Ellen T. Roche

Mechanical circulatory support (MCS) alters cardiac flow dynamics and can contribute to thrombosis. Although commonly used for heart failure with reduced ejection fraction, its application in patients with preserved ejection fraction (HFpEF), nearly half of all heart failure cases, remains limited. Left atrial (LA) inflow cannulation may be advantageous for HFpEF patients due to their reduced ventricular volume, but the resulting LA flow disturbances are not well understood. This study evaluates how MCS affects LA blood flow dynamics in a HFpEF patient to clarify mechanisms influencing thrombosis risk. The effects of cannula length and drainage flow rate on thrombosis-related flow characteristics were quantified. To assess the importance of atrial mechanics, simulations incorporating LA contraction were compared with rigid-wall models, addressing the common simplification of rigid geometries. Predictions were validated against two-dimensional magnetic resonance imaging measurements obtained using a soft-robotic LA model. MCS increased washout by 50% under rigid-wall conditions and 27% with wall contraction compared with unassisted flow. However, thrombus risk increased with cannula advancement, driven by up to 151% higher shear stress. LA wall contraction reduced stasis by 68% and improved washout but also elevated predicted wall shear stress. Cannula advancement and its proximity to pulmonary veins and the mitral valve altered local flow structures and shear stress distributions, demonstrating the benefit of patient-specific positioning. Overall, MCS can enhance LA flow, but lower insertion depth and reduced pump flow may mitigate cannula-related thrombus risk. Even minimal atrial contraction notably affected washout, residence time, and shear stress patterns, highlighting its importance in thrombosis assessment.

Intraperitoneal (i.p.) therapy improves outcomes in abdominal cancers but remains underutilized due to complications from repurposed catheters. In this work, we present a replenishable therapeutic implant for repeated, localized delivery of therapies to the peritoneal cavity. In an ovarian cancer mouse model, expanded natural killer (eNK) cells were delivered once weekly and interleukin-15 (IL-15) thrice weekly. This regime reduced tumor burden compared to standard i.p. injection. The implant supports co-administration of chemotherapy, cytokines, monoclonal antibodies, or other protein-based therapies. Negative pressure applied via the port enabled longitudinal sampling of peritoneal fluid without additional surgical intervention. By reducing procedural burden and improving adaptability, the implant can help increase patient retention and therapeutic efficacy in ovarian cancer and other intra-abdominal cancers.

Intraperitoneal (i.p.) therapy improves outcomes in abdominal cancers but remains underutilized due to complications from repurposed catheters. In this work, we present a replenishable therapeutic implant for repeated, localized delivery of therapies to the peritoneal cavity. In an ovarian cancer mouse model, expanded natural killer (eNK) cells were delivered once weekly and interleukin-15 (IL-15) thrice weekly. This regime reduced tumor burden compared to standard i.p. injection. The implant supports co-administration of chemotherapy, cytokines, monoclonal antibodies, or other protein-based therapies. Negative pressure applied via the port enabled longitudinal sampling of peritoneal fluid without additional surgical intervention. By reducing procedural burden and improving adaptability, the implant can help increase patient retention and therapeutic efficacy in ovarian cancer and other intra-abdominal cancers.

Epicardial delivery of therapies has the potential to prevent adverse remodeling and promote in situ regeneration after myocardial infarction (MI) but further optimization of bioagent dosing and transport to heart muscle is required to maximize their therapeutic potential. Replenishable reservoir systems have enabled localized bioagent delivery to the epicardial surface but therapy transport from these systems is constrained by semipermeable membranes and fibrous capsule formation. Our approach to improved therapy delivery from epicardial reservoir systems is multi-pronged. First, we introduce a membrane-free reservoir system by incorporating a gelatin scaffold into a flexible polymer implant to promote direct integration with the epicardial surface and act as a replenishable depot to encourage myocardial-directed transport. Next, we perform in vitro and ex vivo validations and multi-scale computational simulations to characterize biomaterial, tissue, and organ-level transport of therapy, considering both native tissue architecture, and the effect of blood vessel clearance. As an in vivo use case of our system, we investigated the functional effect of multi-dose regimens of human follistatin-like 1 protein (FSTL1) in a rat model of myocardial infarction (MI). Groups receiving multiple doses of FSTL1 show increased cardiac performance (ejection fraction and fractional shortening), and decreased chamber stiffness 28 days after MI. Multi-dosing increases ventricular wall thickness and reduces infarct size. We demonstrate a dose-dependent increase in blood vessel number and density in the infarct zone. Finally, we establish a computational and experimental framework for patient-specific modeling to optimize implant parameters such as reservoir size and shape, infarct location, and dosing regimens, with a vision for clinical-imaging guided bioagent delivery strategies that can be modified on a per-patient, therapy-specific basis to optimize dosing regimens of various bioagents. This study highlights the potential for integrating personalized computational models with replenishable delivery systems to improve bioagent transport from biomaterials and enhance post-MI therapeutic outcomes.

