Pressure Myography and Cardiac Flow Simulator for Mechanical Characterization of Native and Engineered Blood Vessels

Summary:

Pressure myography, which is the standard for assessing vascular mechanics and vasoreactivity, is expensive, has low throughput, and is limited to static fluid flow. Here, we developed HemoLens, an open-source 3D-printed pressure myography system for < $750. HemoLens features compact micromanipulators, incremental in-line pressure control, physiological temperature regulation, and modular pulse pressure control between normotensive and hypertensive levels. HemoLen’s efficacy was demonstrated by delineation of physiological reactivity and pathological mechanical phenotypes using native mouse arteries and bioprinted acellular scaffolds. Engineered hypertensive vessels demonstrate increased burst pressure (464 mmHg) and reduced dynamic compliance reminiscent of diseased arteries. Together, HemoLens lowers the barrier to entry in pressure myography research by serving as a comprehensive low-cost system for native and engineered vessel characterization.

Graphical Abstract

INTRODUCTION

Arterial vasoreactivity which is the ability of small-diameter blood vessels to regulate luminal diameter via constriction and dilation in response to pharmacologic and endogenous stimuli plays a crucial role in the systemic blood pressure control^1,2. These vasoactive blood vessels are unique in that their cellular and extracellular matrix (ECM) components are structured to withstand the hemodynamic forces applied by changing blood flow and pressure during the cardiac cycle^3,4. Given that arterial stiffness is a defining characteristic of hypertension, the resulting increased blood pressure subsequently leads to downstream effects such as endothelial dysfunction and vascular smooth muscle cell dedifferentiation^5,6. This drives pathological ECM remodeling that promotes systemic vessel stiffening from overproduction of fibrillar collagen, wall thickening, and reduced luminal diameters limiting the vessels’ ability to expand and contract⁴.

Animal models, particularly mice, have been instrumental in advancing our understanding of hypertension and cardiovascular conditions; however, these models often fail to accurately replicate human clinical outcomes^7,8. As a result, there is growing interest in tissue engineering approaches which employ human-specific cells and biomaterials to create biologically relevant models better mimicking human pathophysiology^9,10. One such approach, 3D bioprinting, enables the precise spatial deposition of ECM-based biomaterials to fabricate a range of tissue-engineered blood vessels (TEBVs)¹¹, including acellular vascular conduits¹², cellularized scaffolds¹³, and perfusable microvascularized functional tissues^14,15 for both disease modeling and regenerative medicine applications. However, as the physiological relevance of TEBVs improves, there is a growing need for affordable implementation of standardized, reproducible methods to benchmark their biomechanical properties against native tissues. Moreover, since mechanical dysfunction such as increased arterial stiffness is a hallmark of vascular diseases like hypertension, accurate quantification of the mechanical properties of both native and engineered vessels is essential.

Clinical in-vivo techniques such as pulse wave velocity and Coronary Flow Velocity Reserve (CFVR) provide information on the mechanical behavior of blood vessels, but do not yield direct insight into the underlying biomechanics of the vessel wall^16,17. While ex-vivo characterization allows for tight environmental control to discriminate between cellular and extracellular remodeling processes¹⁸. Across both native and engineered blood vessels, a range of mechanical testing techniques have been utilized to characterize material properties^19–21; nonetheless, pressure and wire myography remain the gold standards for evaluating vascular mechanics and graft compatibility, providing quantitative measures of burst pressure and vascular tone^21,22. Although widely used, commercial myograph systems are prohibitively expensive (often exceeding $40,000), low in throughput, and lack modularity or options for customization. Recent open-source myography systems, such as “Vasotracker”, provide simplistic hardware design and excellent software for monitoring vessel mechanics, but still require custom CNC manufacturing and standalone microscopic imaging systems driving costs above $6000²³. Thus, there is an emerging need for a customizable low-cost biomechanical myography platform designed for testing an expansive range of vessel diameter (0.1 – 5 mm) that uses affordable manufacturing processes, integrated and accessible imaging systems, and can evaluate both conventional burst pressure and functional vasoreactivity under physiological conditions.

We created a modular pressure myography platform called HemoLens with magnetically swapable components and high performance 3-axis micromanipulators built from 3D-printed and low-cost linear motion components. By designing a motorized dynamic pressure control system we enable quantification of vascular dynamic compliance in both native and Freeform Reversible Embedding Suspended Hydrogels (FRESH) 3D bioprinted engineered vessels. We validate HemoLens performance across a range of physiological pressures and vessel types to demonstrate its precision, reproducibility, and adaptability for vascular biomechanics studies. HemoLens combines integrated system design and open-source software to create a customizable open-source platform for high-fidelity vascular mechanical assessment.

RESULTS

Vascular Hemodynamic Forces and HemoLens Design Principles for Myography

When evaluating mechanical properties of native or engineered blood vessels the cellular and extracellular matrix structure, composition, and alignment drive function. In small diameter blood vessels such as mesenteric and carotid arteries, a tri-layered vascular wall provides mechanical strength and function. The outer collagen rich adventitial layer provides mechanical reinforcement to prevent vessel rupture and resist changes in intraluminal pressure (Figure 1A,B)^24,25. The middle tunica media comprises vascular smooth muscle cells (VSMCs), elastic laminins, and scaffolding proteins that control vascular tone and confer compliance against pulsatile hemodynamic forces (Figure 1A). The innermost tunica intima, contains the vascular endothelial cells (EC), which form a selectively permeable barrier along the lumen to sense and respond to changes in blood flow shear stress (Figure 1A,B)²⁶.

Figure 1:

Vascular Hemodynamics and HemoLens Overview A) Graphical representation of the three major layers of the blood vessel and their extracellular components. B) Graphical representation of the major hemodynamic forces experienced by resistance arteries. C) Schematic of Pressure myography showing maximum pressure required to burst a vessel, and representative cartoon graphs of burst pressure testing (i) and myogenic tone (ii). D) Wire myography schematic showing the rupture of the blood vessel and representative graphs of rupture testing (i) and myogenic tone (ii) experiments conducted using wire myography. E,F) CAD model and photo of the HemoLens system highlighting novel components of the system

To isolate the functional consequences of altered vascular biomechanical properties related to developmental and disease progression, researchers perform ex vivo characterization with either pressure and/or wire myography. Pressure myography involves cannulating excised vessels to either measure vascular tone in response to vasoreactive drugs at physiological pressure (120 mmHg) or determine the maximal amount of hydrostatic pressure achievable prior to vessel rupture. (Figure 1C). Alternatively, in wire myography, an excised vessel is cut into small sections, and two small wires are inserted into the lumen. The wires are then slowly pulled apart until reaching a tensile force that simulates physiological pressure. These devices can be used to assess vasoreactivity at a given tensile force or can be utilized to determine the vascular mechanical strength prior to vessel rupture (Figure 1D). While wire myography can be higher throughput than pressure myography, pressure myography more accurately recapitulates the fluid mechanics and circumferential strain of the vascular system. Additionally, since pressure myography is considered the ISO standard for clinical applications of vascular biomechanics²⁷, we chose to design and build HemoLens as a low-cost alternative to commercial pressure myography systems that would recapitulate vascular intraluminal pressure and shear stress under physiological conditions (Figure 1 E–F). Specifically, HemoLens facilitates 1) real-time systemic pressure monitoring, 2) optical vessel diameter tracking, 3) a variable intensity light source, 4) accurate XYZ micropositioning for easy vessel attachment, alignment, and tensioning, 5) intraluminal perfusion and media bath circulation at physiologically relevant temperatures, and 6) discrete static and dynamic pressure control. The overall design incorporated magnetic attachments to enable modular component addition, providing experimental flexibility and customization between native and engineered vessel mechanical testing. As with most open-source hardware design, our goal was to reduce the cost of the platform without sacrificing performance and reproducibility. In total, HemoLens costs $746 and requires no additional components other than a personal computer with USB input (Table S1). At this price point, multiple systems can be built and run in parallel to easily scale pressure myography towards a high throughput methodology.

