Microfluidics-based electronics have shown great potential in numerous applications, such as energy harvesters, bio-inspired devices, and wearable healthcare sensors. Fabrication challenges and material incompatibilities pose barriers to achieving the full potential of this emerging field. Elastomeric flexible microfluidic channels are conventionally fabricated by casting on a micro-machined mold, followed by a bonding step. Difficulty fabricating multilayers, and weak bonding to many substrates limit device capabilities and scalability. Integration of transducers within microfluidic channels is also challenging, especially for materials requiring high temperature annealing.
Novel processing techniques for the realization of scalable and flexible microfluidic electronics have been developed to suit a wide range of substrates. Biocompatible Polyethylene-glycol (PEG) is Inkjet-printed as a sacrificial layer. Elastomers, the channels’ structural material, are cast on PEG and cured. PEG can then be removed through heating above its phase-change temperature. This process allows easy modulation of the shape and dimensions of the channels on any substrate, and eliminates the need of manual realization of multilayers interconnection. Integration of ferroelectric and magnetic sensors is demonstrated with a photonic annealing process that enables film heating to over 900°C with insignificant substrate heating. Combined, these techniques enable realization of a new generation of wearable, flexible microfluidic electronics.
An essential component in the study of inner ear drug delivery is cochlear fluid
pharmacokinetics. A popular method for simulating the drug diffusion in the inner ear
(Plontke et al., Audiology and Neurotology, 12:37-48, 2007) uses a finite element
method model that relies on assumed values of the diffusion coefficients within the
scala tympani (ST), scala vestibuli (SV), scala media (SM), and spiral ligament (SL),
as well as parameters describing the permeability between them and clearance to
other regions. In this work, we investigate inverting this model to estimate diffusion
parameters given empirical concentration values derived from micro-computed
tomography (μCT) imagery of the mouse cochlea. Solving the inverse problem over
the entire set of 3-D regions can be computationally intensive, so we focus instead on
inverting the reduced 1-D problem proposed by Plontke et al. Given initial estimates
of the diffusion, permeability, and clearance parameters, we use an iterative nonlinear
least-squares technique to minimize the mean-squared error between the predicted
concentration values and those derived by averaging over regional cross-sections of
the μCT scans. Extracting the appropriate cross-sectional data from each of the four
regions in the μCT scans is a difficult task that requires accurately
identifying/segmenting the regions. So, we employed a 3-D subject-atlas image
registration technique (Z. Xu, MS Thesis, Rochester Institute of Technology, 2016).
This aligns the high-resolution atlas with pre-identified regions to the μCT scans,
enabling labels for the regions to be propagated. However, voxels are inherently
discrete, hence the registration will not be perfectly accurate, so it is likely that the
labeled regions may include data from bones near the region boundaries. For this
reason, we weight voxels based on their proximity to the center of their labeled
region, so voxels that may contain bone have less influence on the derived
concentration values. Using a set of μCT scans comprising a baseline scan and
subsequent scans of a mouse during micro-infusion delivery of a contrast agent
(ioversol) to its cochlea, we illustrate how our techniques can be used to robustly
estimate diffusion, permeability, and clearance parameters for Plontke et al.’s reduced
1-D problem. These results will accelerate pre-clinical and clinical development of
infusion paradigms, to facilitate optimal drug delivery to the cochlea, for preventing
or treating acute and chronic types of hearing loss.
Supported by the National Institute On Deafness and Other Communication Disorders
of the National Institutes of Health, under Award # R01 DC014568.
