Share this article on:

A Novel Approach to the In Vitro Hydrodynamic Study of the Aortic Valve: Mock Loop Development and Test

Vismara, Riccardo*†; Fiore, Gianfranco B.*; Mangini, Andrea†‡; Contino, Monica†‡; Lemma, Massimo‡; Redaelli, Alberto*; Antona, Carlo†‡

doi: 10.1097/MAT.0b013e3181d9c295
Adult Circulatory Support

The aortic root functional unit (ARFU) is a complex structure whose functions are strictly dependent on the biomechanical interaction among each of its anatomically defined elements. The classical approach to the in vitro study of aortic hydrodynamics does not take this complexity into account. We propose a novel methodology based on the possibility to house whole natural ARFU samples in a purposely designed pulsatile mock loop, allowing for aortic surgery simulation. To point out the usability and potentialities of the device, the mock loop was tested with untreated porcine ARFU samples and with one ARFU prosthesized with a state-of-the-art bioprosthesis. The sample holder design was proved to allow the clinician to house and treat the ARFU sample in the mock loop with easiness and repeatability. The valve leakage with the prosthesized ARFU was comparable with literature data, and Effective orifice areas were consistent with the constructor's data. In contrast, the recorded pressure drops exceeded the data from the manufacturer and were quite aligned with in vivo postop echo-Doppler data acquired in implant recipients. This result suggests that our apparatus and methodology provide a way to investigate aortic hydrodynamic phenomena that resemble in a close way to those taking place in the final recipients' circulation.

From the *Department of Bioengineering, Politecnico di Milano; †FoRCardioLab, Fondazione per la Ricerca in Cardiochirurgia ONLUS; and ‡Cardiac Surgery Unit, H.L. Sacco, Università di Milano, Milan, Italy.

Submitted for consideration November 2009; accepted for publication in revised form February 2010.

Reprint Requests: Riccardo Vismara, PhD, Department of Bioengineering, Politecnico di Milano, via Golgi 39, 20133 Milan, Italy. Email:

The in vitro approach to the study of hemodynamics relies on the use of mock loop devices to recreate realistic hemodynamic conditions in the laboratory. Its potential is well recognized by the scientific community.1–3 When dealing with new cardiovascular devices and/or procedures, performing preliminary in vitro investigations allows one to design precisely targeted in vivo tests, thus limiting the amount of animal experiments needed. These reasons make in vitro tests mandatory in the design/evaluation phase of devices, such as prosthetic heart valves (PHVs), vascular prostheses, and stents. In vitro methods can also conveniently contribute to address some specific issues related to novel surgical approaches. Past in vitro investigations concerning conservative/reparative surgical procedures applied on the mitral valve showed to be able to supply determinant quantitative information.1,4

It is commonly accepted that the aortic valve has to be considered as a functional biomechanical complex, wherein all anatomically defined elements (the ventricle-aortic junction, the leaflets, the sinus of Valsalva, and the sinotubular junction) play intimately interconnected roles, which result, in physiologic conditions, in the proper functioning of the so-called aortic root functional unit (ARFU).5,6 Recent computational investigations contributed to clarify the underlying theoretical framework, which involves a close interaction of the dynamics of the soft biological structures of the ARFU with one another and with the local hydrodynamics of blood.7–10 Such mutual interplay is somewhat unbalanced whenever any of these elements is altered due to pathological conditions (e.g., the leaflets are dysfunctional) or surgical correction (e.g., valve functionality restoring with conservative surgery or prosthesis implantation). Hence, an in vitro approach based on the use of whole ARFU samples is liable to provide sound experimental evidence and help drive the setting of theoretical investigations.

However, the classical in vitro approach to the study of aortic hemodynamics, dating back to 1960s, is historically linked to the evolution of mechanical PHVs11,12 and focuses only on the mechanical behavior of the prosthetic device. Through the years, this approach has been improved: new instruments (such as echo-Doppler and high-speed video13,14) were introduced, diverse aspects of the human circulation were investigated (pediatric or pulmonary hemodynamics, atrial fibrillation, and leaflet calcification),14–17 and specific adaptations were introduced for the evaluation of novel cardiovascular devices (endoluminal prostheses, stented valves, and rotary pumps).3,15,18 Still, very scarce attention was given to the biomechanics of the structures surrounding the aortic leaflets. Only recently, mock loops able to house biological aortic valve samples were designed.19–21 In these works, the ARFU integrity was partially maintained, and they can be considered as first steps toward a novel approach to the in vitro study of the aortic valve fluid dynamics.