Raw data from experimental crimping compression testing of 3D printed nitinol braided stents are provided as supplementary materials for the mentioned study, submitted for evaluation to the journal of Sicence Advances.

Raw data from experimental crimping compression testing of 3D printed nitinol braided stents are provided as supplementary materials for the mentioned study, submitted for evaluation to the journal of Sicence Advances.

Abstract Mitigating adverse tissue remodeling after a heart attack or myocardial infarction (MI) is critical to prevent the development of heart failure. Among various post-MI treatment strategies, mechanical reinforcement of the infarcted region with epicardial patches has promise due to its consistent improvement of chronic cardiac function and its drug- or biologic-free nature. However, despite the variety of patch materials studied to date, the lack of a programmable platform that predictably modifies early-stage cardiac biomechanics to different degrees has prevented further optimization of this strategy. Here, we introduce the matrix-mimicking bioadhesive epicardium (MMBE), a platform that can be rationally designed to achieve a wide range of anisotropic mechanical properties to offer quantifiable mechanical reinforcement of the heart upon application. The platform synergistically combines fully programmable direct-ink-writing of extracellular matrix-inspired crimped fibers and a bioadhesive for sutureless integration to the epicardium. The MMBE platform achieves an array of matrix-mimicking mechanical properties and acute modulation of cardiac biomechanics using numerical analysis, in silico studies and experimental characterizations. Furthermore, the feasibility of the MMBE platform in an in vivo rat model of MI is demonstrated. The MMBE platform can be used to systematically identify patch design parameters that alter post-MI remodeling without introducing confounding biological variables.

Introduction: Given the precise temporal coordination of natural biological processes, administering therapeutic agents at specific times can be used to enhance efficacy in a range of applications. To achieve such controlled drug delivery, various stimulus-responsive techniques (e.g., ultrasound, temperature changes, and electromagnetic radiation) have been developed. However, many of these current methods exhibit limitations, such as premature leakage prior to stimulus activation or delayed and prolonged responsiveness to stimuli. Our research introduces a soft robotic pressure-actuated drug delivery pump aimed at improving therapeutic efficacy through precisely-timed drug administration. Methods: This device utilizes silicone - a low-modulus material - for both the therapeutic reservoir and the actuation chamber to create a biocompatible and conformable interface, facilitating controlled drug release and offering the potential to be adapted as an implantable drug delivery system. Two ports in the actuation chamber allow the therapeutic reservoir to be refilled. We actuated the pressure reservoir of the device in the range of 28.5 - 59.8 mmHg and tested: the pressure-dependent release from the device; repeated release; baseline release, and the ability to deliver a wide-range of therapeutics. Results: Importantly, the system demonstrated a reliable On/Off mechanism - confirmed by actuating to ∼80% of opening pressure over 5 days - which addresses a key limitation in many existing technologies. In vitro, the device was used to deliver a range of therapeutics and had non-significant differences versus manual delivery of therapeutics in relevant assays: antibiotics (doxycycline; reduced E. coli viability by 49.6% vs. 49.8%); adeno-associated virus (AAV; transduced 73.5% vs. 76.2% of cells); dexamethasone (2D fibroblast scratch wound closure 50.9% vs. 51.0%); and successful delivery of viable cells (viability of 83% vs. 100%). We additionally developed a finite element model to model the pressure/volume release trend, and demonstrated the effect of membrane stiffness on release. Discussion: Our results demonstrate that the device can consistently administer therapeutics and molecules of various sizes and functions while maintaining their bioactivity, showcasing its potential for repeated, precisely-timed therapeutic delivery.