Abstract : Multifunctional polymers have attracted attention as the need grows for materials to simultaneously demonstrate different capabilities . Of those highly-demanded nanocomposites, magnetothermally-responsive materials demonstrate a promising application in drug delivery systems . The quick reacting nature of magnetothermally-responsive polymers inspires us to present a one-of-a-kind nanocomposite valve that utilizes both the magnetic force attraction and the thermal-induced phase change for operation. Valves that show good performances are typically micromachined, however, they cannot be easily integrated with other components and fail to provide a simple solution for interconnecting channel systems [3,4]. Researchers have shown great interest in flexible magnetic valves, but have yet overcome the integration difficulty and the high power consumption . The proposed valve based on magnetothermally-responsive nancomposites offers unique features and capabilities including easy integration, biocompatibility, scalability and a one-way passive operation. The valve is fabricated by incorporating Fe3O4 nanoparticles in polyethylene-glycol (PEG). Fe3O4 is biocompatible and can be easily dispersed and manipulated with external field due to its superparamagnetic nature. 2 µm Parylene encapsulates the nanocomposite to offer biocompatibility, enclosure and flexibility. The presented valve is designed to fit 0.7 mm in diameter channels, however, it can be customized to fit any microfluidic system. After the valve is injected in the channel and steered to the desired location, it is heated above the phase change temperature of PEG (e.g. body temperature). The valve expands and gets locked in place with a NdFeB permanent magnetic nanocomposite to attract the nanoparticles, sealing at one edge of the valve while keeping the other loose. The nanocomposite actuator, operated in liquid state, is stretched by the nanoparticles attracted toward the magnetic field source and sticks to the inner wall of the tube to realize one-way sealing. The presented magnetothermally-responsive actuator demonstrates a reliable operation as a passive one-way valve that exhibits steady flows from the inlet side, i.e. 5 μL/min at 14 kPa and robust sealing from the outlet side, i.e. 66 nL/min at 14 kPa min due to the difficulty for the flows to deform the actuator’s outlet side where the nanoparticles are attracted toward the ring magnet to fully stretch the actuator at. The demonstrated performance indicates the valve is suitable to be integrated with implantable micropumps or drug delivery systems to prevent back flow of body fluids to the microfluidic systems.
Won the Best Poster Award
Abstract : Microfluidics-based electronics have shown great potential in numerous applications such as energy harvesters, bio-inspired devices, and wearable healthcare sensors [1,2]. Microfluidic electronics have been realized using conventional microfluidic fabrication techniques. However, fabrication challenges and material incompatibilities pose barriers to achieving the full potential of this emerging field and limit the device capabilities and scalability .
This paper demonstrates a novel and simple processing technique for the realization of scalable and flexible microfluidic microsystems by inkjet-printing polyethylene-glycol (PEG) as a sacrificial template. Planar and out-of-plane structures can be directly printed on a wide range of substrate materials, including nonplanar surfaces, and embedded in a variety of soft (e.g. elastomeric) or rigid (e.g. thermoset) polymers. PEG can be removed through heating above its phase-change temperature after the formation of the structural layer. The developed technique allows easy modulation of the shape and dimensions of the pattern with the ability to generate complex 3D architectures without using lithography, and at low cost and short manufacturing time. The method produces robust structures that can be realized on electrodes and transducers directly without requiring any bonding or assembling steps which often limit the materials selection in conventional microfluidic fabrication [4.5].
The inkjet printing of PEG was optimized to print features with 50 μm width resolution, and 6 μm thickness per printed layer, with the ability to print multilayers and align features. Multilayer Polydimethylsiloxane (PDMS) microfluidic channels were created using this technique on several substrates such as glass, Polyethylene terephthalate (PET), PDMS and non-planar surfaces (including the tip of a cardiac catheter) to demonstrate the capability of the concept to directly realize microfluidic electronics and lab-on-a-chip devices on any surface. Moreover, the biocompatibility of PEG makes it attractive for medical and biological applications. By utilizing liquid metals (i.e. E-GaIn) as the filling material of the channels, flexible and stretchable devices have been realized such as pressure sensors with a sensitivity of 0.24 kPa-1, and strain gauges that exhibit changes up to 10 %ΔR/R0.
