l/min at a pressure of 100 mmHg, conditions achieved in the implantable centrifugal pump at approximately 2800 rpm [17].

An assay with estimation of the results of a haemolysis assay is performed by Particle Image Velocimetry (PIV) which is a benchtop experiment for studying fluid dynamics that has greater detail than sensors and devices that measure pressure and volume [18]. The analysis is performed on the bench using a rapid-fire camera and a laser pulsed at a uniform rate. The region illuminated by the laser is captured by the camera shot and the image is processed in computer programs. In the images, it is possible to detect particles that serve as positioning of the flow vectors [18].

In 2013, the Thematic Research project supported by the Sao Paulo State Research Support Foundation (FAPESP) was initiated, entitled “Sistemas Propulsores Eletromagnéticos para Coração Artificial Implantável e Dispositivos de Suporte Circulatório Mecânico” with the following objectives (FAPESP 2018): application of advanced engineering methods for the construction of high-performance devices, as well as systems to evaluate them from a clinical, experimental and technological point of view. This initiative was continued with the project “Sistemas Propulsores Eletromagnéticos para Coração Artificial Implantável e Dispositivos de Suporte Circulatório Mecânico” and enabled a series of relevant research in the line of VAD [14,15,19-28].

Among the VADs under development in the project is highlighted the Implantable Centrifugal Blood Pump (ICBP) [28,30] and the Apic-Aortic Blood Pump (AABP) [15]. The main system for evaluation of VADs under development in the project is the Cardiovascular System Simulator (HCS) built at the Institute Dante Pazzanese of Cardiology (IDPC), which is a tool that allows the acquisition of data from the physical connection of a VAD to the simulated cardiovascular system, under physiological conditions or with alteration of some of the cardiovascular parameters.

Due to the FAPESP Thematic Research Project (2013/24434-0), there is a technical cooperation agreement (no. 14/2016, DOU of 18/08/16) between IDPC and the Federal Institute of São Paulo, Sao Paulo campus (IFSP SPO) [31]. It is highlighted from this partnership the development of new VAD control techniques through the Automation for Life Laboratory (ALL) of IFSP SPO [32,34-39]. Based on existing models/projects in IDPC and works in biomaterials [32,40-43], a new centrifugal VAD is under development in the Bioengineering and Biomaterials Laboratory (BIOENG) of IFSP SPO, goal to extend the life of the device without magnetic levitation (non-maglev) by the very low wear of the bearings by tribological interaction [17,44,45].

Aiming to perform the evaluations of the VADs developed at IFSP and also as a validation option to the VADs of the FAPESP Thematic Research Project (2013/24434-0) inserted in the scope of the IDPC-IFSP SPO cooperation agreement (nº 14/2016, DOU de 18/08/16) (IFSP 2018), this work aims to present the design of a Hydrodynamic Performance Bench (HPB).

Materials and Methods

Actuator

The first object of study was the choice of an actuator for the HPB. The choice conditions followed the following technical criteria: (i) a commercial catalogue motor, because there is immediate availability, easy maintenance, specified operating parameters and robust actuation guarantee by the manufacturer; and (ii) a brushless direct current motor (BLDC; type of motor generally used in VAD) with physical characteristics similar to those of an implantable motor, because the performance values of the VAD prototype will be close to the values of an "in vitro" test where the actuator is encapsulated in the VAD (step closer to a commercial product) and the control strategies applied in VAD control in HPB can be evaluated, and with that, easily shaped and modified to be the most suitable for this application.

Mechanical construction

The second object of study was the design of the mechanical components of the HPB. The conditions of choice followed the following technical criteria:

(i) Rapid prototyping, the components were designed to be easily prototyped by affordable 3D printers of the Fused Deposition Modeling (FDM) type, drastically reducing raw material expenses and machining processes.

(ii) Modularity, the components were designed to be assembling able and replaceable with new design updates both from VADs and the HPB itself.

