Recent advances in microfluidics and bioelectronics for three-dimensional organoid interfaces

Abstract

Organoids offer a promising alternative in biomedical research and clinical medicine, with better feature recapitulation than 2D cultures. They also have more consistent responses with clinical results when compared to animal models. However, major challenges exist in the longevity of culture, the reproducibility of organoid properties, and the development of non-disruptive monitoring methods. Recent advances in materials and microfabrication methods, such as 3D printing and compressive buckling, have enabled three-dimensional (3D) interfaces of microfluidics and bioelectronics to manipulate and monitor these biological models in exciting ways. These advanced systems have great potential for applications in drug delivery, personalized medicine, and disease modeling. We conclude with important future considerations to generate longevity using further technological development in organoid and spheroid models.

I. ORGANOID CULTURE OVERVIEW

The development of in vitro models capable of replicating complex cellular assemblies of the human body has been a longstanding goal in biotechnology. While two-dimensional (2D) cultures have been standard in a laboratory setting, three-dimensional (3D) multicellular systems more accurately recapitulate the spatial organization, cell–cell interactions, and biochemical gradients present in native tissues, making them more reliable platforms for biological investigation.^1,2 Advantages of 3D in vitro models over planar ones include improved insight into tissue and disease development, enhanced understanding of cell–cell interactions within the extracellular matrix (ECM) for the design of advanced biomaterials, and improved prediction accuracy in preclinical drug testing.³ Moreover, the use of 3D tissue models can potentially reduce reliance on animal models. In this context, the Food and Drug Administration (FDA), especially the recent FDA Modernization Act 2.0, has approved the use of alternative systems, including organoid models, as suitable proof-of-concept drug screening models, to improve the efficiency of drug development and transition into clinical trials.⁴ These advancements underscore the importance of developing and refining high-fidelity organoid and spheroid models, positioning them as critical tools in the future of drug discovery and biotechnology.

A. Sources and generation

The general structure of 3D biological model development is (1) selection of the source cell type, (2) generation of a multidimensional structure, (3) period of extended growth, and (4) evaluation. These models are described as organoids, spheroids, and assembloids based on the cell type and generation method. Major cell types include immortalized cell lines, sourced primary cells, and induced pluripotent stem cells (iPSCs). Each cell type has different properties, including growth rate, accessibility, and resource requirements, that may make it an advantageous choice. Primary cells are challenging to source and often require customized protocols to effectively grow in laboratory settings.⁵ Similarly, iPSCs require extensive protocols, including substantial growth factors, trained personnel, and time to reach sufficient cell numbers that recapitulate in vivo niche conditions.^6,7 In contrast, immortalized cell lines are more accessible, less resource-intensive, and proliferate quickly; however, they lack the diversity of cell types and structural complexity found in other sources. The selection of cell types depends on balancing available resources and the desired model complexity. These considerations are essential for both defining the model as a spheroid, organoid, or assembloid and designing an appropriate generation process to replicate the desired spatial structures or dynamic cell interactions.

3D structure formation typically begins after the introduction of relevant growth factors or genetic modifications that initiate specific developmental or disease states, leading the cells toward the desired differentiation. A common approach to achieving 3D growth is spontaneous aggregation [Fig. 1(a)], achieved by (ia) culturing cells on low-attachment substrates to promote clustering,⁸ (ib) confining cells within low-attachment, small-surface-area wells in 96- or 384-well plates to enhance cell–cell interactions,⁹ or (ii) exploiting gravity-driven aggregation in hanging-drop cultures.¹⁰ Less passive approaches include membrane-based methods [Fig. 1(b)], wherein a synthetic ECM scaffold is introduced to promote cell assembly^11,12 or thermoresponsive gels are utilized to lift cells upon gel dissolution.¹³ The ECM, generally Matrigel or a synthetic matrix, provides a layer with interspersed cells that support cell grouping into organoid formations and can have variable structures that limit reproducibility between batches.¹⁴ Mechanical methods, represented in Fig. 1(c), include centrifugation into a cell pellet and bioreactor systems, where continuous stirring groups the cells into aggregates during culture. After formation, a period of growth continues to allow further properties and interactions to develop. Most critically, organoids and spheroids experience growth restriction due to the limited diffusion of oxygen and nutrients in the core region, resulting in constrained size and maturation.

FIG. 1.

Formation methods for spheroid and organoid models. (a) Methods for spontaneous generation of three-dimensional organoid structures including (ia) low-attachment culture surfaces, (ib) low-attachment individual well formation in a 96- or 384-well plate, and (ii) hanging drop. (b) Membrane-based aggregation methods using (i) charged basement membranes or (ii) thermal lifting processes. (c) Agitation-based aggregation of organoids based on (i) initial centrifugation to form a pellet or (ii) continuous mechanical stirring through a bioreactor. Created in BioRender.com.