Biomedical engineering advances in protective and restorative inner ear biotherapies have created new opportunities to address vestibular disorders, deafness, and noise induced-, sensorineural, and age-related hearing loss. To avoid unwanted, systemic side-effects, controllable, programmable drug delivery systems are essential for therapeutic development in animal models, and for future human translation with implantable, subcutaneous or behind-the-ear systems. The mouse animal-model system offers significant advantages for studies involving specific transgenic, knock-in, and knockout variants that model different human diseases and communication disorders. However, the small size of the mouse necessitates key advances in pump miniaturization for implantation and scaling of flow rates and volumes to match cochlear anatomy. Here we present a biocompatible, fully implantable, scalable, and wirelessly controlled peristaltic micropump. The microfluidic core is based on commercially available catheter microtubing (250μm OD, 125μm ID) providing a leak-free, biocompatible flow path while avoiding complex microfluidic interconnects. Direct write micro-scale printing provides structure around the microtubing for mechanical support and actuation. Micropump features are fabricated directly on the backside of a printed circuit board assembly providing control electronics for sequenced actuation and wireless control. Peristaltic flow is achieved by compressing the microtubing via expansion and contraction of a thermal phase-change material integrated adjacent to the microtubing and within the printed micropump structure. A refillable reservoir was integrated with the micropump while maintaining a form factor optimized for subcutaneous implantation in the mouse. Pump characterization results and initial in vivo implantation tests are presented. Initially targeting the small murine cochlea, this low-cost design and fabrication methodology is inherently scalable to larger animals and for clinical applications in children and adults by appropriate scaling of the microtubing size and actuator volume.
Supported by the National Institute On Deafness and Other Communication Disorders of the National Institutes of Health, under Award # R01 DC014568.
Abstract : Lead Zirconate Titanate (PZT) has been widely used in actuators , sensors , and energy harvesters , due to its high piezoelectric response. PZT aerosol jet printing and photonically pulsed sintering technique has been demonstrated, which significantly shortens the processing time, enhances the film piezoelectric performance, and allows realizing devices on flexible and non-planar substrates . However, it is presented that the depth of photonic sintering did not extend through the full thickness of the film. Therefore, it limits the film to obtain a further higher piezoelectric response. In this paper, a fully photonically sintered PZT film is reported giving rise to a superior piezoelectric performance. The commercially available nano-scaled PZT powders (average diameter = 480 nm) were used to mix with DI water, sintering aid (Cu2O and PbO with molecular weight percentage of 1:4), and binder (polyvinylpyrrolidone) to formulate the aerosol-jet printable ink. Aerosol jet printing technique (Aerosol Jet 300, Optomec, Inc., USA) was adopted to deposit the film on a stainless steel substrate resulting in an approximately 5 μm thickness film. After drying at 200 °C for 1 hour in the atmospheric environment, the film was sintered photonically with the optimized sintering parameter combination (voltage = 550 V, duration = 170 µs, frequency = 2 Hz, number of pulses = 20 × 2) in less than 1 minute. With assistance of Scanning Electron Microscopic images, a fully sintered PZT film was observed. After poling the samples with 20 kV/cm electric filed for 1 hours at 170 °C in the atmospheric environment, the piezoelectric voltage constant (g33) was measured yielding -23.8×10-3 V-m/N. This greatly improves the previousely reported result (17.9×10-3 V-m/N) . This optimized fabrication process enables to use short time to obtain an enhanced piezoelectric response providing more potentials for better piezoelectric performance devices.
Advanced micropumps that enable chronic delivery of therapeutic agents with controlled delivery profiles require integrated electronics and wireless communication capabilities. Traditional assembly approaches require high temperature soldering processes that preclude direct integration with micropumps designed with temperature sensitive polymer components. Here we present low temperature assembly of commercial, off-the-shelf surface mount electronic components electrically connected to the substrate interconnects through a conductive adhesive. The anisotropic conductive adhesive process uses a magnetic field to align conductive particles from 0.5 to 3 μm within the adhesive in the z-direction during curing, providing electrical bridges between the electronic components and the substrate, and insulating between pads in the x-y directions. The isotropic silver adhesive has uniform high conductivity, but must be placed precisely to prevent horizontal bridging. Micropump electronics are demonstrated including embedded microprocessor, power regulation, feedback control of thermal actuators, and Bluetooth low-energy wireless communications. The overall dimensions of the complete assembly are 8.2 mm x 8.8 mm x 1.8 mm. The controller is assembled using the smallest commercially available parts, wafer-level chip-scale packages at 0.4 mm contact pitch and 0.15 mm contact spacing. Assembly is carried out by applying the adhesive to the substrate, placing the components, and then curing at a maximum temperature of 70 °C. Integrated micropump electronics assembled using isotropic and anisotropic conductive adhesive processes are demonstrated for feasibility and performance. The approach enables use of power efficient, commercial off-the-shelf electronics and integration over low-temperature substrates. Future work will leverage this approach to integrate these electronics directly on a micropump with interconnects defined using direct-write technologies.