Mechanotherapy - therapy which uses mechanical forces to produce a remedial or prophylactic effect - has great potential to improve therapeutic outcomes in the fields of regenerative medicine and drug delivery due to its adaptable and tunable nature. In particular, numerous in vivo studies have demonstrated the ability of mechanotherapies to improve functional muscle regeneration and modulate fibrosis. However, the cellular interactions that underlie these tissue level responses are poorly understood. To further harness the potential of mechanotherapies and inform their design and development, a more comprehensive understanding of immune cell responses to mechanical loading is required. Here, we review findings from preclinical investigations of mechanotherapies as both a treatment for muscular injury and as an immunomodulating component of medical implants. We then discuss the mechanosensitive nature of immune cells, emphasizing how mechanical loading and microenvironmental stresses can influence immune signalling pathways in the context of tissue regeneration and fibrosis. Finally, we offer our perspective on the future of the field, including the challenges facing mechanotherapeutic device development and the potential to further broaden the therapeutic targets of mechanotherapies. STATEMENT OF SIGNIFICANCE: Tissue regeneration following injury requires precise regulation of the innate and adaptive immune responses to restore tissue function and prevent fibrosis. Fibrotic tissue alters mechanical properties and impairs physiological function, making fibrosis a critical barrier to effective healing. Mechanotherapy, which uses mechanical forces for therapeutic effect, offers a promising approach to modulate immune responses and improve regenerative outcomes. Emerging evidence suggests that mechanical stimulation can attenuate fibrosis, yet the immune response to mechanical loading is highly dependent on loading parameters such as magnitude, frequency, and duration. Understanding how these parameters influence healing remains a key challenge. This review explores the relationship between mechanotherapy, immune modulation, and fibrosis, highlighting the potential for mechanotherapy to guide wound healing and improve clinical outcomes.

Heart failure with preserved ejection fraction (HFpEF) is a growing health problem worldwide, accounting for half of all heart failure cases. HFpEF patients present with diverse underlying causes and symptoms, making diagnosis and treatment challenging. Current pharmacological therapies are inadequate, while approved device-based therapies have shown limited success due to patient heterogeneity. This underscores the need for improved pre-clinical models, critical for guiding the design and development of effective therapeutic devices. This paper presents an overview of current pre-clinical HFpEF models, including in-silico, in-vitro, ex-vivo, and in-vivo approaches, aimed at advancing the understanding of HFpEF physiology and the development of device-based therapies. We examined each model's ability to replicate key HFpEF characteristics, discuss their respective strengths and limitations, and highlight their role in supporting the creation of clinically relevant solutions. Additionally, the potential of emerging advancements is explored.

Advanced imaging of the heart, including cardiovascular magnetic resonance imaging (CMR), has revolutionized the diagnosis and prognosis for cardiovascular disease1–3. For the past 40 years, CMR has primarily relied on the acquisition of numerous breath-held 2D images resulting in complex scanner operation, patient discomfort, long scan durations, and cumbersome image interpretation4,5. These limitations constrain CMR use to major academic hospital systems and severely limit patient access to CMR, which makes up < 1% of total cardiovascular imaging despite being represented in two thirds of all AHA/ACC guidelines6,7. By leveraging advanced multidimensional physics and artificial intelligence, we overcome these challenges by developing a 30-minute end-to-end automated CMR exam (AutoCMR) that delivers 4D anatomical, functional, and tissue characterization of the whole heart in a single click without breath-holds. AutoCMR was rigorously validated in three cohorts: preclinical large animals, patients scanned in an academic hospital setting with over 40 years of CMR experience, and patients scanned in a community health center with no prior CMR experience. While providing simplified CMR acquisition and automated analysis, we demonstrated that AutoCMR was not significantly different than conventional CMR in imaging biomarkers and human interpretation. With its 4D whole thoracic coverage, we further showcased that AutoCMR can enable next generation patient analytics including personalized digital twins, 3D printing, virtual reality, and automated clinical structured summaries. Due to its inherent scalability, we anticipate AutoCMR will promote the democratization of CMR, increasing patient access for all including underserved health communities, while enabling powerful downstream cutting-edge technologies aimed at personalized medicine.

Soft Robotic Diaphragm-Assist Device In article number 2401087, Diego Quevedo-Moreno, Ellen T. Roche, and colleagues present a soft robotic diaphragm-assist device using fabric-based pneumatic actuators with a 2-step control system for optimized synchronization and support. The system detects breathing initiation to trigger assistance and regulates pressures for inhalation augmentation. Validation is conducted on a custom-built respiratory simulator. This research advances soft robotics in respiratory care, laying the groundwork for next-generation therapeutic devices for diaphragm dysfunction.