As mentioned earlier, the classification of multicellular structures such as spheroids or organoids depends on their structural complexity. Nomenclature has been widely debated between spheroids and organoids, especially in subfields such as brain organoids, where a strong understanding of relevant terms for different parts of the brain being cultured becomes necessary to verify results between research groups.¹⁵ While our focus is primarily on 3D organoid interfaces, we also include spheroid interfaces in our discussion. As the breadth of organoid-based studies has expanded, analytical methods have also extended to fully capture, describe, and compare intricate multidimensional cellular properties.

B. Traditional analysis methods

Analytical methods are essential for establishing the identity and characteristics of cells, chemicals (such as reactive oxygen species), and processes (such as stem cell differentiation or cell maturation) that are key to forming organoid structures. These methods can be categorized according to their focus on compositional, structural, or genetic information, as illustrated in Fig. 2. Given that these methods have been reviewed extensively elsewhere,^1,16,17 we provide only a brief overview here.

FIG. 2.

Evaluation methods for analyzing organoid structures. (a) Composition-targeted methods, including examples of targeted labeling and evaluation of protein content. (b) Structure-targeted methods, including whole imaging with fluorescent labels or tissue sectioning for slide-based staining. (c) Sequence-targeted evaluation, based on the dissociation of genetic material from individual cells for eventual DNA or RNA sequencing. Created in BioRender.com.

Composition-targeted methods utilize labeled molecules that bind to proteins released to the culture media by the cells or extracted from the cell structure, producing either colorimetric or fluorescent signals.¹⁶ These assays are critical for assessing organoid function, as they can quantify metabolites or track the flux of signaling molecules such as calcium. Common assays include the 3–(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT) assay, which measures metabolic activity, and the Western Blot assay, which quantifies specific proteins, as shown in Fig. 2(a). Structure-targeted methods typically employ high-resolution imaging techniques to visualize organoids, either through label-free approaches such as bright-field microscopy or by utilizing specific stains used in immunofluorescence or histology analyses.¹⁶ Such techniques provide detailed spatial and temporal insights into the organoid structure, but often require specialized equipment or skills due to the complexity of analyzing 3D structures.¹⁷ For example, specialization is required to process an organoid, which can involve physically sectioning it or digitally reconstructing its three-dimensional structure from stacked image slices, as illustrated in Fig. 2(b). Sequence-targeted analysis methods [Fig. 2(c)] involve sequencing genetic materials extracted from either the bulk solution surrounding an organoid or from individual cells through techniques including quantitative polymerase chain reaction (q-PCR) or single-cell ribonucleic acid (RNA) sequencing.¹⁷ These techniques generate insights into the cell types and physiological expression post-differentiation but also require advanced equipment and expertise.

C. Common organoid challenges

Despite their widespread use, traditional analytical methods face several limitations, particularly regarding culture longevity and spatiotemporal resolution. For example, many of these techniques necessitate removing organoids from their culture environment, thus halting further development and precluding longitudinal analysis of individual organoids.¹⁸ Moreover, these methods are resource-intensive, requiring reagents, specialized equipment, and skilled personnel for processes such as staining, advanced microscopy, and genetic sequencing.¹⁹ The intrinsic heterogeneity of organoids, reflected in their variability in size, structure, and cellular composition, also complicates reproducibility, often necessitating large sample sizes for statistically meaningful results.²⁰ Additionally, the laborious nature of organoid production and the associated data analysis pose further challenges.

Given these limitations, there is an urgent need for new tools and technologies that enable user-friendly, real-time, and in situ monitoring of key biological processes, including proliferation, differentiation, and region-specific tissue patterning in functional organoid models. This review presents recent advances in integrating innovative microfluidics and bioelectronics with organoid systems, highlighting how these technologies overcome current analytical limitations. We define the engineered physical and functional boundaries where organoid tissues interface with microfluidic and bioelectronic systems as 3D organoid interfaces. We focus on innovative platforms that leverage engineering tools to enable sensing, perfusion, and scalability, thereby supporting key biological objectives such as reproducibility, physiological relevance, and structural and functional complexity (Fig. 3), which are essential for establishing organoids as standardized models. Finally, we explore the convergence of these technologies with emerging applications and propose strategies for building integrated 3D platforms that improve both organoid growth and characterization.

FIG. 3.

Summary of biological and engineering objectives for organoid technologies.