In this work, we describe our approach to the in vitro study of ARFU hemodynamics, based on the design and development of a dedicated mock loop able to house whole ARFU samples and to allow surgery to be performed directly on them, with maneuvers familiar to the clinicians. The mock loop was equipped with an owned programmable pulse duplicator, an adjustable hydraulic afterload, fluid-dynamic measurement systems, and a high-speed camera, so as to allow biomechanical and fluid-dynamic evaluation of natural and modified configurations of the ARFU with pulsatile flow. The mock loop usability was evaluated with four untreated porcine ARFUs, as well as with one sample with a state-of-the-art bioprosthesis implanted, in comparison with classical in vitro data or in vivo postop measurements from the literature.

Back to Top | Article Outline

Materials and Methods

Design Criteria

The main design specifications of the experimental apparatus were as follows. 1) The mounting procedure of ARFU samples should be simple and easily repeatable. The sample housing section should allow surgeons to simulate surgery on the sample directly, so as to allow performing tests with the sample untreated, then apply surgery and run the tests again without dismounting. The in vitro conditions for surgery simulation should be similar to those of the operating room in terms of access to the valve components. The housed sample should be easily accessible, even with the mock loop in operation, for on-the-run adjustment. 2) For the sake of versatility, the sample housing section should be adaptable to samples of different size. 3) The ARFU samples under test should be subjected to physiologic-like mechanical conditions. 4) Pre- and postvalvular fluid dynamics should replicate ventricular and postvalvular hemodynamics during the whole simulated cardiac cycle. 5) The simulator should be suitable for studying the kinematics and dynamics of valvular opening and closure by including view accesses for video instrumentation.

Back to Top | Article Outline

Architecture of the Simulator

A schematic of the simulator and an overall photograph are shown in Figure 1. Letters a through e identify the main components of the simulator. The sample holder (detail a in Figure 1, also shown in Figure 2A) features two parallel polymethylmethacrylate plates, whose distance is adjustable (range 25–70 mm) to fit to the ARFU sample length. An annular, dacron-reinforced silicone patch is fixed to each plate by a bayonet coupling; the ARFU sample is meant to be sutured to such patches to adequately interface the natural tissue with the stiff structure. The sample holder is then submerged in the main reservoir (detail b in Figure 1, also shown in Figure 2B) and fixed to its bottom with another bayonet coupling. This connects the sample's interior with the ventricular chamber below, keeping the sample's outer surface wet and at atmospheric pressure. The main reservoir also serves as an atrial reservoir, connected to the ventricular chamber (detail c, Figure 1) through a one-way service valve (in red in Figure 1), acting as the mitral valve. Such close connection of the two chambers without interposition of any additional conduit minimizes the fluid inertia during diastolic filling. A polypropylene membrane in the ventricular chamber separates the working fluid (saline solution) from the service fluid (deionized water) of the pulsatile pumping device (detail d, Figure 1).22 Downstream of the ARFU sample, a three-element afterload (detail e, Figure 1) simulates the systemic hydraulic impedance.22

The simulator was mainly made up with transparent polymethylmethacrylate by means of traditional and computer numerical control machining. Visual accesses were provided to allow video recordings to be taken both from the aortic and the ventricular sides of the valve.

Back to Top | Article Outline

Experimental Assessment of the Simulator

Porcine ARFU samples were hosted in the simulator by surgeons to evaluate device usability, and simulated surgical sessions were performed on them to assess that the simulator design implied no limitations to surgical work. The potentiality of the simulator in evaluating valvular hemodynamics was checked by means of experimental tests on fresh porcine ARFU samples. Samples were harvested from fresh swine hearts obtained from the local abattoir and the calibers of the aortic annuli were measured via no-go gauges with a 1-mm diameter resolution. Four untreated samples where tested: diameter range 21–24 mm, mean 22.25 mm. For each sample, different hydrodynamic conditions were set with the simulator: series 1) beat rate: f = 70–110 bpm (step 10 bpm) with an imposed stroke volume of 70 ml; series 2) imposed stroke volume: 70–98 ml (step 7 ml) at f = 70 bpm.