(iii) External electromagnetic coupling, to preserve modularity, the drive between HPB and VADs will be performed without physical contact or direct field, but indirectly through the axial magnetic field produced by a disc with 3 pairs of neodymium N42 magnets (the quantity and type of magnets can be freely changed) fixed on the actuator rotational axis, which interacts electromagnetically with the VAD rotor magnets.

The components will be developed following the criteria using the Computer Aided Design (CAD) program SolidWorks^® (Dassault Systèmes, Vélizy-Villacoublay, France). The HPB components designed in 3D CAD were built on the C2 3D printer (3D Applications, Guarulhos, Brazil) by additive manufacturing in FDM using polylactic acid (PLA) filaments.

Transducers

The third object of study was the design of the transduction system of interesting physical phenomena for VAD performance evaluation. A sensor is a device whose output can be quantified and changes with one or more physical phenomena. This output information can be used for process monitoring and control. A transducer is a device that converts one form of energy into another. Measuring the physical variables associated with the resulting energy form allows estimation of the physical variables associated with the input energy [46]. The main variables of this project are the flow rate and the differential pressure of the VAD.

In HPB, the VAD flow will be obtained by means of the HT-110 flow transducer (Transonic Systems, Ithaca, USA). For this, it is only necessary to make the electrical connection between the analog outputs of the HT-110 flow meter with reference of the read in probe flow with the supervisory control system of the HPB. The working range of the HT110 flow transducer is 1 to 100 l/min.

In HPB, the differential pressure of the VAD will be obtained by means of two TruWare pressure transducers (Edwards Lifesciences, Irvine, USA), one located in the inflow cannula and the other in the outflow cannula of the VAD (differential pressure = pressure of the inflow cannula - pressure of the outflow cannula). To do so, it is necessary to perform an engineering adaptation of the TruWare pressure transducer, since this device is originally designed to be used together with a multiparametric hospital console (not available for use in HPB). The adaptations for use of the TruWare pressure transducer are:

(i) Amplify the signal, the pressure transducer has a resolution of 5.0 μ V/V/mmHg ± 1%, this value must be amplified to a standard working range in electronics (0 to 5 V) however within the interesting range for VAD analysis (0 to 300 mmHg). For this, the INA122 signal amplifier (Texas Instruments, Dallas, USA) will be used;

(ii) Straight line equation, with the established range of interest, it is necessary to identify the mathematical relationship of the pressure straight line in input measurement unit (mmH₂O) by the desired measurement unit (mmHg) and

(iii) Processing, with the straight line equation established, it remains to acquire the input signal and process it through an algorithm containing the straight line equation and filters to reduce noise associated with signal amplification (physical filter with capacitors and digital filter with average and Kalman).

Data acquisition and system control

The fourth object of study was the design of the HPB control and data acquisition system. For the satisfactory acquisition and storage of HPB results, a supervisory program with the following technical criteria is required:

(i) Human Machine Interface (HMI): for the operation of HPB as a test bench it is necessary that the process data are intuitively inserted, aiming the overall integrity of the tests.

(ii) Stable control of the actuators: the HPB actuator drive is performed by means of an electronic controller (driver), for this to occur it is necessary that the transmission and acquisition standards are respected, in a homogeneous flow, avoiding oscillations in the system that compromise the operation and the analyzed data.

(iii) Acquisition of data from sensors: HPB aims to collect key data to perform the performance evaluation of VAD. This is possible through signal transducers such as current (A), voltage (V), flow (l/min) and pressure differential (mmHg). Moreover, the control of the HPB is exerted on a complex variable (non-Newtonian fluid in acceleration) subject to momentum, disturbances and noises throughout the test, so a closed-loop actuation system with feedback of the controlled variable is mandatory.

(iv) Data storage, the validation of the performance of the VAD will be performed later with the use of value map and relations of the properties, so it is necessary that the data throughout a series test or long duration are faithfully stored in spreadsheets of later simple handling.

The HPB supervisory program was developed in graphic language through the Labview^® program (2015, National Instruments, Austin, USA) and implemented in a development and application platform (PXIe-8840, National Instruments, Austin, USA) with the signal acquisition/activation and amplification modules (NI PXIe-6361 and NI PXI4022, National Instruments, Austin, USA).