II. MICROFLUIDIC ORGANOID INTERFACES

Current human organoids face limitations in structural maturity, functionality, size, and heterogeneity. They are largely attributed to the lack of stable, perfusable vascularization—an essential feature of native tissue that enables efficient nutrient and oxygen delivery, metabolic waste removal, and distribution of growth factors necessary for sustaining organoid viability, growth, and complexity.^1,19,21 For example, most organoids develop a necrotic core as the increased metabolic demands for cells located in the center cannot be met by passive diffusion of nutrition (effective distance of diffusion ≤~300 μm). In addition, tissue growth and identity rely on specific growth factors, including Wnt, Sonic Hedgehog, and Transformation Growth Factor β (TGF-β),¹⁶ which play a vital role in regulating stem cell differentiation and enabling activities (e.g., protein signaling) that coordinate the formation of complex organoid structures. Spatiotemporal control over growth factor distribution is essential to mimic the in vivo microenvironment, which is particularly relevant for applications in personalized medicine and drug development.^16,22

Achieving the level of precision to mimic the human body in terms of perfusion and multi-organ modeling is a complicated goal, especially considering the need for reproducible conditions to make statistical conclusions. Microfluidic technologies have been introduced to address challenges around three primary objectives: (1) enabling precise control over organoid formation for reproducible size and development, (2) improving continuous perfusion throughout organoid development to generate enhanced longevity of culture and maturity of organoid, and (3) developing systemic flow designs for more complex models to better mimic interactions in the human body.²³ Generating this realism often involves the employment of advanced microscale fabrication techniques, including 3D printing, photolithography and molding, compressive buckling, and kirigami-based folding to create the necessary spacing between organoids for their growth and reproducibility, as well as organoid organization and size.^24–27 In this section, we will discuss recent microfluidic technologies that advance precision delivery, sustained perfusion, and integrated multi-organ systems.

A. Reproducible formation

The development of organoid-on-a-chip devices has significantly advanced the precision and speed of organoid formation. In particular, polymer-based microfluidic chips offer both high customizability and predictable liquid handling. These systems can be patterned with microscale features using techniques, including photolithography, two-photon lithography, micro-milling, and 3D printing. Computational tools such as COMSOL allow for the fine-tuning and prediction of fluid dynamics, along with properties including temperature distribution, concentration gradients, and flow velocity. These precisely engineered microfluidic platforms address key challenges in organoid isolation and reproducibility by enabling the prediction and control of cell aggregation in designated features, such as microwells, over time. This is in contrast to traditional bulk spheroid formation methods that often produce heterogeneous populations, requiring large sample sizes to achieve statistical significance, creating an inefficiency that poses challenges in clinical and pharmaceutical contexts.

One notable device by Prince et al., aiming to rapidly form large numbers of spheroids with minimal handling, utilized polydimethylsiloxane (PDMS) microfluidic chips with microwells to generate MCF-7 spheroids of uniform diameter (100 or 300 μm) within 72 h [Figs. 4(a) and 4(b)]. The chip design also incorporated a tunable secondary layer of cascading microfluidics that created a concentration gradient of doxorubicin (DOX) across samples for rapid screening of dosage efficacy on cancer cells.²⁸ This PDMS microwell system was further applied to patient-derived breast cancer organoids to evaluate the patient-specific efficacy of Eribulin [Fig. 4(c)].²⁸ A similar microwell-based design enables the formation of pancreatic duct-like organoids with a consistent size (~50 μm in diameter) over a 30-day period.³⁰ The spatial separation of microwells limits interaction between organoids, improving reproducibility and enabling high-quality imaging by forming organoids through shear-force and gravity-induced cell localization. Beyond controlled formation, recent efforts have also focused on post-formation size standardization.²¹ For example, Yan et al. reported a sheath-flow microfluidic device integrated with real-time imaging and analysis to detect organoid size and automatically sort them into various collection chambers, yielding a more size-homogenous final organoid population.²¹

FIG. 4.

Reproducibility-targeted microfluidic platforms using microwell arrays and automated organoid generation. Microwell systems: (a) Experimental workflow for MCF-7 or patient-derived spheroid formation with Eribulin stratification. (b) PDMS chip layout to generate uniform-sized spheroids within a 72 h time frame.²⁸ (c) Patient-specific tumor spheroid viability in response to Eribulin exposure.²⁸ Automated organoid formation: (d) Schematic of biopsy processing to a 96-well culture plate. (e) Microfluidic droplet printer design.²⁹ (f) Growth of patient-derived organoids over 7 days. Scale bar, 400 μm.²⁹ Parts (a) and (d) created in BioRender.com. Parts (b) and (c) reproduced with permission from Prince et al., Adv. Healthc. Mater. 11, e2101085 (2022). Copyright 2022 Wiley-VCH. Parts (e) and (f) reproduced with permission from Jiang et al., Cell Rep. Med. 1, 100161 (2020). Copyright 2020 Elsevier.