Mechanical Design and Validation of HemoLens’ Components

Integration of physiological temperature control and perfusion

Temperature is an essential variable for physiological blood vessel function, including vasoreactivity, and preconditioning prior to mechanical testing^28–30. Ex-vivo vascular research is typically conducted at physiological temperature (37 °C) via superfusion, which uses a closed-loop pump system to continuously recirculate an externally heated solution, creating a controlled flow that perfuses stationary biological samples³¹. HemoLens’s vessel bath chamber is constructed from a heat-resistant 3D printed insert with a gasketed glass slide compressed in between the bath chamber and the HemoLens base plate to achieve water tightness and visible light illumination (Figure S1). Additionally, HemoLens was designed to allow for independent control of the bath chamber and vessel perfusion while maintaining physiological temperature for both systems. To achieve bath fluid recirculation, two perfusion ports were integrated into the 3D design to connect tubing via Luer lock coupling with a superfusion process (Figure S1). To prevent bath chamber deformation from direct heating, we used a low-cost water bath to warm a circulating media reservoir at 19 mL/min (Figure S1). Similarly, perfusion within the vessel is achieved using a low-volume pump with a controllable flow rate (Table S2), and temperature is maintained by submersion of the pump into the heated water bath (Figure 2A).

Figure 2:

Validation of HemoLens’ components. A) Temperature reading for both superfusion lines, bath (black), and vessel (grey). Both maintain this temperature with a stability of ± 0.32 °C over time. B) Micromanipulator highlighting movement axes. (i) Calculated linear movement for each axis. n=10 for all axes. Data are represented as mean ± SD, C) incremental pressure ratchet system (iPRS). Inset showing internal stopping mechanisms. n=3 for each syringe type. Data are represented as mean ± SD D) Camera gantry system with call-outs to important features. (i) Image showing a rubber test tube with Vasotracker lines overlayed. (ii) Linear elastic response of the rubber test vessels with incremental pressure increases. E) RAMP system with call-outs to important mechanical components. (i) Graphical representation of RAMP’s mechanism of action for cyclic pressure increases. The RAMP arm swings toward the arrested tubing (1). Contact with the tubing increases the systemic pressure of the vessel (2) as the RAMP arm continues to roll against the tubing. Maximum pressure is reached at the apex of contact between RAMP and the tubing (3). As the arm is released, pressure begins to decrease over time (not shown). F) Representative waveform of RAMP’s cyclic pressure generation set to physiological 80/120 mmHg (black) and pathological 80/160+ mmHg (red) at 60 beats per minute. Solid red and black bars are used to highlight the distance between baseline and peak pressures. G) Calculated mean pressure increase per RAMP pulse of physiological pressure range, 43.0 ± 3.16 mmHg (white, n = 279 peaks), and pathological pressure range, 89.2 ± 6.45 mmHg (red, n = 673 peaks). Data are presented as mean ± SD, **** = p<0.001. H) Calculated mean pressure peaks of physiological pressure range, 121.5 ± 2.8 mmHg (white, n = 279 peaks), and pathological pressure range, 167.4 ± 6.27 mmHg (red, n = 673 peaks). Data are presented as mean ± SD, **** = p<0.001

Next, fluid temperature equilibration time was measured to ensure a physiological temperature of 37 °C was maintained for vessel viability and vasoreactivity²⁸. To determine the equilibration time necessary to achieve physiological temperature stability for each perfusion line we measured both the circulating bath and vessel temperature for 1 hour. With a water bath temperature set point of 47 °C, HemoLens perfusion systems reached 37 °C within 40 minutes. A stable temperature of 36.63 ± 0.32 °C was maintained over an extended period (Figure 2Ai, See Note S1 for more information). With two individually controlled perfusion lines and rapid establishment of physiological temperature, HemoLens offers users quick experimental startup time and multiple avenues of superfusion for the maintenance of physiological temperature.

Design and testing of high-precision 3D-printed micromanipulators

In all myography systems, precise control of 3-axis positioning is essential to mount vessels onto the testing device to achieve the proper tension without damaging the vessel. Additionally, as the trend toward generating smaller diameter engineered vascular grafts continues, careful and precise manipulation of these vessels will only grow in importance^32,33. Commercial pressure myography systems employ optical grade micromanipulators, but implementation of these components into open-source devices increases the overall cost as they often are the most expensive component costing >$1000. Therefore, to lower the cost of HemoLens we designed and built a compact 3-axis (X, Y, Z) manual micromanipulator for ≤ $65 per unit by leveraging widely available miniature linear rails and carriages from the open-source 3D printing community (Figure 2B, Video 1). Beyond its 3-axis positioning, the micromanipulator incorporates customizable dovetailed arms that hold standard Luer Lock Birmingham needle gauges for vessel suturing (Figure 2B). This compatibility with standard gauges gives HemoLens the versatility to work with vessels that have inner diameters from 100 μm (pulled pipettes) to 5.5 mm (5-gauge) without additional modification (Table S3).

Commercial 3-axis micromanipulators report precision as the smallest travel distance producible per axis revolution. We employed the same metric to analyze the precision and repeatability for each of the 3 axes of HemoLens’ custom micromanipulators. A complete turn of each axis resulted in 0.99 ± 0.016 mm of average linear travel per axis, and a quarter turn resulted in 0.25 ± 0.011 mm of linear travel (Figure 2Bi). These experimental values closely match the expected 1 mm pitch of the M6 bolts used as axis leadscrews suggesting that there is minimal mechanical backlash and high repeatability. With these results, we highlight our manipulator as a low-cost, high-performance, and reliable alternative to commercial 3-axis micromanipulator stages well suited for vessel cannulation and myography applications.

Incorporation of real-time in line pressure monitoring

Control and monitoring of itraluminal pressure is the critical feature of pressure myography that separates it from other vasoactive and mechanical characterization techniques^22,27. Many commercial myography systems use proprietary processes to both monitor and automate pressure control; however, these engineered controls are expensive and are normally restricted to sub-burst pressure levels during experimentation. Moreover, since static pressure is often set at the start of an experiment and held constant throughout, we wanted to provide users with a simple, modular, and non-motorized pressure control system. To achieve real-time pressure monitoring we implemented a previously developed open-source pressure monitoring system from Vasotracker²³. Two in-line pressure sensors were attached to the vessel perfusion line before and after the vessel cannulation arms. The range and sensitivity of the flow-through pressure sensors can be tuned to measure small changes in physiologic vasoreactivity or expanded by adjusting the Wheatstone bridge and software configuration to characterize mechanical burst pressure (Table S4). Following calibration of the pressure sensors, repeatable pressure monitoring within ± 3.43 mmHg was achieved within the working range of the pressure sensors (< 500 mmHg) (Figure S1).