The manipulated variable of the supervisory system is the motor speed (rpm) acting via an Escon 50/4 EC-S power controller (Maxon Motor, Sachseln, Switzerland) on the BLDC EC45 N339281 motor (Maxon Motor, Sachseln, Switzerland).

The controlled variable is the “VAD flow” (l/min) provided by an HT-110 flow transducer. The process data are: pressure differential (mmHg) supplied by a TruWare pressure transducer, current (mA) and voltage (V) supplied by the power controller itself, and the speed of the motor feedback (rpm) also supplied by the power controller.

For accuracy and confidence in maintaining the chosen speed throughout the tests, a controller with closed-loop stability algorithm must be implemented [47]. The advantage of using a Proportional-Integral (PI) control as a stability filter is its easy digital implementation and to its versatility, providing satisfactory results even in the presence of noise and disturbances [37]. Using the Ziegler-Nichols Threshold Sensitivity Method to adjust the Proportional- Integral-Derivative parameters [37] a PI control algorithm was created for the current and speed actuator, this allows system stability even with significant variations in process characteristics. For the highest efficiency in controlling the HPB actuation, the chosen motors were parameterized and calibrated for the power controller in the Escon Studio^® program (Maxon Motor, Sachseln, Switzerland).

To evaluate the HPB supervisory control system, two tests were performed: Test 1 to evaluate the transmission and data acquisition of the supervisory control system; and Test 2 to evaluate the stability of the chosen motor speed maintenance.

Test 1 was performed to evaluate if the motor speed values chosen in the HPB supervisory control system are being effectively reproduced in the motor (reading by means of a tachometer) and also if the process data is being processed and faithfully reproduced in the HMI screen. This test was performed with the BLDC 475521 motor (Maxon Motor, Sachseln, Switzerland) as actuator, because although this motor does not have the physical dimensions indicated for encapsulation inside a centrifugal VAD, it has a wide working range (0 to 62000 rpm) enabling more comprehensive control tests.

Test 2 was performed to evaluate the stable maintenance of the motor speed in the presence of pumping resistance variations caused by the physiological system. For this the speed was adjusted increasing with 1000 rpm step (1000 - 4000 rpm) and decreasing with 1000 rpm step (4000 - 1000 rpm) in the HPB supervisory control system and the response of the motor without load and with load was examined by means of contact tachometer. Further, the response of the control in the presence of disturbance (inducing a load on the motor) and noise (proximity to electromagnetic devices) will be checked. This test was performed using the definitive actuator of HPB.

Particle image velocimetry module

The fifth object of study was the design of a VAD flow test module by particle image velocimetry (PIV). This test is effective to study the turbulence and stagnation areas of the VAD. In this module, the supervisory control system processed the image of the pump rotor rotation speed (transparent housing model of the analysed VAD) captured by a fast pulse camera coupled with a high intesity laser (540 nm and 5000 mW) and the data acquisition [17]. The fast pulse camera is a Hero 8 Black (GoPro, San Mateo, USA).

The laser serves to give fluorescence to the particles captured with the camera for mapping the velocity field, and to do this, it must be triggered like a strobe to give the effect of "freezing" the movement of the rotor during pumping.

The fluid used is water with reflective particles of lamellar Polyvinyl Chloride (PVC) or silver with a particle size between 0.1 and 0.5 mm.

The PIVLab toolbox works with image pre-processing, evaluation and post processing.

For post-processing the user determines the acceptable speed limits, this value can also be determined automatically by comparing each speed component with upper and lower limits, using the average speed and the standard deviation of the speed. Noise reduction in the data is done by mediated filtering, considering a penalized least squares method.

Results

The motor chosen for HPB actuation for meeting the two criteria defined in the actuator choice was the BLDC EC45 N339281 motor (Maxon Motor, Sachseln, Switzerland). Figure 1 shows the BLDC EC45 motor N339281.