Supported by the National Institute On Deafness and Other Communication Disorders of the National Institutes of Health under Award Number R01DC014568.
Controllable drug delivery systems will become increasingly important as protective and restorative biotherapies for noise induced-, sensorineural-, and age-related hearing loss; deafness, and vestibular disorders are developed. Miniaturized pumps create opportunities for therapy development in animal models, and will be essential for fully implanted or behind the ear systems for human translation. In particular, since mice have become such a successful biomedical, genetic and bioengineering lab animal of choice, development of micropumps that can be implanted subcutaneously in mice is quite important. This will allow in vivo testing of novel drug treatments in transgenic, knock-in, knockout and various strains of mice that model different human diseases. Here we present progress on a scalable micropump platform based on direct, phase change actuation of either microtubing or micro-fabricated diaphragms for peristalsis. The first approach integrates commercially available 250 μm outer diameter/125 µm inner diameter polyurethane catheter tubing that serves as the micropump flow channel, reservoir interconnect, and cannula connection to the drug delivery site. Direct-write, 3D print technologies are used for definition of micropump features, heaters, and accurate delivery of paraffin wax for phase change actuation. In the latter case, the fluid pumping chambers are formed by a spin-coated PDMS (polydimethyl siloxane) layer peeled off from a 3D printed mold. Then the chambers are bonded with directly-written heaters/paraffin wax actuators embedded in the PDMS diaphragm to integrate a micropump. Results with a proof-of principal device are presented showing an ability to deliver volumes at less than 10 nL per actuation event. This can be scaled readily to higher flow rates by increasing the tubing/paraffin reservoir size and frequency of peristalsis. To test for biocompatibility, procedures consistent with the ISO 10993 standard for implantable medical devices were used. An epithelial cell line was cultured under standard conditions. The micropump components were sterilized with ethylene oxide and then inserted into the center of the cell culture wells. Biocompatibility data analyses are ongoing, but assays of cell cytotoxicity so far reveal no toxicity. Future work will integrate the micropump with a drug reservoir and control electronics for wireless communication and in vivo testing in the mouse model system.
Supported by the National Institute On Deafness and Other Communication Disorders of the National Institutes of Health under Award Number R01DC014568.
Dynamic contrast enhanced (DCE) Micro-computed tomography (µCT) imagery of the rodent inner ear has been successfully used to noninvasively characterize drug delivery inside the cochlea (Haghpanahi et al., Annals of Biomedical Engineering, 41(10):2130- 2142, 2013). Using a series of manually segmented and labeled histological sections of a mouse inner ear as an atlas, Haghpanahi et al. registered selected 2-D atlas images to slices of the µCT image, allowing the atlas labels to be propagated to identify and segment the intracochlear structures in the subject mouse. With this technique, drug concentration profiles can be quantified in each segmented structure. In this work, we present two improvements in the atlas-based registration step; i) full automation, which improves analysis by removing the need for an expert user to select specific slices from the µCT image for registration, and ii) full three-dimensional registration, which improves segmentation accuracy by incorporating all atlas images instead of a user-selected subset. To enable full automation, we carry out the following steps: (i) threshold the µCT image to identify bone regions (+700 to +3000 HU), (ii) identify the sagittal plane by maximizing the Jaccard index (a measure of symmetry) between the left and right bone regions, and (iii) localize the region containing the inner ear by searching for geometric features matching those of the atlas that are preserved when reflected through the sagittal plane. Once the inner ear region has been automatically localized, we perform 3-D atlas-based registration by first identifying the 3-D affine transformation that optimally aligns the atlas with the inner ear, and then refining the registration result by performing 3-D deformable registration. Using the 3-D deformation aligning atlas to subject, we propagate the manually identified region labels from the atlas to yield estimates of the locations of the scala media, scala tympani, and scala vestibule. Finally, we refine the predicted boundaries of the different scalae via a level-set energy minimization approach. Using multiple µCT subject images, we illustrate that our fully automated approach that uses the full three dimensions of data from the atlas and subject images enables more accurately segmented scalae regions than are attainable using the previous approach of estimating a collection of 2-D transformations between atlas and subject slices and interpolating the results.