Design and validation of a manual incremental pressure system

Next, the HemoLens incremental Pressure Ratchet System (iPRS) was designed to function as a low-cost manually advancing syringe pump (Figure 2C). In a closed-loop manual pressure control system, as pressure rises, back pressure accumulates at the pressure-generating site. This pressure increase can push the syringe plunger backward, ultimately reducing the overall systemic pressure. To avoid this pressure loss during experimentation, an internal one-way locking “ratchet” mechanism was designed to prevent anti-clockwise movement when vessels are pressurized (Figure 2C insert, Video 2). A manual clutch disengagement system was incorporated at the top of the ratchet mechanism to allow easy syringe retraction to the starting position or to quickly depressurize the vessel. HemoLens’ iPRS ratcheting pressure control module is fully customizable, enabling users to select various syringe diameters for incremental pressure tuning.

To characterize the incremental pressure repeatability across a range of syringe diameters we measured discrete pressure intervals for 3 mL (8.66 mm ID) and 1 mL (4.78 mm ID) BD syringes. Rubber tubing was mounted and sutured onto HemoLens as a minimally compliant synthetic test vessel, and pressure was increased in a stepwise manner by turning the lead screw on iPRS. Each counterclockwise turn could be audibly perceived as a “click” to provide user feedback in response to incremental advances. Each click increased pressure by 64.37 ± 2.95 mmHg for 3 mL syringes and 7.62 ± 0.19 mmHg for 1 mL syringes. (Figure 2Ci). While this difference in pressure exceeds the theoretical ~4.6X change for an adiabatic process between these syringes, the variations in plastic syringe plunger style and overall syringe rigidity likely influence the experimentally observed pressure change. To modulate incremental pressure with iPRS, smaller syringe sizes, different thread pitches, number of teeth within iPRS’ ratcheting mechanism, and bolt diameters can be tuned to provide tailored pressure resolution per click if desired. With this iPRS system, we demonstrate both high precision and broad range control over pressure increases in a low-cost, compact design that does not require motorization.

Utilization of modular low-cost optical vessel tracking

Most myography devices employed for investigating vascular reactivity and mechanical properties utilize inner diameter (ID) and outer diameter (OD) optical imaging of the mounted vessel to compute changes in diameter, wall thickness, and compliance. To perform this function, commercial and open-source systems conventionally rely on external microscopy systems with in-line lighting solutions, which increases overall system cost and limits throughput based upon optical system avalibility^19,34. To eliminate optical imaging as a hardware constraint, HemoLens utilizes new advances and price reductions in CMOS sensors and LED lighting to achieve integrated, low-cost imaging. We designed a custom imaging gantry for an ArduCTablam USB camera with an M12 ArduCam Lens (Figure 2D, Video 3). The camera mount was then attached to a 50 mm linear rail for precise control of camera focus. A magnetic camera holder was designed to attach and remove the camera gantry from HemoLens’ base. (Figure 2D). This removable design improved vessel mounting by eliminating overhead obstruction during alignment and suturing. A variable intensity lighting system was incorporated into the HemoLens base for oblique or in-line lighting (Figure S1).

To ensure HemoLens’ reliability and performance as an optical pressure myography system, compliant rubber tubing (ID of ~1.4 mm, OD of ~3.5 mm) was mounted and optically tracked using the Vasotracker software (Figure 2Di). A stepwise pressure increase was applied via iPRS while monitoring vessel ID, OD, and inlet and outlet pressure (Figure S1). For the synthetic vessel, we observed a linearly increasing change in diameter (R² values of 0.98 ID and 0.98 OD) as a function of pressure up to 450 mmHg (Figure 2Di). These results demonstrate HemoLens’ capability to conduct conventional real-time pressure myography with a low-cost and modular CMOS sensor.

Generating cyclic physiological pulsatile pressure with RAMP

Pressure myograph platforms often rely on static pressure to measure vessel reactivity and burst pressure²⁰. This neglects the dynamic native vessel environment, characterized by cyclic pulsatile flow that generates pressure waves between 80 mmHg (diastolic) and 120 mmHg (systolic) in normotensive adult humans. In hypertensive pathological conditions, the systolic pressure can exceed 145 mmHg^1,35. To model these dynamics, a low-cost aquarium pump was used to establish closed-loop luminal flow through our cannulated test vessel with an estimated shear stress of 1.90 Dynes/cm^{2 36}. Systemic baseline pressure was achieved by attaching a custom 3D-printed occlusion collar to the outlet tubing to manually restrict flow until 80 mmHg was achieved. To incorporate pulsatile pressure into HemoLens, the Regular Adjustment of Modulated Pressure (RAMP) system was designed (Figure 2E). RAMP consists of a NEMA 17 stepper motor attached to a 3D printed bracket and occlusion arm that partially occludes the outlet tubing line via a swinging motion to create a brief pulsed increase in systemic pressure (Figure 2Ei, Video 4, See Note S2–3 for further information).

To confirm the repeatability and physiological accuracy of the RAMP system, pulsatile pressure readings were measured using a similar synthetic test vessel as above. When the outlet collar was tightened to achieve a 80 mmHg baseline, RAMP was programmed to create 60 beats per minute, or 1 Hz cyclic pressure increases in physiological (~80 to ~120 mmHg) and pathological (~80 to ~160+ mmHg) ranges by changing the RAMP arm length. Our results show mean peak pressures reached 121.5 ± 2.80 mmHg and 167.4 ± 6.27 mmHg, with corresponding pressure increases of 43 ± 3.16 mmHg and 89.2 ± 6.45 mmHg per pulse, respectively (Figure 2F–G). The measured frequency of the RAMP pulse wave sequence was empirically determined to be 60 beats per minute (bpm) or 1 Hz (Video 4), which is within the frequency range to mimic a normal human heartbeat^37,38. The slight deviation between our programmed beat rate and measured frequency during vessel testing is likely due to a brief delay in occlusion arm rotation during tubing compression.

RAMP’s tunable beat rate is set between 0–220 BPM via the C++ Arduino code or startup splash screen, while its maximum pulsatile pressure is controlled by adjust tube compression with variable length occlusion arms to accommodate various tubing sizes (Figure 2E–F, Figure S1). Therefore, as both the rate and degree of occlusion can be tuned, researchers can use RAMP to mimic physiological and pathological pressure conditions with peak pressure ranges between 90 – 220 mmHg (Figure S1), easily incorporating human heartbeat pulsatile pressure into their vascular research.