Beyond chip-based systems, another promising approach for achieving organoid reproducibility and rapid precision forming is droplet organoid generation.³¹ These droplets generally consist of an inner culture chamber with cells that will form the organoid and a shell layer that is typically produced from alginate or collagen. Formation of these droplets can be achieved through microfluidic planar focusing or automated droplet printing methods. Figure 4(d) shows an example of using a 3D droplet printer to generate Matrigel-embedded patient-derived organoids (PDOs) in a 96-well plate.²⁹ The 3D printing process leverages thermally controlled phase transitions to form droplets that solidify prior to printing, as shown in Fig. 4(e), which creates nearly identical droplets in size and composition. The resulting organoids retained patient-specific characteristics including size, structure, and protein expression that replicate the native tissue [Fig. 4(f)]. Several comprehensive reviews focusing on droplet-based organoid technologies are available.^31,32 While droplet systems have the benefit of creating large quantities of uniformly sized organoids rapidly, limitations include restricted growth time and the difficulty of introducing differentiation factors without removing the protective shell layer. Another notable recent example showed the formation of iPSC droplets that could be differentiated into either neural, cardiac, or liver organoids (LOs) from the same initial high-throughput droplet formation.³³

Both chip-based and droplet-based organoid culture systems offer cost-effective solutions by minimizing reagent consumption during culture.³⁴ Notably, organoid-on-a-chip platforms also enable continuous nutrient perfusion and waste removal, which supports long-term viability and physiological relevance. This advantage leads into the next discussion: organoid perfusion chips, which use compact, localized microfluidic designs to further enhance culture conditions and enable real-time monitoring through integrated sensors.

B. Enhancing perfusion strategies for organoid culture

To address the limitations of nutrient and oxygen delivery in 3D organoid cultures, various perfusion strategies have been developed. Among these, micropillar-based systems, which can either function solely as a mechanism for spacing (typically made of solid and rigid PDMS) or serve as permeable, synthetic vasculature (made of soft polyethylene glycol diacrylate) for organoids to grow around, have shown promise.^35,36 The inter-pillar spacing plays a critical role in regulating the size of the organoid, with one example supporting the formation of 200 μm spheroids that remain isolated for up to 30 days of culture.³⁷ Beyond controlling size, the spacing of organoids also simplifies the ease of individual imaging and minimizes the inter-organoid interactions during differentiation, improving nutrition consistency.^35,38 Grebenyuk et al. demonstrated the use of polyethylene glycol diacrylate (PEGDA) micropillars in a perfused grid-like array for in situ differentiation and maturation of organoids [Figs. 5(a) and 5(b)]. Fluorescence staining for apoptotic factor Caspase 3 and hypoxia marker H1F1-α in Fig. 5(c) showed reduced cell death and oxygen limitation, suggesting optimal nutrient and oxygen access. Similar staining after 62 days under flow conditions showed enhanced maturation and native brain tissue fidelity in these organoids.³⁶ More recently, micropillar-based perfusion was reported to achieve consistent exposure to bisphenol S (BPS) in developing neural organoids, where soft PDMS micropillars maintained stable organoid positioning under continuous flow for 34 days.⁴⁰ Overall, the increased perfusion and spatial arrangement in these pillar systems optimize nutrient and oxygen delivery, waste removal, and size consistency, improving the reproducibility of properties between organoids.

FIG. 5.

Approaches to improving perfusion through organoids. Micropillar system: (a) Formation overview from seeded embryoid bodies to fully developed tissue with perfusion. (b) Flexible, permeable polyethylene glycol diacrylate (PEGDA) micropillars capable of providing synthetic vascularization.³⁶ (c) Visualization of tissue growth with and without perfusion through the flexible pillars, showing the lack of cl. Csp3 around the center and limited presence of H1F1α after 62 days Scale bars, 1 mm.³⁶ Vascularization: (d) Overview of the vascular incorporation process described by Quintard et al. (e) Blood vessel organoid (BVO) positioned in a cyclic olefin copolymer (COC) chip prior to incorporation of human umbilical vein endothelial (HUVEC) cells Scale bars, 1 cm for the upper left image, 1 mm for the upper right image, and 100 μm for the lower images.³⁹ (f) Angiogenesis imaging of formed endothelial networks in the on-chip culture chamber over a period of 30 days Scale bars, 200 μm.³⁹ Parts (a) and (d) created in BioRender.com. Parts (b) and (c) reproduced with permission from Grebenyuk et al., Nat. Commun. 14, 193 (2023). Copyright 2023 Nature Publishing Group. Parts (e) and (f) reproduced with permission from Quintard et al., Nat. Commun. 15, 1452 (2024). Copyright 2024 Nature Publishing Group.