The components developed to meet the three criteria defined in the mechanical construction of HPB are presented in Figure 2. With all the mechanical components of HPB available, the verifications of the fixation between the components and the mechanical interaction of transmission and absence of contact were performed, as well as the resistance of the material to wear and fatigue in working regime. Figure 3 shows the assembly of the HPB components built on a 3D printer.

Figure 4 shows the electrical schematic for the adaptation of the TruWare pressure transducer for use in the HPB. Figure 5 shows the tests for the experimental survey of the straight line equation with the mathematical correlations of pressure transduction. Figure 6 shows the straight line equation for use of the TruWare pressure transducer as a pressure variable within the HPB supervisory control system.

Figure 7 shows the reconfiguration of the Escon 50/4 EC-S power controller ports aiming the integration to HPB's supervisory and Figure 8 shows the calibration of the motors inside the Escon Studio^® virtual environment, where speed levels were adjusted in the program (blue colour) and the motor tried to follow (red colour) without overloading the expected current levels, from this relation the optimized gains of the Proportional-Integral (PI) control were identified.

In the supervisory control system of HPB (Figure 9) the motor speed is adjusted and sampled the real-time value of the process data of the motor feedback speed (rpm), Motor current (mA), VAD flow rate (l/min) and Differential pressure between the inlet and outlet cannula (mmHg).

Figure 10 shows test 1 of the HPB supervisory control system using the BLDC 475521 motor (Maxon Motor, Sachseln, Switzerland). Table 1 shows the results of test 1 of the HPB supervisory control system.

In Figure 11, the transfer functions with PI gains for current and speed of HPB actuator and the system response with PI control following the reference set in the supervisory control system of HPB. And in Figure 12 is shown the test 2 of HPB supervisory control using the definitive actuator of HPB inserted in the assembly of the mechanical components of HPB. Table 2 shows the results of test 2 of the HPB supervisory control system.

Figure 13 shows the flow test by PIV in the HPB of a blood pump for cardiopulmonary bypass developed at the institute Dante Pazzanese of Cardiology.

In 2018, FAPESP made a public call in Brazil for the development of paracorporeal radial centrifugal VAD within a fostering program called PITCHGOV. PITCHGOV is a program of the Government of Sao Paulo/Brazil aiming to solve challenges of public management through the connection between public power and startups. At HPB, the performance curve of the paediatric VAD developed for this project was raised [48]. Figure 14 shows the hydrodynamic performance test of a prototype pediatric VAD developed and its respective performance curve.

Santos developed an Intelligent Local System (ILS) for VAD, which is an intelligent embedded system with integration of real-time data processing, reconfigurable architecture and with communication protocols based on the concepts of Healthcare 4.0. To legitimize the feasibility of ILS as a VAD controller solution, the physiological multi objective control (MOPC) technique for VAD proposed by Leao T, et al. [34] was chosen as a proof of concept. For the correct operation of the MPOC technique of VAD is required the formation of the LUTs (value maps) of the estimators for each specific motor of the VAD to be controlled. In HPB, the map of values of the VAD prototype to be controlled by the MOPC embedded in the ILS was raised [39]. In Figure 15, the test in HPB for determining the LUTs of the estimators of a physiological control of VAD.

Discussion

The project of HPB construction presented satisfactory results, being highlighted that: the actuator was adequate; the mechanical construction was functional in the interaction of the actuator with the VAD; the fast prototyping reduced the cost and the necessary time for the project development; the mechanical components of HPB being modular allowed generalization of the use for performance evaluation of different types of developed VAD.

The data acquisition and system control project presented satisfactory results, being highlighted that The transducers presented reliable results; The PI control actuation was stable throughout the test, with less than 2% error in steady state (1.263 ± 0.211; with statistical significance, ƿ= 0.983; α= 0.05) between the set value and the effective value, which is an indication that in the performance of VADs tests for a long time the data will continue to be complete; The data acquisition was satisfactory with real-time sampling of the main variables necessary for the classification of the VAD performance (0.6 ± 0.1; with statistical significance, ƿ= 0.986; α= 0.05), with all data stored in spreadsheets.