Supported by the National Institute On Deafness and Other Communication Disorders of the National Institutes of Health under Award Number R01DC014568.
While pressure and flow waveforms, calculated from computationally inexpensive one dimensional analysis, reflect the real picture of the process related to the inviscid flow, local three-dimensional viscous flow analysis brings an additional insight, highlighting the area of elevated level of circulations and wall shear stress. The latter according the experimental evidence is leading to the atherosclerotic lesions and triggers unfavorable biological events in the development of the intimal hyperplasia. To validate the implementation of the code in an arterial network, analysis has been applied to the simple “fork” type bifurcation. Three-dimensional Navier – Stokes simulation has been used to optimize the graft artery junction geometry from the view point of wall shear stress level minimization. We managed to reduce recirculation zone in the bifurcated area by minimization the maximum values of a wall shear stress gradient across all vessels varying the bypass vessel orientation angle within the frame of implied geometric constraints. It was shown that the typically used for the bypass T-junction topology is not the best. The best orientations of a bypass configuration was found for a set of specified flow rates across each bifurcated vessel.
Blast related traumatic brain injury (bTBI) is characterized by neurological impairment including loss of consciousness, disorientation , memory loss, impaired cognition, and headache. Of significant concern is mild bTBI where symptoms are wide ranging, non-specific, and difficult to distinguish from psychological co-morbidities. The Blast Gauge System was created to provide an objective measure of blast overpressure exposure to assist in identifying soldiers at risk of bTBI, and to enable correlation of singular and repetitive exposure to acute and chronic injury for future refinement of thresholds. While capturing full blast overpressure and acceleration waveform data for post-event analysis, this system uses two thresholds to aid in triage. These thresholds were developed based on recent human and large animal studies correlating blast overpressure exposures to neurological effects and changes in physiological, histochemical, immunohistochemical, and protein biomarkers. These thresholds are consistent with those identified for blast-related upper respiratory and gastrointenstinal tract trauma.
Currently the Neurocognitive Assessment of Blast Exposure Sequela in Training (NC-BEST) and Investigating the Neurologic Effects of Training Associated Blast (I-TAB) studies have incorporated the Blast Gauge into their protocols to allow for a correlation between objective measures of blast exposure during combat training and physiological and behavioral measurements. These studies will allow for an increased understanding of blast exposure dose response curves in relation to multiple blast variables, repetitive blast exposures, and individual genetic and physiological characteristics associated with development of chronic neurocognitive sequela. The results will guide refinement of our current blast exposure thresholds, and may help determine whether certain structural changes in weapon systems, training areas, or CONOPS can mitigate or minimize these overpressure exposures. This is critical, as training associated blasts currently account for 67% of all recorded blast exposures in military personnel. Anecdotally, concussive-symptoms are commonly reported by instructors and trainees participating in Breacher and heavy weapons training programs. Tate et al. recently reported that 11 of the 21 participants in a 2-week breacher course had positive biomarkers indicative of bTBI with associated neurobehavioral symptoms and decreased cognitive functioning.
A series of laboratory and field tests were conducted using the BG with industry standard pencil probes (PP) as controls. Three different models of the BG were used: H1 (High Sensitivity), S1 (Standard Sensitivity, DARPA), and S2 (High-G). Laboratory tests were conducted with two shock tubes. The WRAIR/NMRC shock tube used 4 different membrane thicknesses (OP range 4 – 15 psi, n=9–12 trials per level while the DSTO shock tube used 2 thicker membrane levels (OP range 30 – 70 psi, n = 3 trials per level). Three distinct field tests were conducted with high explosive charges or during weapons systems training: 1. Two iterations of an explosive door breach were conducted (OP range: 2–11 psi) using the same charge weight (CW); 2. Within a blast chamber, high explosives were detonated with BG and PP sensor locations varied during 5 separate detonations at the same CW (OP range: 4–72 psi); 3. Field artillery (M119, 105 mm Howitzer) firing with 3 different propellant charge levels (OP range: 0.9–3.5 psi). Within the optimal sampling range (H1:0.25–30 psi, S1:2.5–100 psi, S2:2.5–100 psi), BG readings were 1–19% lower than PP readings and hypothesis tests showed no statistically significant difference (SSD) between BG and PP in 9 of 11 conditions (p > 0.05). Above the optimal sampling range, BG was 1 –30% lower than PP and hypothesis tests showed no SSD in 3 of 5 conditions (p > 0.05). These differences are likely due to the slower sampling rate of the BG and piezoresistive properties of the sensing element. This data supports the repeatability and reliability of the BG for the measurement of both reflected and incident overpressures in the field environment. Additional examples of the utility, interpretation challenges and limitations associated with OP measurement (eg. detonation in confined spaces) will be discussed.