Native Vasculature Testing using HemoLens

Wildtype and Diseased Vessel Characterization with HemoLens

The pathological hallmark of hypertensive vessel stiffening is the overproduction of collagen and degradation of the elastin within the vessel wall following damage to the endothelium³⁹. The stiffer a vessel becomes, the more impaired its ability to dilate and contract to accommodate fluctuations in blood flow and pressure⁴⁰. To demonstrate HemoLens’ capability to elucidate static and dynamic responses to pressure changes in diseased vessels, carotid arteries were excised from both wildtype (WT) and Sickle Cell Anemic (SCA) mice. Arteries from SCA mice are known to exhibit endothelial damage-associated vascular remodeling and vessel stiffening and are expected to show hypertensive-like mechanical phenotypes^41,42 (Figure 3A). In this study, there were no differences between the inner and outer diameters of WT and SCA carotid arteries when measured by stereomicroscopy (Figure S2). Each excised mouse artery was mounted onto HemoLens for evaluation (Figure 3B, See Note S4–S5 for further information). To ensure the mounted vessels were vasoreactive post excision, excised WT mouse carotid arteries were used within 3 hours of excision and exposed to increasing concentrations of Phenylephrine (PE) and Acetylcholine (ACh) as previously described^43,44. A noticeable increase in constriction was observed at 10⁻⁶ M PE, reaching a maximum constriction at 10⁻⁵ M PE (Figure 3C). When exposed to ACh, vessel relaxation increased linearly with increasing concentrations of ACh (Figure 3D, See Note S6–S7 for further information). The highest range of the working concentration of ACh was deemed the point of maximum relaxation. However, true maximum relaxation experiments are beyond the scope of this work. These results highlight HemoLens’ ability to control the critical environment conditions (temperature, salt concentrations, and pressure) required to maintain vasoreactivity and ensure experimental repeatability⁴⁵.

Figure 3:

Native artery characterization with HemoLens A) Schematic showing carotid artery excision from mouse and mounted on HemoLens with important callouts. B) Images of mounted carotid arteries with and without Vasotracker trace lines overlayed (Inner diameter= red, outer diameter = blue). Vasotracker trace lines can be tuned by changing pixel size, altering camera magnification, and via algorithm selection. C) Normalized constriction percentage of Wiltdtype (WT) artery exposed to cumulative concentrations of Phenylephrine with maximum constriction occurring at 10–5 M. Data shown are from a single representative experiment for illustrative purposes, not biologically significant data. D) Normalized vessel relaxation percentage of WT artery exposed to cumulative concentrations of Acetylcholine. No maximum relaxation was recorded. Data shown are from a single representative experiment for illustrative purposes, not biologically significant data. E) Pressure-Diameter curves demonstrating a lower elastic response to internal pressure increases of stiffened Sickle Cell Anemic (SCA Stiff) vessels (red) compared to WT vessels (blue) show a peak. F) Representative wave forms of WT (blue) and SCA Stiff (red) vessels under physiological RAMP generated pressure waves at 60 beats per minute. Dashed lines represent inner diameter measurements. G) Average diameter change per pulse in each vessel. WT vessels show how higher changes per pulse for both inner and outer diameters, *** = p<0.01 H) Calculated dynamic compliance of each vessel under physiological RAMP generate pressure waves. WT vessels (0.0045 ± 0.0008 mmHg-1, n= 61) are significantly more dynamically compliant than SCA vessels (0.0012 ± 0.0002 mmHg-1, n= 42). Data are represented as mean ± SD, *** = p<0.01

Stepped intraluminal pressure increase is considered a gold standard for evaluating vascular mechanics such as strength, burst pressure, and compliance of excised and engineered vessels²⁷. Therefore, we conducted mechanical testing on our excised mouse vessels using HemoLens as previously described²². The intraluminal pressure of the mounted vessels was increased stepwise with iPRS and held static for 20 seconds at each pressure increment (See Note S6 for further information). Both WT and SCA vessels displayed an increase in inner and outer diameter as a function of increasing intraluminal pressure (Figure 3E). Interestingly, the SCA vessel’s diameter−pressure response curve plateaued early at 125 mmHg of internal pressure, whereas WT vessels showed higher distensibility and plateaued at 200 mmHg of internal pressure (Figure 3E). Vessel diameters remained unchanged beyond 200 mmHg for both conditions (Figure S2).

Wildtype vessels have higher dynamic compliance than diseased vessels

Blood flow within the cardiovascular system is dynamic. As blood pressure and heart rate vary cyclically, a vessels diameter distends and recovers in response to hemodynamic demands^46,47. This is a phenomenon described as dynamic compliance (DC)⁴⁸. RAMP was designed to generate cyclic pressure waves at human cardiac frequencies, facilitating dynamic compliance analysis of a diverse range of vasculature. The dynamic compliance of mouse carotid arteries was evaluated under physiological pressure variations induced by RAMP. WT vessels showed a higher average distention of 124.3 ± 23.59 μm for outer diameter and 129 ± 22.19 μm for inner diameter (Video 5–6). In contrast, SCA vessels only dilated to an average of 43.07 ± 8.50 μm OD and 40.62 ± 6.61 μm ID (Figure 3F–G). The pulse frequency in these native vessels was found to be 1.06Hz and 1.02Hz for WT and SCA, respectively. The calculated dynamic compliance for carotid arteries is significantly higher in WT vessels (4.445⁻³) compared to SCA vessels (1.160⁻³) (Figure 3H). This outcome, in conjunction with the incremental pressure results, underscores HemoLens’ capability to distinguish various disease relevant mechanical phenotypes. Furthermore, it enables the observation of the inherent pathological vascular stiffening under pulsatile pressure associated with SCA⁴².

Testing of Engineering Vascular Scaffolds using HemoLens

FRESH bioprinting of small diameter collagen vascular scaffolds

3D bioprinting is a biofabrication method that has gained immense popularity in recent years due to its ability to create patient specific geometry and control scaffold microstructure^49–51. To demonstrate HemoLens’ applicability beyond native vessels and its usefulness in biofabrication workflows, a small-diameter collagen vascular scaffold was designed as a proof-of-concept bioprinted test vessel. To increase experimental throughput and improve printability, we designed a single scaffold model containing five free-floating artery-like vessels supported within a collagen frame. Each individual vessel was designed to have a 450 μm inner diameter and a 1200 μm outer diameter (Figure 4Ai). The model was sliced into individual layers for machine pathing, visualized to confirm appropriate model settings, and exported as 3D printer Gcode (Figure 4Aii).