Conclusion

The Hydrodynamic Performance Bench built in the Automation for Life Laboratory (ALL) of the Federal Institute of Sao Paulo campus São Paulo (IFSP SPO) was able to assist in the analysis of the development of VAD prototypes, based on this it is possible to affirm the technical success of the project.

In future works it will be designed the total automation system of the VAD hydrodynamic performance curves test. Also, the use of the particle image velocimetry module will be expanded.

Acknowledgments

The authors would like to thank the Adib Jatene Foundation (FAJ), Institute Dante Pazzanese of Cardiology (IDPC), Sao Paulo Research Foundation (FAPESP), Higher Education Personnel Improvement Coordination (CAPES), University of Sao Paulo (USP), and Federal Institute of Sao Paulo (IFSP) for partially supporting this research.

Conflict of Interest

The authors declare that they have no conflicts of interest with the contents of this article.

References

1. Ministerio Da Saude/Datasus Sistema De Informação Do Ms, Disponível Em: <Http://Www2.Datasus.Gov.Br/Datasus/Index.Php>, Accessed On: January 7th, 2018
2. Ghista, Dhanjoo N, and Anant P. Rao. "Mitral-valve mechanics—stress/strain characteristics of excised leaflets, analysis of its functional mechanics and its medical application." J Med Biol Eng 11 (1973): 691-702.
3. Cooley, Denton A. 100,000 hearts: a surgeon’s memoir. University of Texas Press, (2021).
4. DeVore, Adam D, Priyesh A. Patel, and Chetan B. Patel. "Medical management of patients with a left ventricular assist device for the non-left ventricular assist device specialist." JACC Heart Fail 5(2017): 621-631.
5. Bacal, Fernando, João David De Souza Neto, Alfredo Inácio Fiorelli, and Juan Mejia, et al. "II Diretriz brasileira de transplante cardíaco." Arq Bras Cardiol 94(2010): e16-e76.
6. Nosé, Yukihiko, M. Yoshikawa, S. Murabayashi, and T. Takano. "Development of rotary blood pump technology: Past, present, and future." Artif Organs 24(2000): 412-420.
7. Kirklin, James K, and David C. Naftel. "Mechanical circulatory support: Registering a therapy in evolution." Cir Heart Fail 1(2008): 200-205.
8. Miller, Leslie W, and Joseph G. Rogers. "Evolution of left ventricular assist device therapy for advanced heart failure: A review." JAMA Cardiol 3 2018): 650-658.
9. Nosé, Yukihiko. "Design and development strategy for the rotary blood pump." Artif Organs 2 (1998): 438-446.
10. Muradbegovic, Mirza, Steven Taub, Elena Rizzo, and Ludwig K. von Segesser,et al. "Ultimate test bench for pediatric biventricular assist device based on artificial muscles." Asaio Journal 57(2011): 62-67
11. Khan, Talha Irfan, Haris Sheh Zad, Ismail Lazoglu, and Ozlem Yalcin. "Development of a novel shrouded impeller pediatric blood pump." Artif Organs 21(2018): 142-149.
12. Brittai S.B, J.L. Hajjar, and S.S. Nidadavolu. "In-vitro hemostasis test platform." In 2011 IEEE 37th Annual Northeast Bioengineering Conference (NEBEC) (2011): 1-2.
13. Fonseca, Jeison, Aron Andrade, Denys EC Nicolosi and José F. Biscegli, et al. "Cardiovascular simulator improvement: pressure versus volume loop assessment." Artif Organs 35(2011): 454-458.
14. Leme, Juliana, Cibele Da Silva, Jeison Fonseca, and Bruno Utiyama da Silva, et al. "Centrifugal blood pump for temporary ventricular assist devices with low priming and ceramic bearings." Artif Organs 37(2013): 942-945.