Atrial fibrillation (AF) poses significant clinical challenges due to the complex and variable geometry of the left atrial appendage (LAA), whose structure complicates the development of personalized interventions like LAA occlusion (LAAO) for stroke prevention Current reliance on animal models and cadavers for the assessment of left atrium (LA) and LAA to study AF-related disease and interventions raises reproducibility concerns, necessitating the development of high fidelity, physiologically relevant tools. To address this, we present a multimodal framework combining a soft robotic benchtop simulator, a lumped parameter model (LPM), and finite element analysis (FEA) to replicate LA function in sinus rhythm, atrial flutter, and AF. The system integrates 3D-printed, patient-specific LA geometries with soft robotic actuators to reproduce realistic wall motion and hemodynamics. A compact, magnetic resonance imaging (MRI)-compatible flow loop, driven by a soft robotic left ventricle (LV), eliminates bulky and non-physiological pulsatile pumps, allowing precise flow measurements and LAAO device testing under clinically relevant conditions. Complementary LPM and FEA models provide mechanistic insights, quantifying systemic hemodynamic changes and LA wall stress during disease and interventions. The models effectively replicate the clinical markers of atrial dysfunction and arrhythmia disorders, and their physiological accuracy is demonstrated through validation against human imaging and porcine models. The soft robotic LV’s ability to drive the mock flow loop is validated in a hybrid synthetic-biological configuration within a swine circulatory system, where the soft robotic ventricle replaces native ventricular contraction to sustain systemic circulation. This scalable and versatile framework integrates experimental and computational techniques to advance cardiovascular biomechanics, supporting device development, clinical research/training, and personalized AF treatment to improve patient outcomes.

The diaphragm is a critical muscle for respiration, responsible for up to 70% of the inspiratory effort. Standard treatment for patients with severe diaphragm dysfunction is permanently tethering the airway to a mechanical ventilator, which greatly impacts patient autonomy and quality of life. Soft robots are ideal to assist in complex biological functions, such as diaphragm contraction. This article introduces a soft robotic diaphragm‐assist device designed as a therapeutic treatment for diaphragm dysfunction, moreover a clinically relevant respiratory simulator is designed and proposed as a validation and testing tool for this treatment. The device uses fabric‐based pneumatic actuators to provide targeted mechanical assistance during inhalation. A two‐step control system is implemented to optimize synchronization and support: 1) detecting breath intention from the pleural pressure signal to trigger the device and 2) regulating the device's input pressure to assist in inhalation. Using the respiratory simulator, the device demonstrated the ability to restore pleural and abdominal pressures and significantly increased transdiaphragmatic pressure during simulated conditions of diaphragm dysfunction. This research advances the field of soft robotics in respiratory care, providing a foundational platform for the development of next‐generation therapeutic devices aimed at improving the quality of life for patients with diaphragm dysfunction.

Coronary artery disease remains a significant global health concern, leading to high morbidity and mortality, with conditions like spontaneous coronary artery dissection (SCAD) presenting unique challenges in diagnosis and treatment. SCAD, a critical cause of acute coronary syndrome and myocardial infarction, complicates percutaneous coronary intervention (PCI) due to the fragility of the affected vessels. To improve PCI techniques specifically tailored to treat SCAD, we developed an ex-vivo biorobotic coronary artery flow simulator to model SCAD in a benchtop setup. The simulator integrates intraventricular balloons, McKibben soft robotic actuators, and a synchronized flow loop to replicate physiological coronary flow and myocardial wall motion in an explanted porcine heart. After recreating healthy conditions of ventricular wall motion and coronary flow, we induced a coronary dissection model, performed stent implantation, and assessed the intervention deployment using intravascular optical coherence tomography (IV-OCT). This platform can potentially enable precise testing of interventional strategies, improve PCI techniques, examine catheter or PCI-induced dissection, train clinicians, and facilitate testing of novel coronary imaging and intervention devices.

Despite its various potential uses, untethered pneumatics have little practical use due to the speed, power, and controllability limitations of power systems, and system-design restrictions (dead volume contained in pneumatic circuits and hardware needed to handle high actuator pressures). In this work, we present a compact monopropellant engine that localizes pressure generation to actuators. Our approach minimizes dead volume by eliminating pneumatic circuitry and allows high-pressure actuation from a low-pressure fuel source. We introduce the architecture, present a hardware implementation, and prove our pressure-amplification working principle experimentally. We also experimentally demonstrate fast response time (<30ms), high-pressure capability (>100kPa), high flow rate capacity 140 SLM/kg), and the ability to control the rate of gas output and volume of produced gas. Finally, we interface the engine with two distinct soft actuators to demonstrate its plug-and-play ease of use, operating at both high speeds and high pressures, and estimating a promising power output range (~5-25W/kg). With this design, we hope to move closer towards self-contained pneumatic muscles that interface with low-pressure embodied energy storage, which may increase the viability of pneumatic actuation for untethered robotic applications.