A state-of-the-art pericardial bioprosthesis (21-mm tissue annulus diameter Carpentier-Edwards 2900 Perimount) was implanted in one ARFU sample, which underwent hydrodynamic tests with the same protocol. The experimental results obtained with the prosthesized ARFU were compared with literature data from conventional in vitro tests carried out by the manufacturer and echo-Doppler data from human recipients of the same PHV.

For the experimental campaign, the simulator was equipped with piezoresistive pressure transducers (PC140 series, Honeywell Inc., Morristown, NJ) and with an ultrasound flow meter (HT110R, Transonic Systen Inc., Itacha, NY) for hydrodynamic measurements (pressures P1, P2 and flow rate Q in Figure 1, left). Data were acquired with an A/D converter (USB-6009, National Instruments, Austin, TX). For all the experiments, the effective orifice area (EOA) was calculated according to the following equation23:

where Qrms (L/min) is the root mean square systolic flow rate, ΔPm (mm Hg) is the mean systolic pressure drop across the sample, and k is a conversion factor (k = 3.1 to yield the EOA in cm2). Valve leakage was estimated by extracting the telediastolic backflow (tDBF) from the measured flow rate tracings and by integrating the tDBF over the diastolic interval to obtain the valve leakage volume (VLV).

High frame rate videos displaying the opening-closing kinematics of the physiologic and surgically treated ARFU samples were recorded with a charge-coupled device camera at 1200 fps and 680 × 480 resolution (Phantom Miro2, Vision Research, Wayne, NJ).

Back to Top | Article Outline


Each procedure concerning sample mounting and surgery simulation (mimicking the surgical maneuvers of reparative approaches and prosthesis implantation) was carried out at least three times by four experienced surgeons and deemed satisfactory by the clinicians in terms of manageability, control, and repeatability of the parameters of the surgical procedure.

Figure 3 reports an example pressure tracing as measured at the inlet of the hydraulic afterload (P1 in Figure 1) together with the corresponding flow rate course for a 21-mm untreated ARFU sample at f = 70 bpm, with a mean flow rate Q = 4 L/min. In general, simulations pointed out behaviors that resemble the physiological ones. The only exceptions are the flow oscillations that take place at early diastole; these are a transient effect of the interaction between the inertia of the fluid downstream of the valve and the elastic nature of the ARFU walls, stimulated by the sudden valve closure. Because mass oscillations are quite dampened in the second half of the diastole, the tDBF was assumed as a reliable indicator of valve leakage.

The leak flow measured for the untreated samples was negligible. The VLV values calculated for the prosthesized ARFU lied in the range 5.8–9.4 ml, as those previously obtained from in vitro tests with a pericardial bioprosthesis of the same family.24,25

Figure 4 plots the systole-averaged pressure drop, ΔPm, measured across the samples versus the mean flow rate Q. Pressure drop values for the untreated ARFU samples were lower than those measured with the prosthesized ARFU: values (mean ± SD) range from 2.6 ± 1.3 mm Hg for the 24-mm sample to 6.2 ± 0.99 mm Hg for the 21-mm sample, compared with 21.3 ± 4.5 mm Hg for the prosthesized ARFU. The pressure drops we measured for the prosthesized ARFU were sensibly higher if compared with the in vitro values reported by the manufacturer (range: 1.2 ± 0.6 mm Hg at Q = 2.1 L/min to 11.2 ± 0.4 mm Hg at Q = 7.1 L/min). Conversely, our values were quite comparable with those from the literature, recorded in vivo on humans by means of transesophageal echo-Doppler examinations [range: (13 ± 5) − (20.3 ± 9.08) mm Hg].26–30

Effective orifice areas for the untreated ARFU samples ranged from 3.05 ± 0.36 cm2 for the 21-mm sample to 4.66 ± 0.77 cm2 for the 24-mm sample. The EOA for the prosthesized ARFU was 1.84 ± 0.20 cm2, in accordance with the data from the manufacturer [range: (1.8 ± 0.0) − (2.0 ± 0.4) cm2], although with some overestimation with respect to the in vivo data available from the literature for the same bioprosthesis [range: (1.1 ± 0.2) − (1.47 ± 0.36) cm2].27–30

Table 1 reports the overall spectrum of the values obtained in the tests for VLV, ΔPm, and EOA for all tested samples (mean ± SD).