Figure 4:

FRESH printed scaffolds evaluated with HemoLens. A) Simplified workflow of FRESH biorprinting of test scaffolds. (i) CAD model showing designed 450 μm ID and 1200 μm OD. (ii) Visualized GCODE/Machine pathing. (iii) Graphical representation of FRESH bioprinting. (iv) Stereomicroscope image of FRESH printed test scaffold array. SB= 1000 μm. (v) Post processing measurements of the printed scaffold using the measurement tool in LasX v3.9 (Leica). OD = 1172 ± 25.88 μm (n=5), ID = 430 ± 11.88 μm (n=5). (vi) Photo showing excised test vessel during HemoLens mounting step. B). Burst pressure testing showing crosslinked collagen scaffolds have nearly double the burst pressure of native collagen vessels, 464.34 mmHg and 218.18 mmHg, respectively. C) Representative waveforms of each scaffold condition under RAMP generated pressure waves. Crosslinked collagen vessels under physiological pressure changes (blue). Crosslinked collagen vessels under pathological pressure conditions (80/180+ mmHg) (red), show a similar pattern to native collagen vessels at physiological conditions (black). D) Spectra graph of a Fast Fourier Transform (FFT) on the wave forms of C showing that crosslinked collagen vessel at pathological (red) and native collagen vessel at physiological (black) have similar frequencies, while no uniform frequency is seen with crosslinked collagen vessel at physiological conditions (blue). E) Calculated diameters change per pulse for native collagen vessel under physiological conditions and crosslinked collagen vessel under pathological conditions are shown to be non-significantly different. F) Calculated dynamic compliance for native collagen vessel under physiological conditions (3–4 ± 9.9–5 mmHg⁻¹, n= 140) is significantly more dynamically compliant and crosslinked collagen vessel under pathological conditions (6.23⁻⁵ ± 2.83–5 mmHg⁻¹, n= 92). Data are represented as mean ± SD **** = p<0.001.

To create small-diameter artery-like vascular scaffolds, we utilized the Freeform Reversible Embedding of Suspended Hydrogel (FRESH) 3D bioprinting process. FRESH enables the accurate fabrication of collagen-based scaffolds by extruding the biomaterial into a gelatin microparticle support bath (see Note S8 for further information)⁵². The support bath contains a pH buffer to neutralize the acidic collagen biomaterial, resulting in immediate gelation as the layers are printed. Once the printing process is complete, the temperature of the bath is raised to 37ºC allowing for non-destructive print release and retrieval⁵³.

Each vascular scaffold was FRESH printed as previously described from 70mg/mL of collagen-I⁵⁴. The printed scaffolds exhibit a patent, transparent lumen surrounded by a thick opaque (darker in image) outer wall (Figure 4Aiii–iv, See Note S9 for further information). Quantification of the vessel’s inner and outer diameters confirmed that our scaffolds are within a 5% deviation of the original design. The outer diameter was measured to be 1202 ± 21.60 μm, while the inner diameter was 420.6 ± 44.07 μm (Figure 4Av). Prior to experimentation, each individual vessel was dissected from the collagen frame and prepared for myography quantification.

Fabrication and mechanical validation of a bioprinted hypertensive disease model

Due to its excellent biocompatibility, collagen is one of the most widely used biomaterials in tissue engineering^55,56. However, scaffolds made from collagen alone, prior to cellularization, are weak and often require post processing modifications, such as chemical crosslinking to increase overall material strength⁵⁷. We hypothesized that crosslinking would enhance the printed scaffold’s mechanical properties, but it would decrease the dynamic compliance, resembling a stiffened disease phenotype. To test this, our printed vessels were crosslinked with 4% (v/v) paraformaldehyde (PFA). Crosslinking increased FRESH printed collagen vessel’s burst pressure, resulting in a doubling from 218 mmHg to a peak of 464 mmHg (Figure 4B). Vessels were then tested at physiologically relevant dynamic pulsatile pressures produced by RAMP. At a physiological rate, control native collagen vessels showed a consistent pulsatile frequency of 1.02 Hz (Video 7); however, crosslinked collagen vessels displayed a weak and nonuniform sporadic pulse frequency with reduced peak, indicating a stiffer phenotype similar to that of SCA vs. WT (Figure 4C–D, Video 8). After switching the crosslinked collagen vessel to a pathological dynamic pressure regime (peak systolic pressure of 160+ mmHg), we regained a consistent pulsatile frequency profile nearly identical to the native collagen vessels under physiological pressure (Figure 4C–D, Video 9). The average diameter change per pulse was negligible between native collagen vessels at the physiological rate and crosslinked collagen vessels at the pathological rate (Figure 4E). Since a higher pulsatile pressure was required to achieve a similar vascular wall distention, the calculated dynamic compliance of our crosslinked collagen vessel was significantly lower than the native collagen vessel, 6.23⁻⁵ and 3.04⁻⁴, respectively (Figure 4F). These data show that, with HemoLens, we can characterize mechanical phenotypic differences between altered engineered vascular scaffolds, and that by implementing FRESH bioprinting, we can mimic the loss of dynamic compliance observed in vascular disease progression.

Computational Modeling of Vascular Material Properties

Biomachnical characterization of wildtype and disease phenotypes for native and engineered vessels

Prior work by the Humphrey and Mecham Labs have led to constitutive models for vascular biomechanics describing the non-linear mechanical behavior of arterial vessels^58,59. In particular, aortic vessels demonstrate a stress versus stretch response curve containing three zones of interest (Figure 5A)⁵⁸. The first is a decreased incremental elastic modulus at low stretch when elastin dominates mechanical behavior, followed by a sharp increase in modulus at high stretch when collagen dominates the behavior. The intersection between these two regions is termed the physiological region⁵⁸. A balance between these mechanical properties is essential to maintain adequate compliance during pulsatile blood flow and resist vascular damage under high pressure. However, during vascular disease, as vessels become less elastic and stiffen due to increased collagen deposition, a shift in the slopes of these regions is observed^58,60. Therefore, to demonstrate the utility of HemoLens beyond standard diameter tracking, we used biomechanical modeling techniques to extract metrics of vascular function that are commonly used to assess arterial pathophysiology^61,62. Circumferential stress and stretch were calculated for each native sample, showing a much higher collagen dominant region in the SCA mouse vessel compared to the WT mouse vessel, highlighted by the sharp increase in circumferential stress at higher stretch ratios (Figure 5B). By fitting a microstructurally-motivated material model for blood vessels to the stress-strain data⁶³, we extracted material and structural stiffness values^64–67. The calculated material stiffness for the three different loading states (Figure 5C) and the calculated circumferential stress vs. strain behavior show an increased resistance to diameter expansion in the SCA condition compared to the WT condition. At low pressures (20 mmHg), stiffness is slightly higher in the WT sample, but at a normotensive pressure value, stiffness is elevated in the SCA sample. At hypertensive pressure (>120 mmHg), the SCA sample’s stiffness is elevated compared to the WT sample, as reflected by the sharp increase in slope of the stress vs stretch plot at higher values of stretch. Our biomechanical analysis suggests that the material behavior of our SCA vessels is more collagen-dominated than the WT control vessels, which aligns with the known vascular stiffening associated with sickle cell anemia⁴¹.

Figure 5:

Mechanical Analysis of Native Wildtype and Sickle Cell Anemia Mice Vessels and FRESH printed Scaffolds Tested on HemoLens A) Schematic showing the non-linear behavior of arterial vasculature. Adapted from Wagenseil et al 2009. B). Circumferential stretch and stress curve relationship highlighting the increased compliance of Wildtype (WT) vessels within the collagen-dominant region, denoted by the leftward shift of the Cell Anemic (SCA) vessels. C) Pressure diameter relationship of WT and SCA mice displaying characteristics of material stiffening in SCA mouse vessels. D). Circumferential stretch and stress curve relationships between FRESH printed acellular scaffolds. The leftward shift in the stress vs stretch curve is indicative of stiffening at comparable levels of stretch. E) Pressure diameter relationship of FRESH printed acellular scaffolds displaying characteristics of increased material stiffness of crosslinked vessels. F) Summary table of calculated material stiffnesses at various pressure ranges.