Abdominal near-infrared spectroscopy (NIRS) holds promise for early detection of necrotizing enterocolitis and other infant pathologies prior to irreversible injury, but the optimal NIRS sensor design is not well defined. In this study, we develop and demonstrate a computational method to evaluate NIRS sensor designs for infant splanchnic oximetry. We used a finite element (FE) approach to simulate near-infrared light transport through a 3D model of the infant abdomen constructed from computed tomography (CT) images. The simulations enable the measurement of the contrast-to-noise ratio (CNR) for splanchnic oximetry, given a specific NIRS sensor design. A key design criterion is the sensor's source-detector distance (SDD). We calculated the CNR as a function of SDD for two sensor positions near the umbilicus. Contrast-to-noise was maximal at SDDs between 4 and 5 cm, and comparable between sensor positions. Sensitivity to intestinal tissue also exceeded sensitivity to superficial adipose tissue in the 4-5 cm range. FE modeling of abdominal NIRS signals provides a means for rapid and thorough evaluation of sensor designs for infant splanchnic oximetry. By informing optimal NIRS sensor design, the computational methods presented here can improve the reliability and applicability of infant splanchnic oximetry.

Minimally invasive procedures in interventional cardiology are often carried out with tools that lack distal control and visualization methods that introduce noise and uncertainty. This work addresses these complications in the context of cardiac resynchronization therapy, used to treat certain types of heart failure. We introduce a soft robot that actively guides existing interventional tools inside the beating heart. The robot provides controllable distal dexterity through a soft manipulator, as well as passive stability through a stent‐like, expandable stabilization mechanism. Using this platform, existing tools that are normally passive can instead be guided toward a target with high spatial resolution and stability. The operating clinician can thus probe the heart's internal anatomy more methodically than with conventional equipment alone, leading to reduced procedural times. Indeed, results from ex vivo and in vivo studies show reduced overall procedure times, reduced radiation exposure from X‐ray fluoroscopy, and reduced adverse impacts from anatomical probing. The five animal studies further demonstrate that anatomical probing and target localization is the most time‐consuming aspect of cardiac resynchronization therapy and that robotically enhanced dexterity mitigates this primary difficulty.

Preclinical modeling of human circulation has been instrumental in advancing cardiovascular medicine. Alongside clinical research, the armamentarium of computational (e.g., lumped parameter or computational fluid dynamics) and experimental (e.g., benchtop or animal) models have substantially enhanced our understanding of risk factors and root causes for circulatory diseases. Recent innovations are further disrupting the boundaries of these preclinical models toward patient-specific simulations, surgical planning, and postoperative outcome prediction. This fast-paced progress empowers preclinical modeling to increasingly delve into the intricacies of single ventricle physiology, a rare and heterogeneous congenital heart disease that remains inadequately understood. Here, we review the current landscape of preclinical modeling (computational and experimental) proposed to advance clinical management of a prominent yet complex subset of single ventricle physiology: patients who have undergone Fontan-type surgical corrections. Further, we explore recent innovations and emerging technologies that are poised to bridge the gap between preclinical Fontan modeling and clinical implementation.

This study explores two methods of predicting non-cyclic upper-body motions using classification algorithms. Exoskeletons currently face challenges with low fluency, hypothesized to be in part caused by the lag in active control innate in many leader-follower paradigms seen in today's systems, leading to energetic inefficiencies and discomfort. To address this, we employ k-nearest neighbor (KNN) and deep learning models to predict motion characteristics, such as magnitude and category, from surface electromyography (sEMG) signals. Data were collected from six muscles located around the elbow. The sEMG signals were processed to identify significant activation changes. Two classification approaches were utilized: a KNN algorithm that categorizes motion based on the slopes of processed sEMG signals at change points and a deep neural network employing continuous categorization. Both methods demonstrated the capability to predict future voluntary non-cyclic motions up to and beyond commonly acknowledged electromechanical delay times, with the deep learning model able to predict, with certainty at or beyond 90%, motion characteristics even prior to myoelectric activation of the muscles involved. Our findings indicate that these classification algorithms can be used to predict upper-body non-cyclic motions to potentially increase machine interfacing fluency. Further exploration into regression-based prediction models could enhance the precision of these predictions, and further work could explore their effects on fluency when utilized in a tandem or wearable robotic application.