Both the ventricular and the aortic visual access provided for the high-speed visualization were successfully tested. Figure 5 reports some representative snapshots from a video acquired from the aortic side of a 21-mm untreated sample at 1200 fps.

Back to Top | Article Outline


Several aspects of the functional anatomy of the ARFU have been understood only in very recent years.6 With this work, we propose an advanced in vitro approach to the functional study of the ARFU, based on the possibility to house a complete natural ARFU sample, untreated or with surgery applied, in an experimental apparatus designed ad hoc. It is partially inspired by those proposed by Haaf et al.20 in which only the aortic leaflets were housed in a mock loop, Nötzold et al.19 in which ARFU samples were tested in vitro, without any surgical treatment, and Erasmi et al.21 in which sparing operations were tested in vitro although in a very simplified apparatus. Unlike previous works in the field, however, our apparatus was designed for a broad range of applications, involving a wide spectrum of possibilities concomitantly: a) to easily house whole ARFUs of different sizes; b) to replicate and manage physiologic and pathologic fluid dynamic working conditions; c) to reproduce ARFU surgery in vitro in a surgeon-friendly manner and to test its effects on the biological structure with the possibility d) to adjust surgical parameters on the run and e) to investigate ARFU kinematics with high-speed video and echo-Doppler capability.

The replacement of the natural valve cusps with a bioprosthesis was taken as a paradigmatic test case to get an objective evaluation of the potentialities of our device. We calculated the most classical hydrodynamic evaluation indices (pressure drops, EOA, and valve leakage) for untreated ARFU samples and for an ARFU sample with the bioprosthesis implanted. The experimental results with the untreated ARFUs showed to be biomechanically consistent, with EOA values increasing with sample diameter and mean pressure drops values increasing with mean flow rate and decreasing with sample diameters; pressure drops and valve leakage were low in all the tested configurations, as awaited. In the in vitro tests of the prosthesized ARFU, EOA values were very similar to those provided by the manufacturer and comparable with data from in vivo measurement taken from the literature. Valve leakage was higher than that evaluated on nontreated samples in all the tested configurations. Leakage data for the prosthesized ARFU were comparable with data from the literature related to in vitro tests of the sole pericardial bioprosthesis.24,25 This evidence suggests that leakage was not significantly influenced by the interaction between the stented prosthesis and the surrounding biological structure.

In turn, experimental data of pressure drops for the prosthesized ARFU showed that the interaction between the relatively stiff structures of the stented prosthesis and the biological structure of the ARFU affects the energy dissipation of the fluid flowing through the sample. The measured pressure drops were higher than those provided by the prosthesis manufacturer, which were recorded in in vitro tests where the bioprosthesis is housed in a stiff valve holder, according to international standards recommendations. On the other hand, the mean pressure drops we recorded showed to be comparable with the pressure drops measured by means of echo-Doppler examinations on humans after implantation of the same prosthesis.26–29 This suggests that our in vitro set-up better replicates the conditions the prosthetic device has to face in vivo. Indeed, it is conceptually different to test a prosthetic device in vitro in a classical setting—the final result being a characterization of the biomechanics and fluid dynamics of the prosthesis—or to test the prosthesis in vitro once implanted in an ARFU—the result being now a characterization of the biomechanics and fluid dynamics of the prosthesized ARFU, whose behavior depends on the interaction between the artificial and biological structures. Although the data derived by a classical characterization is of undoubted utility (e.g., for design/evaluation purposes), a whole-ARFU characterization leads to data with a higher level informative content, which is liable to constitute a touchstone for biomechanical modeling and possibly enhance the comprehension of the theoretical implications involved. In view of a future use of our apparatus to investigate the hydrodynamics of reparative/conservative surgery, similar considerations apply; the possibility to apply surgery to a previously tested physiological sample, and to test it again without dismounting, will allow straightforward, pre-post biomechanical studies of novel surgical techniques.