15. Dias, Jônatas Cerqueira, Jeferson Cerqueira Dias, Marcelo Barbosa, and José Ricardo Corrêa de Sousa Sobrinho, et al. "Intelligent control system applied to in vitro ventricular assist device test bench for decision support." In 2018 13th IEEE International Conference on Industry Applications (INDUSCON) (2018) :1280-1284.
16. Nishida, Masahiro, Ryo Kosaka, Osamu Maruyama and Takashi Yamane, et al. "Long-term durability test of axial-flow ventricular assist device under pulsatile flow." J Artif Organs 20(2017): 26-33.
17. Hernandes, Mariana Maria Aparecida Pinto. "Análise do escoamento em um dispositivo de assistência ventricular." (2018).
18. Ring, Brock Philip, and Evan C. Lemley. "Design and implementation of a low-cost particle image velocimetry system for undergraduate Research and education." In 2014 ASEE Annual Conference & Exposition (2014): 24-365.
19. Bock, Eduardo, Pedro Antunes, Tarcisio Leao, and Beatriz Uebelhart, et al. "Implantable centrifugal blood pump with dual impeller and double pivot bearing system: Electromechanical actuator, prototyping, and anatomical studies." Artif Organs 35(2011): 437-442
20. Cavalheiro, André CM, Diolino J. Santos Fo, Aron Andrade and José Roberto Cardoso et al. "Specification of supervisory control systems for ventricular assist devices." Artif Organs 35(2011): 465-470.
21. Sá R.C.L., N.C.Cruz, J.R.Moro, T.F.Leão, and A.J.P. Andrade, et al. "Modification surface in medicine: Techniques with plasma in a centrifugal blood pump implantable." Sinergia 18(2017): 91-94.
22. Leão, Tarcísio, Jeison Fonseca, Eduardo Bock, and Rosa Sá, et al. "Speed control of the implantable centrifugal blood pump to avoid aortic valve stenosis: Simulation and implementation." In 5th IEEE RAS/EMBS International Conference on Biomedical Robotics and Biomechatronics (2014): 82-86.
23. Marini, Marisol. "Corpos biônicos e órgãos intercambiáveis: a produção de saberes e práticas sobre corações não-humanos." PhD diss., Universidade de São Paulo, (2018).
24. Silva, Evandro Drigo Da. "Nova abordagem na transmissão de energia transcutânea para dispositivos de assistência ventricular implantáveis." PhD diss., Universidade de São Paulo, (2018).
25. De Andrade, Gustavo C, Aron JP De Andrade, Jeison WG Da Fonseca, and Bruno U. Da Silva, et al. "Development of a Transventricular Assist Device (TVAD)–Influence of the Volute Vanes." IFAC-PapersOnLine  51(2018):158-161.
26. Bessa C.V. X., L.D.R. Ferreira, O. Horikawa, and S. Gama. "On the relevance of temperature, applied magnetic field and demagnetizing factor on the performance of thermomagnetic motors." Appl Therm Eng 145(2018): 245-250.
27. Dias, Jônatas C., Jeferson C. Dias, Marcelo Barboza, and José R. Sousa Sobrinho, et al. "Systemic model of cardiac simulation with ventricular assist device for medical decision support." In Doctoral Conference on Computing, Electrical and Industrial Systems(2018): 235-242.
28. Dias, Jeferson C., Jônatas C. Dias, Edinei Legaspe, and Rodrigo Lima Stoeterau, et al. "Treatment of ventricular assist device test bench data for prediction of failures and improved intrinsic reliability." In Doctoral Conference on Computing, Electrical and Industrial Systems(2019):183-190.
29. Bock, Eduardo, Adriana Ribeiro, Maxwell Silva, and Pedro Antunes, et al. "New centrifugal blood pump with dual impeller and double pivot bearing system: Wear evaluation in bearing system, performance tests, and preliminary hemolysis tests." Artif Organs 32(2008): 329-333.
30. Bock, Eduardo Guy Perpétuo. "Projeto, construção e testes de um dispositivo de assistência ventricular= bomba de sangue centrífuga implantável." (2011).