PURPOSE: Atrial fibrillation (AF) is the most common chronic cardiac arrhythmia that increases the risk of stroke, primarily due to thrombus formation in the left atrial appendage (LAA). Left atrial appendage occlusion (LAAO) devices offer an alternative to oral anticoagulation for stroke prevention. However, the complex and variable anatomy of the LAA presents significant challenges to device design and deployment. Current benchtop models fail to replicate both anatomical variability and physiological hemodynamics, limiting their utility. This study introduces a novel left atrial cardiac simulator that incorporates patient-derived LAA models within a benchtop circulatory flow loop, enabling high-fidelity LAAO device testing and development. METHODS: A rigid, patient-derived left atrium (LA) model was 3D printed from segmented MRI data and modified to accommodate attachment of patient-specific LAA models. A library of LAA geometries was fabricated using silicone casting techniques to replicate the mechanical properties of native tissue. The LA-LAA model was integrated into a circulatory flow loop equipped with a pulsatile pump, pressure sensors, and flow probes, allowing real-time hemodynamic analysis. System tunability was demonstrated by varying heart rate, stroke volume, resistance, and compliance to simulate physiological and pathological conditions. RESULTS: The simulator accurately replicated LA pressure and flow waveforms, closely approximating physiological conditions. Changes in heart rate, stroke volume, and compliance effectively modulated LAP and LA inflow before and after LAAO. Distinct pressure and flow waveforms were observed with different LAA geometries. Hemodynamic analysis revealed increased left atrial pulse pressure after occlusion, with the greatest increase occurring after complete exclusion of the LAA. The simulator facilitated the evaluation of LAAO device performance, including metrics such as seal and PDL, and served as an effective training tool for iterative device deployment and recapture with visual and imaging-guided feedback. CONCLUSIONS: The left atrial cardiac simulator offers a highly tunable and realistic platform for testing and developing LAAO devices. It also serves as an effective procedural training tool, allowing for the simulation of patient-specific anatomical and hemodynamic conditions. By enabling these advanced simulations, the simulator enhances pre-procedural planning, device sizing, and placement. This innovation represents a significant step toward advancing personalized medicine in atrial fibrillation management and improving LAAO outcomes.

Therapeutic proteins, the fastest growing class of pharmaceuticals, are subject to rapid proteolytic degradation in vivo, rendering them inactive. Sophisticated drug delivery systems that maintain protein stability, prolong therapeutic effects, and reduce administration frequency are urgently required. Herein, a mechanoresponsive hydrogel is developed contained within a soft robotic drug delivery (SRDD) device. In a step-change from previously reported systems, pneumatic actuation of this system releases the cationic therapeutic protein Vascular Endothelial Growth Factor (VEGF) in a bioactive form which is required for therapeutic angiogenesis, the growth of new blood vessels, in numerous clinical conditions. The ability of the SRDD device to release bioactive VEGF in a spatiotemporal manner from the hydrogel is tested in diabetic rats - a model in which angiogenesis is difficult to stimulate. Daily actuation of the SRDD device in the diabetic rat model significantly increased cluster of differentiation 31+ (CD31+) blood vessel number (p = 0.0335) and the diameter of alpha-smooth muscle actin+ (α-SMA+) blood vessels (p = 0.0025) compared to passive release of VEGF from non-actuated devices. The SRDD device combined with the mechanoresponsive hydrogel offers the potential to deliver an array of bioactive therapeutics in a spatiotemporal manner to mimic their natural release in vivo.

Aortic stenosis (AS) is the most common valvular heart disease in developed countries. High-fidelity preclinical models can improve AS management by enabling therapeutic innovation, early diagnosis, and tailored treatment planning. However, their use is currently limited by complex workflows necessitating lengthy expert-driven manual operations. Here, we propose an AI-powered computational framework for accelerated and democratized patient-specific modeling of AS hemodynamics from computed tomography (CT). First, we demonstrate that the automated meshing algorithms can generate task-ready geometries for both computational and benchtop simulations with higher accuracy and 100 times faster than existing approaches. Then, we show that the approach can be integrated with fluid-structure interaction and soft robotics models to accurately recapitulate a broad spectrum of clinical hemodynamic measurements of diverse AS patients. The efficiency and reliability of these algorithms make them an ideal complementary tool for personalized high-fidelity modeling of AS biomechanics, hemodynamics, and treatment planning.