An alternative approach to designing perfusion systems is the co-culture of organoids with vascular cells, which establish natural perfusion pathways both on the surface and within the organoid. By creating branching networks, these vascular cells provide biologically relevant perfusion structures for nutrient transport, addressing the limitations of nutrient access within the organoid core.⁴¹ Improved nutrient delivery supports the growth of larger organoids and enables longer culture durations, ultimately promoting the development of mature tissue features and enhancing disease modeling. For instance, vascularized pancreatic organoids exhibit luminal structure formation, fluid secretion, and disease traits consistent with cystic fibrosis.⁴² Vascular cells can initially be separated from the organoids using a semi-permeable membrane in a multi-chamber model while being connected to the nutrient inlet supply.^43,44 Alternatively, vascular cells can be co-cultured directly with organoids in the same culture chamber, as demonstrated in blood vessel organoids [Fig. 5(d)].³⁹ In this design, a goblet-like chamber houses the organoid [Fig. 5(e)], while human umbilical vein endothelial cells (HUVECs) are flowed through adjacent microchannels, subsequently attaching to the chamber surfaces and forming vascular networks. Perfusion of nutrients throughout the organoid was demonstrated by tracking 1 μm fluorescent, which are much larger than most proteins or sugars, circulating throughout the structure [Fig. 5(f)]. Despite its effectiveness in delivering nutrients to the organoid core, this vascular co-culture strategy presents challenges for subsequent analyses, such as isolating the target organoid for optical evaluation. Moreover, co-culture vascularization is better suited to the development of spheroids that are already differentiated, where precise spatial control of growth factors is less critical.

C. Systemic interactions

Co-culture microfluidic chips can also be designed with multiple chambers separated by a permeable membrane, allowing for crosstalk between different types of organoids to study systemic responses. Figures 6(a) and 6(b) show an example of such a device, which integrates microwells with separated chambers to examine the unintended side effects of the antidepressant clomipramine on cardiac organoids (COs).⁴⁵ In this setup, the chambers are separated by a polycarbonate permeable membrane: liver organoids metabolize the drug, and the resulting byproducts are introduced to the cardiac organoids. As shown in Fig. 6(c), cardiac organoids in the co-cultured system including liver organoids exhibited more significant damage, highlighting the importance of multi-organ drug testing to determine effective and safe treatment concentrations for patients. On-chip designs of microfluidic exchange are crucial for modeling multi-organ interaction models and hold promise for ultimately reducing or replacing animal testing.

FIG. 6.

Systemic interaction models. Multi-chamber system: (a) Side-view schematic showing permeable membrane-separated culture chambers for liver and cardiac organoids to study multi-organ effects of clomipramine. (b) Images of the overall device layout and assembled PDMS device with polycarbonate porous membrane separation.⁴⁵ (c) Viability of liver organoids (LOs) and cardiac organoids (COs) after 24 and 48 h of clomipramine treatment, showing significant cardiac effects. Body-on-a-chip system: (d) Schematic overview of a three-organ chip with capecitabine treatment to show the necessity of multiple organoids in a system. (e) Image of the three-organoid adhesive film-based (AFB) microfluidic chip connected through external tubing.⁴⁶ (f) Imaging results showing damage to cardiac and lung tissue in a system containing liver organoids, which does not occur in their absence Scale bars, 100 μm.⁴⁶ Parts (a) and (d) created in BioRender.com.⁴⁵ Parts (b) and (c) reproduced with permission from Yin et al., Lab Chip. 21, 571 (2021). Copyright 2021 Royal Society of Chemistry. Parts (e) and (f) reproduced with permission from Rajan et al., Acta Biomater. 106, 124 (2020). Copyright 2020 Elsevier.

Multi-well microfluidic models can also be used to study the broader impacts of cross-organ toxicity and incorporate more than three-organoid types for body-on-a-chip studies. These platforms enable nutrient and by-product exchange between different organs, representing an essential step toward comprehensive in vitro drug evaluations as alternatives to animal studies. For instance, drugs primarily targeting lungs can have unexpected adverse effects on other organs, most commonly the heart, brain, and liver.^46,47 A flexible adhesive film-based (AFB) chip design shown in Fig. 6(d) demonstrates this principle. The chip was first patterned by a computer-controlled razor plotter and then folded to create channels with a height of 560 μm from four stacked layers. The film structure is sandwiched between a glass slide base and a laser-cut polymethyl methacrylate (PMMA) cover, with external tubing connecting the chambers [Fig. 6(e)]. Using this system, the cross-toxicity of the drug capecitabine was assessed by monitoring liver, heart, and lung organoids. As shown in Fig. 6(f), the metabolic breakdown of capecitabine by the liver organoids increases toxicity in heart and lung organoids compared to systems without a liver organoid.⁴⁶ The platform was also expanded to include six organoid types (liver, heart, lungs, blood vessels, testes, and brain) to evaluate ifosfamide toxicity. Similar outcomes were observed, with systems containing liver organoids exhibiting increased toxicity, suggesting the necessity of metabolizing tissues in multi-organ drug assessments. Importantly, as the number of organoids increases, tailored culture media and growth protocols are often required to sustain the entire system. The complexity of body-on-a-chip designs is crucial for maintaining both safety and accuracy in drug testing, offering a pathway to reduce reliance on animal models.

Overall, microfluidic technologies enhance organoid models by enabling reproducibility, deeper perfusion, and more physiologically relevant interactions. Leveraging these designs also allows integration of diverse monitoring modalities, including imaging, genetic profiling, and sensor-based readouts.