Delivery of potentially therapeutic compounds to the inner ear via the round window membrane (RWM) has advantages over alternative methods. Specifically, the RWM approach minimizes bleeding, with little impact upon functional hearing measures. However, some previous reports show that significant basal-to-apical concentration gradients occur (Plontke et al. Laryngoscope 2007;117:1191-1198). The new surgical approach presented here utilizes polyimide infusion tubing placed onto the niche near the RWM, through a bulla-ostomy. Inclusion of a posterior semicircular canalostomy provides additional opportunities to reduce drug concentration gradients. Methods: Young adult CBA/CaJ mice were divided into two groups: bulla-ostomy approach only (B) and bulla-ostomy + canalostomy (B+C). Cochlear function was evaluated by distortion product otoacoustic emissions (DPOAEs) thresholds and auditory brainstem responses (ABRs) (8-51 kHz). Mice were infused with 50mM salicylate in artificial perilymph at a flow rate of 50nl/min for a total volume of 1000nl (20 min); DPOAEs were measured before and after surgery, and at intervals during the infusion. The mice recovered for 1 week, and re-tested. Results: there was no significant impact on auditory function utilizing the RWM surgical procedure, and DPOAE thresholds were elevated reversibly during the salicylate infusion. Comparing the threshold shifts for both methods, the B+C approach had more of a physiological effect than the B approach, especially at lower frequencies – more apical cochlear locations. Unlike mouse cochleostomies, there was no deleterious auditory functional impact after 1 week recovery from surgery. Regarding the B+C approach, it had more drug efficacy at lower frequencies indicating reduced concentration gradients, underscoring potential benefits for more precise control of delivery of inner ear therapeutic compounds.
Introduction: Arterial compliance is a marker for cardiac burden in atherosclerotic disease, with the pressure Pulse Wave Velocity (PWV) correlated to compliance. Current clinical practice employs pulsed wave Doppler to measure Flow Wave Velocity (FWV) as a surrogate of PWV. We hypothesized that PWV and FWV are not directly related and are affected by left ventricular ejection time (LVET). Furthermore, we proposed that aortic PWV is independent of mean arterial pressure (MAP) in the setting of isolated systolic hypertension.
Methods: Using a physiologically accurate electromechanical cardiovascular simulator, two solid state manometer-tipped pressure transducers and two transit time flow sensors were located at the aortic root and at the aortic bifurcation. PWV and FWV were directly measured while individually varying contractility and thus LVET. The experiments were repeated at various systemic vascular resistances (SVR) and vascular compliances. Automated signal processing and data extraction techniques were used to calculate the key parameters.
Results: As LVET increased, FWV decreased but PWV increased while MAP remained constant for a fixed SVR and compliance. (Figure 1) This trend held consistent at different SVR’s and compliances. The relationship of PWV and FWV with LVET appeared to be exponential and linear respectively. For a constant MAP, the associated PWV varied by up to 50m/s and FWV by up to 10m/s for a change in LVET of 225ms.
Conclusions: In conclusion, our data shows that PWV and FWV appear to be inversely related. Our data also suggest that PWV and FWV are independent of MAP in the setting of isolated systolic hypertension. These findings suggest that FWV measured by pulsed wave Doppler may not be a simple surrogate for true PWV. Future work is needed to elucidate the hemodynamic principals governing the relationship between PWV and FWV.