Abstract Swellable microneedles (MNs) expand to mechanically interlock with wet biological tissue, offering improved adhesion and enhanced drug delivery over non-swellable counterparts. This study numerically evaluates how the material and geometric parameters of swellable MN arrays influence shape change. Using finite element simulation, MNs were subjected to unconstrained swelling, approximated via a thermal-strain analogy. Optimal MN design must support mechanical interlocking to prevent dislodgement. We observed that wet in vivo environments induce unwanted swelling-mediated curvature, hindering contact and interlocking. We quantified this bending and calibrated gel material swellability using experimental data. To counteract curling, we introduced a design approach to shift the direction of the unwanted curling and improve MN array conformability.

Soft Robotic Drug Delivery Micro-CT imaging of the hybrid hydrogel actuator (HHA) prototype showcasing its robust, flexible adhesion to tissue during dynamic actuation. This device enables tunable, mechanoresponsive drug delivery directly to the target site, presenting a transformative approach that integrates precisely controlled drug delivery with mechanical stimulation for enhanced localized therapeutic interventions. More details can be found in article number 2303301 by Ellen T. Roche and colleagues.

Calcification has significant influence over cardiovascular diseases and interventions. Detailed characterization of calcification is thus desired for predictive modeling, but calcium deposits on cardiovascular structures are still often manually reconstructed for physics-driven simulations. This poses a major bottleneck for large-scale adoption of computational simulations for research or clinical use. To address this, we propose an end-to-end automated image-to-mesh algorithm that enables robust incorporation of patient-specific calcification onto a given cardiovascular tissue mesh. The algorithm provides a substantial speed-up from several hours of manual meshing to ~1 min of automated computation, and it solves an important problem that cannot be addressed with recent template-based meshing techniques. We validated our final calcified tissue meshes with extensive simulations, demonstrating our ability to accurately model patient-specific aortic stenosis and Transcatheter Aortic Valve Replacement. Our method may serve as an important tool for accelerating the development and usage of personalized cardiovascular biomechanics.

Tetralogy of Fallot is a congenital heart disease affecting newborns and involves stenosis of the right ventricular outflow tract (RVOT). Surgical correction often widens the RVOT with a transannular enlargement patch, but this causes issues including pulmonary valve insufficiency and progressive right ventricle failure. A monocusp valve can prevent pulmonary regurgitation; however, valve failure resulting from factors including leaflet design, morphology, and immune response can occur, ultimately resulting in pulmonary insufficiency. A multimodal platform to quantitatively evaluate the effect of shape, size, and material on clinical outcomes could optimize monocusp design. This study introduces a benchtop soft biorobotic heart model, a computational fluid model of the RVOT, and a monocusp valve made from an entirely biological cell-assembled extracellular matrix (CAM) to tackle the multifaceted issue of monocusp failure. The hydrodynamic and mechanical performance of RVOT repair strategies was assessed in biorobotic and computational platforms. The monocusp valve design was validated in vivo in ovine models through echocardiography, cardiac magnetic resonance, and catheterization. These models supported assessment of surgical feasibility, handling, suturability, and hemodynamic and mechanical monocusp capabilities. The CAM-based monocusp offered a competent pulmonary valve with regurgitation of 4.6 ± 0.9% and a transvalvular pressure gradient of 4.3 ± 1.4 millimeters of mercury after 7 days of implantation in sheep. The biorobotic heart model, in silico analysis, and in vivo RVOT modeling allowed iteration in monocusp design not now feasible in a clinical environment and will support future surgical testing of biomaterials for complex congenital heart malformations.

Aortic stenosis (AS) is the most common valvular heart disease in developed countries. High-fidelity preclinical models can improve AS management by enabling therapeutic innovation, early diagnosis, and tailored treatment planning. However, their use is currently limited by complex workflows necessitating lengthy expert-driven manual operations. Here, we propose an AI-powered computational framework for accelerated and democratized patient-specific modeling of AS hemodynamics from computed tomography. First, we demonstrate that our automated meshing algorithms can generate task-ready geometries for both computational and benchtop simulations with higher accuracy and 100 times faster than existing approaches. Then, we show that our approach can be integrated with fluid-structure interaction and soft robotics models to accurately recapitulate a broad spectrum of clinical hemodynamic measurements of diverse AS patients. The efficiency and reliability of these algorithms make them an ideal complementary tool for personalized high-fidelity modeling of AS biomechanics, hemodynamics, and treatment planning.