A further major goal of our project was to develop a lab tool for surgical decision support, and the judgment in this respect was committed to experienced surgeon professionals. The sample holder design was proved to allow the clinician to house and treat the ARFU sample in the mock loop with easiness and repeatability. The open-fashioned design of the ARFU holder, with the external ARFU wall directly submerged in an open-to-air reservoir, makes the anatomical structure accessible to the surgeon even when the mock loop is in operation. Compared with other previous solutions,21 this endows the device with great versatility, allowing for a continuous monitoring and adjustment of surgery parameters. Indeed, an immediate acknowledgment about the functional response of the valvular structure to surgical maneuvers may amplify the surgeon's awareness considerably. The high-speed imaging apparatus and echo-Doppler capability of the mock loop further boost these possibilities.

From a technical viewpoint, the presence in the mock loop of the whole valvular complex introduces new issues concerning the correct management of the in vitro hydrodynamics. One of these is related to the interaction between the inertance of the hydraulic conduits downstream of the ARFU sample and the natural compliance of the sample walls. This interaction results in the mass oscillation that was recorded in the flow rate tracings just after the closure of the valve. Hence, valve insufficiency was evaluated using the signal measured at late diastole, when the mass oscillation dampens substantially. The future developments foreseen for the experimental apparatus will involve a design refinement of the sample holder and of the downstream conduits, aimed at removing the unwanted diastolic flow oscillations.

Another limitation of this approach is related to the use of biological samples, which intrinsically challenges the repeatability of the experimental sessions, because of samples variability and the operator-induced mounting variability. An appropriate statistical approach to the interpretation of the experimental results is mandatory in this respect. However, great efforts were spent in reducing the arbitrariness of human intervention by devising a standardized mounting procedure, which was deemed optimal by the surgeon professionals who served as evaluators. The use of a single prosthesized sample is a specific limitation of this study. Nevertheless, it should be considered that the aim of this experimental campaign was to assess any limitation of the simulator and to point out its potentiality and not to systematically evaluate the hydrodynamic performance of a state-of-the-art bioprosthesis.

Back to Top | Article Outline


Supported by FoRCardio Foundation (Fondazione per la Ricerca in Cardiochirurgia), Milan, Italy.

The authors thank Edwards Lifesciences for supplying data concerning in vitro tests of the bioprosthesis and Dr. Dario Comparolo for technical support.