Computer modeling of biological systems has proven to be an invaluable tool in numerous applications. In developing therapies for the cochlea involving direct application of biotherapeutic compounds, multiple parameters of the infusion/delivery system must first be established; concentration of the therapeutic agent, volume and flow rate of the infusion, and which delivery system leads to the optimal concentration profile are all parameters that can be investigated with a biologically accurate fluid dynamic model of the inner ear. Existing models for intracochlear drug delivery utilize clearance and interscala transport parameters based on limited in vivo measurements using implanted ion-selective electrodes [1,2]. Measurement of transport parameters within the fluid spaces along the length of the cochlea would enhance the accuracy of these models.
These preliminary studies show the ability to image delivery of contrast agent to the cochlea and determine concentrations in each region of interest. This will lead to the eventual calculation of interscala transport and clearance parameters. The use of real-time in vivo experimentation to gather these parameters will aid in the construction of a fluid dynamic model of the inner ear with higher accuracy than is currently available.
Advanced deafness therapies that restore normal auditory function will require carefully timed and dosed, site-directed delivery of multiple therapeutic compounds over a period of time. Syringe and osmotic pumps have proven effective for baseline investigations in animal models, but lack the flexibility required for more sophisticated therapy development. An implantable micropump with controllable flow rate and dosing profile is being developed using MEMs technologies. The pump initially targets murine intracochlear drug delivery for deafness therapy research. Key integration technologies have been created to minimize both size and power for this implanted device. A method for coupling fine capillary tubing to microfluidic channels via in-plane interconnects is presented. Capillary tubing inserted into micro-channels etched in the surface of a silicon wafer is sealed in place via room temperature polymer deposition providing a low volume, biocompatible interconnect. These microfluidic connections withstand pressures as high as 827 kPa (120 psi) with average pull test strength of 2.9 N. The interconnects consume less than 20 nl of volume and exit in- plane with the pump, facilitating implantation. Resistive micro-bridges integrated into the middle of the fluid channel provide an optimal means of flow rate measurement via hot-wire anemometry. Multiphysics modeling is used to determine the sensor geometry that provides the most favorable trade-off between sensitivity and power consumption. Analog circuit solutions are explored for in-situ fluid temperature compensation without digital computation. A process for creation of deformable membranes over pump chambers with simultaneous coating of the microfluidic channels has been developed allowing integration of a biocompatible fluid flow path. These integration technologies represent critical steps towards micropump development suitable for implantation in mice.
Compound delivery to the cochlea generally results in strong basal-to-apical concentration gradients, limiting the efficacy of therapeutic approaches for apical structures, and constraining therapeutic dose ranges. These gradients depend on the location of cochlear entry, the resulting fluid flow path, and diffusion / clearance mechanisms, offering few options for control and reduction. In the present study, infusion flow rate was examined as a parameter for reducing concentration gradients within the cochlea for basal turn scala tympani cochleostomy infusions with and without a posterior semicircular canal canalostomy. Fixed volumes of artificial perilymph (AP) and sodium salicylate (10 mM) were infused. Cochlear function was measured via closed-system DPOAE threshold measurements from 8-44 kHz, with threshold shifts compared for both surgical approaches using two flow rates; 16 and 32 nl/min. Neither flow rate had an acute impact on auditory function as demonstrated by stable thresholds during the initial AP infusion and subsequent threshold recovery during the washout phase. The cochleostomy-only approach exhibited statistically significant enhancements in DPOAE threshold shifts with the higher flow rate at 12 and 16 kHz, suggesting higher concentrations of salicylate at these more apical locations. Flow rate did not impact DPOAE thresholds for the cochleostomy-plus-canalostomy approach. Simulations were used to investigate the impact of higher flow rates on salicylate concentration along the length of the cochlear spiral, with clearance rates examined as a potential influencing factor.
Infusion flow rate may be an important variable in optimization of intracochlear drug delivery paradigms. Elucidating the details of threshold shifts will provide information critical for development of more advanced models of cochlear infusate flow patterns and optimization of inner ear infusion protocols.