III. ORGANOID SENSOR INTERFACES

In parallel with highly tunable microfluidic systems, a wide range of sensor technologies have been developed to track organoid development and function. Sensors integrated into developmental or drug-response models must exhibit minimal interference with organoid growth, maintain stable performance under culture conditions, and provide sufficient spatial resolution to capture structural and functional complexity. Desired designs include non-destructive and dynamic monitoring of the same organoid over extended culture periods, often lasting several months, without restricting organoid development.¹⁸ While multiple sensing targets have been explored, most studies focus on either physical properties or chemical markers related to metabolism and signaling.^48,49 Incorporating features such as microscale dimensions or mechanical flexibility allows sensors to interface with organoids while minimizing structural damage and developmental disruption.

A. Physical sensors

Physical sensors measure characteristics of the culture environment or the organoid itself, such as temperature, pressure, electrical signals, or displacement caused by organoid growth. These sensors can be positioned on the culture vessel,^50,51 embedded inside the organoid,^52–54 or wrapped around the organoid surface.^55–58 They are particularly suited for organoid types with intrinsic electrophysiological activity, such as brain^59–61 and heart organoids.^50,62,63 Because electrophysiological properties can be affected by pharmaceutical agents, drug response is a common validation method for such sensors, where firing rate or beating rate may have a proportional response to drug concentration.

The Rogers group developed a flexible multielectrode array [Fig. 7(a)] that enables spatial monitoring of electrophysiological signals in co-cultured cortical and astrocyte spheroids. This device uses 25 low-impedance electrodes to map the field potential across the organoid surface.⁵⁵ Its adaptable and basket-like shape is formed by compressive buckling of photolithographically patterned 2D gold electrodes on a pre-strained PDMS substrate as an assembly platform [Fig. 7(b)]. Using this basket, spatial mapping of electrophysiological activity revealed firing rate reductions following administration of tetrodotoxin (TTX), a sodium channel blocker [Fig. 7(c)]. These results underscore the importance of developing monitoring methods capable of capturing spatial heterogeneity across the spheroid structure, rather than relying on single-point or planar electrodes. In the context of assembloids, such methods are particularly valuable for revealing electrical connections across distinct brain regions.

FIG. 7.

Physical sensing methods for organoid analysis. Multifunctional wrap: (a) Multifunctional mesoscale framework (MMF) for spatial recording of electrophysiological activity on cortical and astrocyte co-culture neural spheroids, integrating microelectrodes with thermal, optical, and electrochemical sensors. (b) Design of the electrode array and holding structure around the pre-formed spheroid.⁵⁵ (c) Multichannel electrophysiological firing recordings around the spheroid before and after exposure to tetrodotoxin (TTX).⁵⁵ Wrapping basket: (d) Experimental overview of differentiated cortical organoids (hCO) and striatal organoids (hStrO) analyzed individually or as assembloids via electrodes. (e) Kirigami-inspired 3D electronic structure design with microelectrodes to collect electrophysiological measurements on neural organoids.⁵⁶ (f) Comparison of firing rates during hCO differentiation, as well as at 138 day before and after exposure to 4-AP, a known stimulant.⁵⁶ Internal mesh: (g) Process of cardiac organoid formation around a flexible mesh structure in vitro. (h) Image of mesh embedded within a cardiac organoid after 40 days of differentiation.⁵⁴ (i) Beating amplitude over time during cardiac organoid differentiation.⁵⁴ Diaphragm stretching: (j) Side view demonstration of a PDMS holder with tip and Pt sensing diaphragm for remote monitoring of cardiac beating activity. (k) Device image of an organoid on a cracked Pt diaphragm.⁶³ (l) Electrophysiological (black) and mechanical (red) recordings of beating activity with increasing carbachol concentration.⁶³ Parts (a), (d), (g), and (k) were created in BioRender.com. Parts (b) and (c) reproduced with permission from Park et al., Sci. Adv. 7, eabf9153 (2021). Copyright 2021 The American Association for the Advancement of Science. Parts (e) and (f) reproduced with permission from Yang et al., Nat. Biotechnol. 42, 1836 (2024). Copyright 2024 Nature Publishing Group. Parts (h) and (i) reproduced with permission from Li et al., Nano Lett. 19, 5781 (2019). Copyright 2019 American Chemical Society. Parts (k) and (l) reproduced with permission from Lyu et al., Nat. Commun. 13, 7259 (2022). Copyright 2022 Nature Publishing Group.