Back to Top | Article Outline


1. Jimenez JH, Soerensen DD, He Z, et al: Mitral valve function and chordal force distribution using a flexible annulus model: An in vitro study. Ann Biomed Eng 33: 557–566, 2005.
2. Yoffe B, Vaysbeyn I, Urin Y, et al: Experimental study of a novel suture-less aortic anastomotic device. Eur J Vasc Endovasc Surg 34: 79–86, 2007.
3. Legendre D, Fonseca J, Andrade A, et al: Mock circulatory system for the evaluation of left ventricular assist devices, endoluminal prostheses, and vascular diseases. Artif Organs 32: 461–467, 2008.
4. He S, Fontaine AA, Schwammenthal E, et al: Integrated mechanism for functional mitral regurgitation: Leaflet restriction versus coapting force: In vitro studies. Circulation 96: 1826–1834, 1997.
5. Hopkins RA: Aortic valve leaflet sparing and salvage surgery: Evolution of techniques for aortic root reconstruction. Eur J Cardiothorac Surg 24: 886–897, 2003.
6. Cheng A, Dagum P, Miller DC: Aortic root dynamics and surgery: From craft to science. Philos Trans R Soc Lond B Biol Sci 362: 1407–1419, 2007.
7. Conti CA, Votta E, Della Corte A, et al: Dynamic finite element analysis of the aortic root from MRI-derived parameters. Med Eng Phys 32: 212–221, 2010.
8. Weinberg EJ, Shahmirzadi D, Mofrad MR: On the multiscale modeling of heart valve biomechanics in health and disease. Biomech Model Mechanobiol 2010 (in press).
9. Beller CJ, Labrosse MR, Thubrikar MJ, Robicsek F: Finite element modeling of the thoracic aorta: Including aortic root motion to evaluate the risk of aortic dissection. J Med Eng Technol 32: 167–170, 2008.
10. Soncini M, Votta E, Zinicchino S, et al: Aortic root performance after valve sparing procedure: A comparative finite element analysis. Med Eng Phys 31: 234–243, 2009.
11. Westerhof N, Elzinga G, Sipkema P: An artificial arterial system for pumping hearts. J Appl Physiol 31: 776–781, 1971.
12. Cornhill JF: An aortic-left ventricular pulse duplicator used in testing prosthetic aortic heart valves. J Thorac Cardiovasc Surg 73: 550–558, 1977.
13. Verdonck PR, Dumont K, Segers P, et al: Mock loop testing of On-x prosthetic mitral valve with Doppler echocardiography. Artif Organs 26: 872–878, 2002.
14. Zapanta CM, Dourte LM, Doxtater BJ, et al: Mechanical heart valve performance in a pulsatile pediatric ventricular assist device. ASAIO J 53: 87–96, 2007.
15. Vismara R, Lagana K, Migliavacca F, et al: Experimental setup to evaluate the performance of percutaneous pulmonary valved stent in different outflow tract morphologies. Artif Organs 33: 46–53, 2009.
16. Mouret F, Garitey V, Bertrand E, et al: In vitro atrial flow dynamics: Normal conditions versus atrial fibrillation. J Biomech 37: 1749–1755, 2004.
17. Bakhtiary F, Dzemali O, Steinseiffer U, et al: Hydrodynamic comparison of biological prostheses during progressive valve calcification in a simulated exercise situation. An in vitro study. Eur J Cardiothorac Surg 34: 960–963, 2008.
18. Giridharan GA, Pantalos GM, Gillars KJ, et al: Physiologic control of rotary blood pumps: An in vitro study. ASAIO J 50: 403–409, 2004.
19. Notzold A, Scharfschwerdt M, Thiede L, et al: In-vitro study on the relationship between progressive sinotubular junction dilatation and aortic regurgitation for several stentless aortic valve substitutes. Eur J Cardiothorac Surg 27: 90–93, 2005.
20. Haaf P, Steiner M, Attmann T, et al: A novel pulse duplicator system: Evaluation of different valve prostheses. Thorac Cardiovasc Surg 57: 10–17, 2009.
21. Erasmi A, Sievers HH, Scharfschwerdt M, et al: In vitro hydrodynamics, cusp-bending deformation, and root distensibility for different types of aortic valve-sparing operations: Remodeling, sinus prosthesis, and reimplantation. J Thorac Cardiovasc Surg 130: 1044–1049, 2005.
22. Lanzarone E, Vismara R, Fiore GB: A new pulsatile volumetric device with biomorphic valves for the in vitro study of the cardiovascular system. Artif Organs 33: 1048–1062, 2009.
23. Walker PG, Yoganathan AP: In vitro pulsatile flow hemodynamics of five mechanical aortic heart valve prostheses. Eur J Cardiothorac Surg 6(suppl 1): S113–S123, 1992.
24. Gerosa G, Tarzia V, Rizzoli G, Bottio T: Small aortic annulus: The hydrodynamic performances of 5 commercially available tissue valves. J Thorac Cardiovasc Surg 131: 1058–1064, 2006.
25. Bottio T, Tarzia V, Rizzoli G, Gerosa G: The changing spectrum of bioprostheses hydrodynamic performance: Considerations on in-vitro tests. Interact Cardiovasc Thorac Surg 7: 750–754, 2008.
26. Khan SS, Siegel RJ, DeRobertis MA, et al: Regression of hypertrophy after Carpentier-Edwards pericardial aortic valve replacement. Ann Thorac Surg 69: 531–535, 2000.
27. Dellgren G, David TE, Raanani E, et al: Late hemodynamic and clinical outcomes of aortic valve replacement with the Carpentier-Edwards Perimount pericardial bioprosthesis. J Thorac Cardiovasc Surg 124: 146–154, 2002.
28. Goetzenich A, Langebartels G, Christiansen S, et al: Comparison of the Carpentier-Edwards Perimount and St. Jude Medical Epic bioprostheses for aortic valve replacement—A retrospective echocardiographic short-term study. J Card Surg 24: 260–264, 2009.
29. Eichinger WB, Botzenhardt F, Keithahn A, et al: Exercise hemodynamics of bovine versus porcine bioprostheses: A prospective randomized comparison of the mosaic and perimount aortic valves. J Thorac Cardiovasc Surg 129: 1056–1063, 2005.
30. Rosenhek R, Binder T, Maurer G, Baumgartner H: Normal values for Doppler echocardiographic assessment of heart valve prostheses. J Am Soc Echocardiogr 16: 1116–1127, 2003.
Copyright © 2010 by the American Society for Artificial Internal Organs