We also examined the material stiffness of our printed vessels treated with and without PFA crosslinking agent. As expected, there was a leftward shift in the stress vs stretch curve that was indicative of stiffening at comparable levels of stretch (Figure 5D). However, at comparable pressures, the degree of stretch was much less in the crosslinked samples (Figure 5E), which resulted in lower values of material stiffness for pressures of 90 and 120 mmHg (Figure 5F). Material stiffness at low pressure (20 mmHg) was elevated with crosslinking to the highest value of all samples tested, which appeared consistent with the decreased non-linearity of its response (Figure 5F).

Discussion

Tissue-engineered vascular scaffolds and Tissue-engineered blood vessels (TEBVs) must meet the mechanical requirements of the native vascular environment before they can be considered for clinical applications³. Therefore, mechanical characterization becomes a key metric in developing TEBVs, with intraluminal burst pressure being considered a gold standard²⁷. In addition to burst pressure, these scaffolds must remain elastic to maintain their dynamic compliance, expanding and contracting with internal pressure changes to regulate blood flow and maintain natural blood pressure with each pressure cycle. Thus, we developed HemoLens as a bespoke platform for the mechanical assessment and characterization of native vasculature and TEBVs. Its core technical innovation in system design is a microscope-free hardware platform that leverages custom-designed 3D-printed tools, allowing for sub-millimeter micromanipulation, customizable pressure generation via RAMP, and an affordable CMOS imaging system. HemoLens reduces equipment costs and enhances modularity. To power this hardware, we have generated open-source Arduino C++ code to improve calibration and reliability of Vasotracker pressure sensing protocol, integrated bath and vessel perfusion pump control, and implemented stepper motor-driven pulsatile pressure via RAMP.

As a low-cost system, HemoLens balances both performance and cost. For instance, switching from a plastic syringe used in this report to a glass syringe should provide a more linear relationship between syringe size and incremental pressure steps. For finer control over temperature regulation, thermistors and temperature probes can be added to HemoLens to establish real-time closed loop feedback control. This addition may require the use of larger or additional Arduino boards. Given that most components are fabricated from 3D-printed plastic, it is anticipated that some material wear will occur, particularly in the locking spring of iPRS’ ratchet mechanism. However, since HemoLens was designed for modularity and customization, researchers can address these limitations by modifying part design or simply by changing the plastic filament to more durable materials such as polycarbonate or carbon fiber PETG^68–70. Additionally, some of HemoLens’ design choices prioritize cost reduction at the expense of mechanical control. The iPRS ratchet system, for example, requires manual control, limiting experimental automation. Motorization of the manual iPRS pressure control can be achieved by attaching a stepper motor to the leadscrew driving the ratchet system. These types of modifications can be integrated into the electronics firmware and added to the Vasotracker open-source software to provide a fully automated experience. If researchers require higher precision micromanipulation, the resolution of our micromanipulators can be increased by employing higher quality lead screws with a finer thread pitch. To enhance the capabilities of HemoLens, future modifications can be implemented to incorporate load cell force sensors into either of the micromanipulators. This will enable the provision of force readouts, facilitating tensile strength testing, suture retention assessments, and recording tensile stress during myography.

Mouse models remain a major benchmark in vascular tissue engineering, with native mouse vessels used as the control group for many in-vivo studies⁷¹. To this end, we used HemoLens to evaluate and delineate the phenotypical differences between WT and SCA mouse vessels. We further provide an example of HemoLens’ applicability in vascular research by presenting evidence of vasoconstriction and vasodilation of WT vessels in response to Phenylephrine and Acetylcholine, respectively. Using standard myography protocols and the RAMP system, we show that WT vessels are more compliant than SCA vessels in both static and dynamic conditions, which agrees with previously published work^40,72. However, our isolated vessels did not reach critical intraluminal pressure to cause vessel rupture for true native burst pressure testing. This is likely attributed to the rupture pressure of mouse arteries being significantly higher than the 500 mmHg working range limit of the pressure sensors employed in our system⁷³. Further, since most vascular tissue engineering is dedicated to creating biological alternatives from humans⁷⁴, HemoLens was tuned to a frequency of 1Hz to closely mimic human in-vivo conditions³⁷. If researchers aim to test the maximum burst pressure or replicate the frequency of murine heart rate of 10Hz⁷⁵ for native or engineered vessels, HemoLens’ modularity allows users to upgrade the inline pressure sensors, adjust RAMP’s motor control to provide faster rotation, and tune the Wheatstone bridge dynamic range to meet their needs.

To overcome limitations in animal models like size and genetics, the field of tissue engineering is increasingly using vascular substitutes with human derived cells and materials to study human-specific diseases^7,76–79. For this reason, we FRESH printed collagen-I vascular scaffolds as a proof-of-concept, small-diameter platform to study hypertensive disease progression. To compensate for the softness of collagen scaffolds, we designed our scaffolds to have a wall thickness larger than the reported wall thicknesses of native arteries, but still within ranges of recently published work on acellular vessels^80,81. To direct our collagen scaffolds towards a human hypertensive mechanical phenotype, we chemically crosslinked them with 4% PFA, which is known to stiffen acellular scaffolds^82,83. This process resulted in a significantly higher burst pressure, and is within reported ranges of other ECM-based small diameter engineering vessels, such as fibrin and collagen^84–88. In addition to chemical crosslinking, cellularization of the non-crosslinked vessels should significantly improve the mechanical performance and shift the vessel phenotype toward more compliant native human vasculature. Collagen and elastin are two of the primary components that affect the elasticity of the vessel wall⁸⁰. Previous reports have demonstrated that elastin deposition and increased fibrillar collagen within vascular grafts significantly shift their phenotype toward a native human vasculature phenotype^89–92. In-vivo, Elliott et al. showed that cellular ingrowth of their endothelialized fibrin vessel after implantation had significantly increased the deposition of elastin, leading the remodeled conduit to have similar elasticity levels to that of native human aortas⁸⁹. Similarly, Huang et al. demonstrated that cultured smooth muscle cells onto PGA scaffolds within a biaxial stretching bioreactor had increased fibrillar collagen production, which resulted in higher suture retention and distention rates similar to human vasculature⁹¹.