31. Bock, Eduardo Guy Perpétuo, Tarcísio Leão, Jeison Fonseca, and Aron Andrade. "Left ventricle failure and blood flow estimation for centrifugal blood pumps." Automation 6(2016): 162-166.
32. Bock, Eduardo Guy Perpétuo, Araki, S., Florentino, P. C. and, Ortega, F, et al. Dilatometria Da Zircônia Pelo Processo De Colagem De Barbotina Em Moldes De Gesso Para Eixos No Dav 2(2018): 113-119.
33. Bock, Eduardo Guy Perpétuo, Tarcísio Leão, Jeison Fonseca, and Aron Andrade. "Left ventricle failure and blood flow estimation for centrifugal blood pumps." Automation (2015): 162.
34. Leao, Tarcisio, Bruno Utiyama, Jeison Fonseca, Eduardo Bock, and Aron Andrade. "In vitro evaluation of multi‐objective physiological control of the centrifugal blood pump." Artif Organs 44(2020): 785-796.
35. Leão, Tarcísio F, Eduardo Bock, Jeison Fonseca, and Aron Andrade et al. "Modeling study of an implantable centrifugal blood pump actuator with redundant sensorless control." In Proceedings of the 2012 44th Southeastern Symposium on System Theory (SSST) (2012):174-178.
36. Santos, Bruno and Tarcisio Leão. "Implementação Em Hardware De Um Controlador proporcional-integral (Pi) para motor bldc utilizado em bomba de sangue centrífuga implantável (Bsci)". Available In: Https://Bit.Ly/2cfousw. (2017).
37. Santos B. and Tarcisio Leão. "Implementação de um controlador inteligente para microbombas de sangue centrífuga implantáveis (Mbsci) como dispositivo de assistência ventricular (Dav)". Available In: Https://Bit.Ly/2v4e3ps. (2018).
38. Santos, Bruno, Tarcisio Leão, and Eduardo Bock. "Intelligent control based on fuzzy logic embedded in fpga applied in ventricular assist devices (VADs)." In Proceedings of the 2019 4th International Conference on Robotics, Control and Automation (2019):138-143.
39. Dos, Santos and Bruno Jesus. "Controle inteligente embarcado para dispositivos de assistência ventricular."(2020)
40. Sá R.C.L., N.C. Cruz, J.R. Moro, T.F. Leão, A.J.P. Andrade, and E.G.P. Bock. "Modification surface in medicine: Techniques with plasma in a centrifugal blood pump implantable." Sinergia 18(2017): 91-94.
41. Saito M., Araki,S., Stoeterau, R., Bock, E. Estudo de análise superficial em materiais poliméricos para uso em biomedicina. Acad Soc J 1(2017): 60-67.
42. Florentino P., Bock, E. Study of tribology of bioinertes materials in blood pumps (2017).
43. Araki S, Stoeterau R., Marques A, Bock E and Florentino P. Friction in micro bearings of an implantable centrifugal blood pump (2017).
44. Marques, Júnior and Antonio Luiz. "Automatização de uma Bancada de Atrito e desgaste para seleção de materiais de uma Bomba de Sangue Centrífuga Implantável." (2016).
45. Rocha and Joaquim Antonio Ferreira da. "Microusinagem de óxido de zircônio pré-sinterizado e Peek no desenvolvimento de dispositivos de assistência ventricular." (2017).
46. Mamishev, Alexander V, Kishore Sundara-Rajan, Fumin Yang, and Yanqing Du et al. "Interdigital sensors and transducers." Proceedings of the IEEE 92 (2004): 808-845.
47. Leao, Tarcisio, Bruno Utiyama, Jeison Fonseca, and Eduardo Bock et al. "In vitro evaalution of multi‐objective physiological control of the centrifugal blood pump." Artif Organs 44(2020): 785-796.
48. Hernandes, M. M.A.P. Silva, E.D. Silva, M.B. Santos and Bruno. J et al. Development of a pediatric paracorporeal centrifugal blood pump. In: Asaio 65th Annual Conference, 2019, São Francisco. Asaio 2019 Annual Meeting Abstracts 65 (2019): 21-21.