A wrapping array of a similar concept was designed using a kirigami-inspired buckling approach, creating a thinner suspended basket shape to monitor either single organoids or assembloids as a part of the growth vessel [Fig. 7(d)]. This array employs a spiral design of platinum and gold patterned between two layers of SU-8 followed by etching of a sacrificial layer on the silicon wafer to release the basket structure [Fig. 7(e)]. The device contains 32 impedance electrodes that can simultaneously record local signals from suspended neural organoids at a sampling rate of 20 kHz, enabling stable long-term monitoring over 179 days of culture.⁵⁶ The high density of contact points enhances spatial resolution, while the flexible design provides a high level of temporal resolution. Drug validation confirmed the increase in the firing rates of the cortical organoids after administration of 4-AP, consistent with expected pharmacological effects [Fig. 7(f)]. The system also demonstrated responsiveness to optogenetic stimulation. Importantly, its thin, net-like architecture allowed visualization of suspended organoids without interference, making it compatible with the traditional evaluation methods such as bright-field imaging and immunofluorescent imaging.

Together, these advances demonstrate that thin, flexible electronic suspension systems can integrate seamlessly with organoid growth and function. By offering minimally invasive, high-resolution, and long-term monitoring, they represent a significant step toward more physiologically relevant readouts in organoid research. However, it should also be noted that a common challenge with wrapping designs is the surface-level nature of the measurements, preventing the probing of the 3D volumetric space of the organoid for comprehensive evaluation. Moreover, functional electrophysiological evaluations generally need to be used in conjunction with imaging methods such as fluorescence microscopy of tagged proteins to obtain phenotypical information. Additionally, there is a limitation in the number of samples that can be evaluated due to the limited system scalability and relatively large footprint of the electronic suspension supports.

Beyond external electrode systems, several internal physical sensor designs have been developed for integration within spheroid structures. Examples include internalized RF antennas for remote signal transmission^59,65 and pointed electrode arrays that partially pierce the spheroid to attach it to external measurement devices.^66,67 One notable approach employed a flexible microelectrode array sheet embedded within cardiac organoids during growth [Figs. 7(g) and 7(h)] to monitor beating throughout different developmental stages.⁵⁴ These electrodes were fabricated by depositing platinum and poly(3,4-ethylenedioxythiophene) (PEDOT) on a stretched substrate, forming a serpentine structure that buckles upon release of the pre-strain in the substrate.⁵⁴ The mesh was connected to an external system for power supply and data acquisition to record signals. This design allowed the continuous growth of the formed “cyborg” organoids over 35 days while capturing electrophysiological signal amplitude and frequency as the cardiac organoid matured [Fig. 7(i)]. Similar mesh electrodes from the same group were adapted with graphene barcoding to spatially corroborated RNA sequence,⁵³ further enabling a deeper understanding of the formation, characterization, and integration of these mesh electrodes.⁵² Similar to suspension meshes, these systems maintain optical transparency, allowing imaging through the organoid structure. The successful incorporation of spatially distributed internal electrodes highlights the potential for minimally disruptive monitoring of electrophysiological and optical properties in growing organoids.

Beyond the measurement of electrophysiological signals, mechanical forces are critical for the development and function of some organoids and are targeted by evaluation methods such as bulk rheology.⁶⁸ However, bulk rheology typically focuses on environmental properties and quantification in individual organoids occurs using techniques such as optimal tweezers, parallel plate potentiometry, laser ablation, or atomic force microscopy (AFM).^69,70 An integrated approach combining mechanical and electronic sensing is demonstrated in Fig. 7(k), where a PDMS tip coupled with a platinum-coated sensing diaphragm is used to track cardiac organoid beating. The culture vessel shown in Fig. 6(l) simultaneously collected both electrophysiological signals and mechanical beating data, demonstrating that treatment with carbachol caused a reduction in beating frequency, which can be monitored remotely through both modalities [Fig. 7(j)]. Similarly, a microneedle array coupled with a deformable diaphragm has been used to the ability to simultaneously track both regular and interrupted beating patterns using a real-time monitoring system.⁶⁴ Please note that complex mapping and convolution are often required to extract true mechanical measurements and provide adequate spatial resolution. Nevertheless, electro-mechanical coupling systems already have broad applications in drug screening, including recent examples wherein electrical and mechanical sensors are combined to evaluate the effect of DOX on a cardiac organoid beating pattern.⁶⁴ A summary of representative electrical and mechanical systems with electrophysiological targets can be found in Table I. These systems not only indicate when beating becomes prominent in cardiac organoid development but also reveal alterations to heart rhythm after exposure to common chemotherapeutic treatments, closely recapitulating arrhythmias observed in patients.

TABLE I.

Comparative summary of representative electrical organoid sensors.

Device type	Organoid type	Key metric	Sensing modality	Measurement resolution	Measurement duration	Source
Soft sensing diaphragm	Cardiac organoids	Mechanical beating and electrophysiological activity	Deformation	10 ms	5 h	63
Wrapping electrodes	Neural organoid	Electrical activity	Local field potential	1 ms	1 min	55
Penetrating electrode	Neural organoid	Electrical activity	Local field potential	1 ms	40 min	66
Penetrating electrode	Cardiac organoids	Electrophysiological activity	Electrocardiogram	50 ms	30 s	67
Penetrating electrode array and soft sensing diaphragm	Cardiac organoid	Electrophysiological and mechanical beating	Deformation and local field potential	10 ms	75 min	64

B. Chemical sensors

Chemical sensors are used to monitor molecular changes either in the surrounding environment or within the organoids themselves, tracking parameters such as pH, oxygen concentration, and biomarker levels. Monitoring these factors offers analytical insights into the culture process for both research and industrial applications. While in industrial settings, culture monitoring is typically performed by optical sensors or spectroscopic analysis,⁷¹ research applications often require more detailed analysis, given the high sensitivity of differentiated organoid models to growth factors and environmental conditions.