Conclusions and Outlook

In this report, we present an open-source vascular characterization device called HemoLens. As a myography platform, HemoLens offers micromanipulation down to 250 μm, which rivals commercial systems in performance. Additionally, these capabilities are achieved at 12% of the cost of open-source alternatives and 3% of the cost of commercial systems. The modular imaging system tracks vessel diameter during experimentation, is magnetically removable, and works across a range of vessel sizes. Moreover, iPRS’ manual pressure pump design includes a 3D printed locking clutch that provides consistent and stable discrete pressure increases. Our designed RAMP system provides HemoLens with additional functionality beyond commercial or open-source systems (Table S5) by producing cyclic pressure waves that more closely recapitulate the dynamic environment of the body. Several relevant use cases of standard vascular characterization techniques were conducted on both native vessels and bioprinted scaffolds. While we focused on evaluation of small-diameter vessels in this work, the modular nature of HemoLens enables the characterization of larger-diameter scaffolds and blood vessels without modification. Moreover, although systems capable of delivering cyclic pressure waves for preconditioning and dynamic compliance testing have been developed previously^48,93, this manuscript details the design, construction, and testing of a complete low-cost myography system that incorporates pulse pressure control via intraluminal flow. Future work will focus on integrating force sensors and load cells into the HemoLens platform to expand its functional capabilities. In combination with the existing RAMP system, these enhancements will enable dynamic preconditioning and pulsatile pressure application to promote maturation of engineered vessels and assess implant readiness more comprehensively. Additionally, these upgrades to the system will support uniaxial tensile testing and provide quantitative data to inform constitutive mechanical models for computational analysis. Ultimately, the customizability of the HemoLens platform will allow for continued feature addition and improved user experience to serve as a standardized benchmark for evaluating the mechanical performance of novel biomaterials and fabrication strategies relative to native vascular tissue.

Resource Availability

Lead Contact:

Request for further information and resources should be directed to and will be fulfilled by the lead contact, Daniel J. Shiwarski (djs87@pitt.edu)

Materials Availability:

All STL/STEP files for 3D bioprinter hardware modifications, collagen constructs, and perfusion systems within in this manuscript are available under an open-source CC-BY-SA license at Zenodo.com (https://zenodo.org/records/17280307). Additional open-source 3D models can be found at www.ShiwarskiLab.com. This study did not generate any new unique reagents.

Data and Code Availability:

All custom Python scripts and software used are available at our lab’s GitHub account (https://github.com/STEL-Pitt-BioE/HemoLens). All data are available in the main text and the supplementary materials.

Materials and Methods

Device construction

HemoLens was ideated and modeled in Fusion360 (Autodesk). The primary structure of its components was 3D printed from Polylactic acid (PLA) or Polyethylene Terephthalate Glycol (PETG) with a BambuLab X1 3D printer. Commercial components, including screws, linear rails, and nuts, were added as needed.

Micromanipulator Control

Micromanipulator precision was measured as a function of linear movement repeatability in 1 mm and 0.25 mm increments, with an identical procedure employed for each of the micromanipulator’s three axes. HemoLens’ was placed flat on a laboratory bench. An erasable marker was used to demarcate one starting and three quarter-turn positions for the three knurl thumbs screws through which the X, Y, and Z bolts thread. A digital indicator (Mitutoyo model# S112S) was zeroed and placed perpendicular to the back of the manipulator on HemoLens’ vessel outlet side. The X-axis M6 bolt was manually turned to the first quarter-turn position, and the linear travel displayed on the digital indicator was recorded. This was repeated ten times; the digital indicator was reset to 0 mm after each recorded measurement. The X-axis bolt’s position was reset, and the experiment was repeated ten additional times with a full rotation. This procedure was replicated with the Y and Z-axis bolts, with the digital indicator placed on the corresponding side of the manipulator.

Vessel imaging

Prior to each experiment, an M12 Lens 1/2.3” (ArduCam #LN024) was attached to a 12MP USB camera module (ArduCam #BO433), and the module was attached to HemoLens’ camera gantry. The camera gantry was magnetically attached to the camera stand, which was then magnetically attached to one side of HemoLens’ base platform. Vessels are properly illuminated from below with a USB powered dimmable LED Strip (Topai #COBW1–480-5000K) that activates when HemoLens’ is turned on. With the vessels mounted, Vasotracker software (V1.4.0, vasotracker.com) is started, and OpenCV was used to establish the connection between the software and the camera system according to Vasotracker’s user manual³⁴. Focus was established by adjusting the camera gantry’s position up and down until a clear image of the vessel walls was visible within Vasotracker. Vasotracker’s diameter-tracking algorithm was engaged, which plotted vessel ID and OD. The image scale was adjusted in Vasotracker to corroborate previously measured nominal dimensions using a reference calibration slide.

In general, our accuracy in edge detection of the vessel wall and lumen diameter is determined by the pixel size within our images and the Vasotracker analysis parameters. From a hardware standpoint, our camera’s 12 MP sensor has a resolution of 3840(H) by 3032(V) pixels with a 1.55 micron physical pixel size. Using a calibration slide we measured our pixel size to be 2.4 micron/pixel. This yields an effective magnification of 0.65X. To perform the vessel diameter tracking, Vasotracker software utilizes an edge detection algorithm that looks for a sharp change in the boundary condition between the inner and outer wall that can be sensitive to sub-pixel changes. While many of the parameters can be tuned to meet specific needs, in practice, we find that our observed maximum error is ≤ 2X our pixel size allowing for minimum detection limit of ~ 5 μm. For most vessels this is adequate. However, if increased sensitivity to small changes in vessel diameter is needed, a higher magnification lens can be utilized.

Static Pressure Experiments

All iPRS experiments were conducted with 3 mL syringes (BD #309657). However, syringes (BD) up to 10 mL are compatible with the iPRS. Before vessel mounting, a 3 mL syringe (BD) was filled with 1x PBS and placed inside the ratchet system housing. The syringe was held via a designed magnetic clamp. A 3-way stopcock with 1/16” tubing was attached to the syringe and connected to a 30-gauge needle attached to the micromanipulators. All air was purged from the tubing lines prior to experimental start. For vessel mounting, all tubing lines were sealed by setting each stopcock to the closed position to ensure no air or water escaped during mounting. Vessels were mounted as described in Note S5. After the vessels were mounted, stopcocks were set to the open position, and the ratchet was continuously turned counterclockwise to plunge the syringe and perfuse 1x PBS through the vessel and into a collection beaker to remove any remaining air. The outlet was then closed, ensuring a baseline pressure close to 0 mmHg. Systemic pressure was increased in discrete increments by rotating the knurled handle of iPRS counterclockwise. The pressure was held for 20 second intervals and increased until either burst pressure or systemic pressure of ~500 mmHg was reached. The absolute maximum value of the pressure sensors is dependent on the potentiometer adjustments and specific sensors used. Empirically tested values may vary. Vessel diameters and systemic pressure were tracked using VasoTracker pressure myography software. This process was repeated for both native and FRESH printed samples. Software settings were adjusted to ensure the starting outer and inner diameters were aligned with the measured values of each carotid or printed vessel at atmospheric pressure. The maximum recorded pressure before rupture was taken as the burst pressure.

RAMP system

The RAMP system was engaged for pulsatile pressure experiments. After vessels were mounted and sutured, and HemoLens’ heating was prepared pressure sensors (Honeywell #26PCDFG5G) were placed near the vessel inlet and outlet, connected to the 30-gauge needles via additional 1/16” tubing. HemoLens’ perfusion pump (TOPINCN #JT-180A-12) was primed with 5 mL PBS and activated via HemoLens’ LCD. Vessel outlet tubing was threaded through a designed circular clamp. The location of the clamp is dependent upon the relationship between flow and pressure, and specific perfusion pump utilized. We found that tubing with an inner diameter less than 1 mm resulted in high baseline pressure and required the clamp to be placed before the P1 inlet pressure sensor to slow inlet flow and drop pressure across our vessel. Alternatively, when using larger diameter tubing, we observed a low baseline pressure under flow requiring the clamp to be placed after the P2 outlet sensor to slow outlet flow and increase pressure. In either case, the clamp was manually cinched by tightening two M3 screws to achieve a baseline of 80 mmHg.