Current clinically used electronic implants, including cardiac pacing leads for epicardial monitoring and stimulation of the heart, rely on surgical suturing or direct insertion of electrodes to the heart tissue. These approaches can cause tissue trauma during the implantation and retrieval of the pacing leads, with the potential for bleeding, tissue damage, and device failure. Here, we report a bioadhesive pacing lead that can directly interface with cardiac tissue through physical and covalent interactions to support minimally invasive adhesive implantation and gentle on-demand removal of the device with a detachment solution. We developed 3D-printable bioadhesive materials for customized fabrication of the device by graft-polymerizing polyacrylic acid on hydrophilic polyurethane and mixing with poly(3,4-ethylenedioxythiophene):poly(styrenesulfonate) (PEDOT:PSS) to obtain electrical conductivity. The bioadhesive construct exhibited mechanical properties similar to cardiac tissue and strong tissue adhesion, supporting stable electrical interfacing. Infusion of a detachment solution to cleave physical and covalent cross-links between the adhesive interface and the tissue allowed retrieval of the bioadhesive pacing leads in rat and porcine models without apparent tissue damage. Continuous and reliable cardiac monitoring and pacing of rodent and porcine hearts were demonstrated for 2 weeks with consistent capture threshold and sensing amplitude, in contrast to a commercially available alternative. Pacing and continuous telemetric monitoring were achieved in a porcine model. These findings may offer a promising platform for adhesive bioelectronic devices for cardiac monitoring and treatment.

Epicardial drug delivery offers the potential to increase drug concentration at the target site, decrease systemic side effects, and reduce overall drug usage and cost. However, controlled drug delivery to the epicardium remains challenging, limiting exploration as a feasible administration route. Existing epicardial delivery systems lack precise control over drug dosing and responsiveness to local physiological cues. To address these limitations, we present SmartSleeve, a sutureless, soft robotic drug-delivery device that enables switchable on-off drug delivery in response to epicardial electrocardiogram (ECG) sensing. SmartSleeve is composed of an elastomeric soft robotic actuator with self-sealing therapy reservoirs, coupled to an adhesive bioelectronic interface. With SmartSleeve, we demonstrate controlled epicardial delivery of therapeutic agents in response to real-time ECG changes in animal models. SmartSleeve provides a sophisticated platform for studying epicardial drug delivery, with the potential to transform cardiac therapy delivery in clinical applications, including postoperative arrhythmia management and chronic inotropic support.

Our understanding of cardiac remodeling processes due to left ventricular pressure overload derives largely from animal models of aortic banding. However, these studies fail to enable control over both disease progression and reversal, hindering their clinical relevance. Here, we describe a method for progressive and reversible aortic banding based on an implantable expandable actuator that can be finely tuned to modulate aortic banding and debanding in a rat model. Through catheterization, imaging, and histologic studies, we demonstrate that our platform can recapitulate the hemodynamic and structural changes associated with pressure overload in a controllable manner. We leveraged soft robotics to enable noninvasive aortic debanding, demonstrating that these changes can be partly reversed because of cessation of the biomechanical stimulus. By recapitulating longitudinal disease progression and reversibility, this animal model could elucidate fundamental mechanisms of cardiac remodeling and optimize timing of intervention for pressure overload.

Calcification has significant influence over cardiovascular diseases and interventions. Detailed characterization of calcification is thus desired for predictive modeling, but calcified heart meshes for physics-driven simulations are still often reconstructed using manual operations. This poses a major bottleneck for large-scale adoption of computational simulations for research or clinical use. To address this, we propose an end-to-end automated meshing algorithm that enables robust incorporation of patient-specific calcification onto a given heart mesh. The algorithm provides a substantial speed-up from several hours of manual meshing to $\sim$1 minute of automated computation, and it solves an important problem that cannot be addressed with recent template registration-based heart meshing techniques. We validated our final calcified heart meshes with extensive simulations, demonstrating our ability to accurately model patient-specific aortic stenosis and Transcatheter Aortic Valve Replacement. Our method may serve as an important tool for accelerating the development and usage of physics-driven simulations for cardiac digital twins.

The field of durable mechanical circulatory support (MCS) has undergone an incredible evolution over the past few decades, resulting in significant improvements in longevity and quality of life for advanced heart failure patients. Despite these successes, substantial opportunities for further improvements remain, including in pump design and ancillary technology, peri- and post-operative management, and the overall patient experience. Ideally, durable MCS devices would be fully implantable, automatically controlled, and minimize need for anticoagulation. Reliable and long term total artificial hearts for bi-ventricular support would be available; and surgical, peri- and post-operative management would be informed by the individual patient phenotype along with computational simulations. In this review, we summarize emerging technological innovations in these areas, focusing primarily on innovations in late preclinical or early clinical phases of study. We highlight important considerations that the MCS community of clinicians, engineers, industry partners and venture capital investors should consider to sustain evolution of the field.