Direct delivery of compounds to the inner ear in animals is most commonly achieved by absorption or direct injection through the round window membrane, or infusion through a basal turn cochleostomy. These methods provide direct access to cochlear structures, but with a strong basal-to-apical concentration gradient consistent with a diffusion-driven distribution. This gradient limits the efficacy of therapeutic approaches for apical structures, and puts constraints on practical therapeutic dose ranges. A surgical approach involving both a basal turn cochleostomy and a posterior semicircular canal canalostomy provides opportunities for facilitated perfusion of cochlear structures to reduce concentration gradients. A cannula inserted into scala tympani at the cochleostomy site allows controlled delivery of solutions into perilymph, with the canalostomy in the PSCC serving as a fluidic exit port. The theoretical flow pattern is from base to apex in scala tympani, through the helicotrema, from apex to base in scala vestibuli, and then through the semicircular canals. This contrasts with infusion at the basal turn of scala tympani with no exit hole (cochleostomy only) where presumably excess fluid is pushed through the cochlear aqueduct.
Infusion (1μl/hr) of fixed volumes of artificial perilymph (AP) and sodium salicylate were used to evaluate the two surgical approaches in the mouse. Cochlear function was evaluated via closed-system DPOAE threshold level measurements. AP infusion confirmed no surgical impact to auditory function, while shifts in DPOAE thresholds were measured during infusion of sodium salicylate (10 mM) and AP (washout). Different fluids were separated by 10 nl air bubbles within the infusion tubing to avoid within-tube diffusion. Frequency dependent shifts were compared for the cochleostomy-only and the cochleostomy-plus-canalostomy approaches.
Intracochlear drug delivery in the mouse is challenging due to the extremely small size and volume of the inner ear. Physiological measures can access the impact of delivered compounds on peripheral auditory function, but infusate distribution characteristics must be inferred. Computer simulations offer opportunities to calculate cochlear concentration gradients for different delivery approaches and are an excellent complement to in vivo studies. Fluid modeling software is available to simulate a variety of delivery protocols representing infusions, perfusions, and extracochlear round window drug applications. Diffusion, volume flow, interscala transport, and clearance mechanisms are modeled allowing determination of drug concentration with position in the cochlea. When coupled with a cochlear tonotopic map, simulated concentration profiles can be correlated with shifts in measured auditory responses as a function of sound frequency. Two different drug delivery approaches are considered in the present investigation involving mice: scala tympani infusion through a basal turn cochleostomy, and this approach coupled with a fluidic exit hole in the posterior semicircular canal. Distortion product otoacoustic emission thresholds were measured from 8-44 kHz before and during delivery of artificial perilymph and sodium salicylate. Shifts in threshold were compared to simulated concentration gradients to highlight the relationship between drug concentration and frequency-dependent changes in auditory function. Model parameters were adjusted to match the profile of these characteristics and to provide insight into mechanisms impacting drug distribution in the cochlea. Strengths and limitations of currently available cochlear infusion models are discussed and critiqued for improvements.
Inner ear disease therapy research involving cochlear infusions in mice is challenging due to anatomical size, and the lack of commercially available cannulae with integrated insertion stops. Historically researchers have created custom cannula by insertion of micro-tubing into larger tubing to create a stepped profile, or by manually applying a small volume of silicone Silastic® a fixed distance from the tip of the micro-tubing. These approaches rely on the effective change in tubing diameter to provide an insertion stop to limit cannula intrusion into the intracochlear space. The profile also facilitates bonding the cannula to the bony tympanic bulla for long term infusion studies. While successful in some applications, the methods of construction make consistency and reliability difficult to achieve, particularly for mice. In the present investigation, a method has been developed to facilitate controlled interfaces for intracochlear infusions. Using microfabrication technologies, precision molds have been created that allow control of the size and shape of a Silastic® insertion stop molded around capillary tubing. The insertion depth is controlled by mold design. The approach allows consistent manufacture of micro-cannulae of varying tubing size specifically designed for intracochlear infusions in mice or other small rodents. Details of the mold fabrication process, molding techniques, and data on dimensional consistency of produced micro-cannulae are presented. Preliminary results of intracochlear infusions in the mouse using micro-cannulae designed specially for the surgical approach and anatomical limitations are analyzed and discussed. This approach has the potential to enhance infusion consistency, and mitigate complications associated with variable insertion depths and leakage at the cochleostomy site.