Once the baseline pressure was established, the RAMP system was placed downstream of the P2 outlet pressure sensor, and the vessel tubing was held taught under RAMP’s occlusion arm by insertion into the two tubing holders designed into the RAMP base. To avoid tubing sliding during RAMP operation, a Luer-lock connector can be attached to each end of a small segment of perfusion tubing. The RAMP system was activated and beats per minute were defined (0 – 220 bpm range, which can be adjusted if needed in the Arduino code) via HemoLens’ LCD.

Dynamic compliance measurements.

With the perfusion pumps on, baseline pressure is adjusted to roughly 80 mmHg using our designed small clamp to slightly occlude the outlet line until the desired baseline pressure is achieved. Baseline pressure is monitored on HemoLen’s LCD screen during adjustment. Once set, the HemoLens system is restarted, and the pump and RAMP system are activated during the normal HemoLens startup. Due to Vasotracker’s limitation in video frame rate, vessel distension was recorded for 1 minute with the Windows 11 camera app (Microsoft), and diameters were calculated using Vasotracker’s offline analyzer²³. FIJI (NIH) or Premiere Pro (Adobe) were used to convert recorded video into a .avi format for Vasotracker’s offline analyzer. Real time pressure during the experiment was recorded using the custom python script “Pressure_Graphing_HemoLens.py” provided in our Zenodo repository. The outer diameter was used in all dynamic compliance calculations for both native and FRESH printed scaffolds as no significant differences were observed between inner and outer diameter measurements. Software settings in the offline analyzer were adjusted to ensure the starting diameters at baseline were consistent with results taken from burst pressure measurements. A custom Python script was used to filter and compile the recorded diameter changes and provide a list of the diameter changes of each peak. Diameter changes were averaged over the number of sampled peaks. Another Python script used fast Fourier transform (FFT) to calculate pulse frequency of vessel distention. Dynamic compliance ($Dc$) (mmHg⁻¹) was calculated as previously described⁴⁸.

$$Dc=\frac{{(D}_{max}-D_{min})}{D_{min}}\frac{1}{{(P}_{max}-P_{\mathrm{m}\mathrm{i}\mathrm{n})}}$$

where $D_{max}$ = the recorded diameter of each peak (μm), $D_{min}$ = the baseline recorded diameter (μm), $P_{max}$ = maximum pressure peak (mmHg), and $P_{\mathrm{m}\mathrm{i}\mathrm{n}}$ = minimum or baseline pressure (mmHg).

Characterization of Mechanical Behavior

To demonstrate the usability of data gathered with the HemoLens platform, we used a biomechanical framework to characterize the mechanical behavior of sample specimens from a variety of testing conditions: normotensive carotid, SCA carotid, printed vessel without crosslinking, and printed vessel with crosslinking. Our framework closely follows prior work for biomechanical phenotyping of murine blood vessels⁶³. Loading protocols in this work were reduced, as our focus was on development and validation of the HemoLens. Therefore, pressure and loaded outer diameter from a single testing cycle of each specimen were used for fitting with a constant axial stretch. A four-fiber family-based (4FF) stored energy density function, extensively used to simulate mechanical behavior of vascular tissue, was used to capture the measured experimental work. We fit the material parameters of the 4FF model by matching the model-generated pressure vs diameter behavior to that gathered experimentally with numerical optimization. Linearized stiffness, which is a measure of the material stiffness in the loaded configuration, was calculated according to prior methods at different pressure states as the main output for comparison^63,64. This material stiffness was calculated at different pressures, which is necessary for non-linear materials to observe differences in behavior across loading states.

FRESH bioprinting of collagen vessels

Test vessels were designed in an array fashion nested within a support frame using Fusion360 (AutoCAD) with 450 μm inner diameter and 1200 μm outer diameter. Machine pathing was visualized and Gcode Cura v5.7 (Ultimaker). For printing, Gcode was uploaded to our custom-built bioprinter. Collagen-I (LifeInk 260, Advanced Biomatrix #5358) was used as our bioink without modification. Once the print was complete, the collagen vessels were allowed to rest at room temperature for at least 30 minutes, then transferred to a benchtop incubator at 37 °C to melt the gelatin support. After 1 hour of incubation, gelatin support was washed away with repeated 50 mM HEPES (Sigma #1003551688) exchanges every 30 minutes thereafter until no gelatin was apparent in the HEPES bath. Washed printed scaffolds were stored in fresh 50 mM HEPES at 4 °C until needed.

Paraformaldehyde fixing of FRESH printed collagen scaffolds

A fixing solution of 4% Paraformaldehyde was created by diluting a 16% Paraformaldehyde (ThermoScientific lot# T15K021) stock concentration with MiliQ water. The fixing solution was stored at room temperature until needed. Printed scaffolds were immersed in 4% PFA for 30 minutes at room temperature. Printed scaffolds were washed 3x in 50 mM HEPES for 3 minutes each and stored in fresh 50 mM HEPES at 4 °C until needed.

Data Analysis:

Custom Python scripts were used to calculate average diameter changes and produce the frequency spectra via fast Fourier transforms (FFT). Please see data availability section for details to accessing custom Python scripts. Statistical and graphical analyses were conducted using Prism v10 (GraphPad) and Excel v16 (Microsoft) software. Statistical tests were chosen based on sample size, normality of the data set, and data requirements. For pressure increases in syringe diameter, a student’s T-test was used to compare pressure increases between 3 mL and 1 mL syringe types. Significance was determined as p<0.05. For average diameter and dynamic compliance, our sample size of pulse peaks (technical replicates) was above n=30, sufficient for the central limit theorem (CLT) to assume our data set reaches a standard Gaussian distribution. A one-way analysis of variance (ANOVA) with Tukey’s multiple comparisons was used to determine the significance of average diameter and dynamic compliance. Significance was defined as p<0.05 for all comparisons.

Supplementary Material

Video_S10

Video 10: Vessel Mounting Instructional Video

Video_S9

Video 9: Crosslinked 70mg/mL Collagen Acellular Scaffold under RAMP 80/160+ mmHg

Video_S1

Video 1: 3D printed Micromanipulator Control

Video_S8

Video 8: Crosslinked 70mg/mL Collagen Acellular Scaffold under RAMP 80/120 mmHg

Video_S7

Video 7: Non-Crosslinked 70mg/mL Collagen Acellular Scaffold under RAMP 80/120 mmHg

Video_S6

Video 6: SCA Mouse Carotid Artery under RAMP 80/120 mmHg

Video_S5

Video 5: Wildtype Mouse Carotid Artery under RAMP 80/120 mmHg

Video_S4

Video 4: RAMP Occlusion Mechanism

Video_S3

Video 3: Modular Camera Control

Video_S2

Video 2: iPRS control