Many on-chip microfluidic systems can employ in-line chemical monitoring, where a portion of culture media is isolated for analysis either in real time or through periodic sampling. In-line chemical sensors often focus on detecting biomarkers that can be quantified using existing cellular assays, such as oxygen, pH, or specific secretions linked to the organ of interest.^72–74 For example, hepatic organoids are monitored for transferrin and albumin, both key proteins synthesized by mature liver tissue,^47,75 while neural organoid maturity can be assessed through neurotransmitters including glutamate or dopamine.^61,76 Another example of an organ-specific biosensor target is creatine kinase (CK-MB), an isoenzyme linked to cardiac cell damage, which can be monitored using an aptamer-based sensor [Fig. 8(a)]. In this design, media from the culture chamber was diverted to the sensor through separate microfluidic connections [Fig. 8(b)], enabling measurement of CK-MB concentration variations in response to doxorubicin dosage and exposure time [Fig. 8(c)].⁷⁷ These in-line sensors can reduce the need to terminate cultures for end point analysis, such as fluorescent assays, immunohistological imaging, or flow cytometry.⁷⁸ However, in-line sensing is a representation of the entire organoid environment, rather than capturing spatial variations around an individual organoid, and additionally, biorecognition elements such as aptamers often have limited stability over time. Despite these limitations, combining in-line chemical monitoring with other physical readouts, such as temperature, provides a more comprehensive view of growth conditions, enabling precise control and understanding of organoid development.

FIG. 8.

Chemical sensing strategies for organoids. Bulk solution sensing: (a) Aptamer-based electrochemical sensor for detecting CK-MB in damaged cardiac organoids. (b) Integration of the sensor with organoid culture chamber on a PDMS chip.⁷⁷ (c) Measurement of cardiac organoid damage with increasing exposure of doxorubicin.⁷⁷ Penetrating electrode: (d) The functionalization process of a penetrating electrode for sensing glutamate in forebrain organoids. (e) Image of the electrode inserted into neural organoid tissue.⁶¹ (f) Differences in electrical current and associated glutamate levels between dorsal and ventral forebrain organoids at different lengths of culture time.⁶¹ Parts (a) and (d) created in BioRender.com. Parts (b) and (c) reproduced with permission from Shin et al., Anal. Chem. 88, 10019 (2016). Copyright 2016 American Chemical Society. Parts (e) and (f) reproduced with permission from Nasr et al., Biosensors 8, 8010014 (2018). Copyright 2018 Author(s), licensed under a Creative Commons Attribution (CC BY) license.

Sensing chemical concentrations within individual organoids, rather than in the bulk solution, requires localized spatial electrode designs including permanently implanted sensors⁷⁹ or temporary microelectrode punctures.⁶¹ For example, Nasr et al. employed a temporary penetrating electrode to monitor glutamate levels in neural organoids modeling distinct brain regions. As shown in Figs. 8(d) and 8(e), a glass pulled tip was coated with Cr/Pt, followed by drop casting of the enzyme glutamate oxidase (GluOx) to selectively detect glutamate. The electrode tip was periodically inserted into brain organoids using a micromanipulator, enabling electrical monitoring of regional glutamate dynamics.⁶¹ Measurements over the course of differentiation revealed higher levels of glutamate in dorsal forebrain organoids compared to ventral forebrain organoids, consistent with in vivo brain architecture [Fig. 8(f)].⁶¹ Longitudinal tracking of biochemical signals in individual organoids can generate data-driven insights into their size, physical properties, and evolving metabolic states.

In contrast to implanted approaches, chemical sensors that detect changes in the surrounding environment of organoids are far more common, largely due to concerns with the long-term stability and functionality. These sensors rely on biorecognition elements, such as enzymes or aptamers, to generate an electrical or optical signal in response to changes in pH or molecular concentrations. However, both the selectivity of the biorecognition element and the long-term stability of the sensor performance can be compromised in biologically complex environments. For example, biofouling, the adsorption of non-specific proteins to the sensor surface, can cause signal shift over time, which is magnified in the case of direct contact between cells and electrodes.⁸⁰ In organoid cultures, such contact can be the result of undirected cell growth onto electrodes or cell shedding during long-term culture, altering the local chemical environment. Signal drift can be partially mitigated through periodic baseline measurements combined with cleaning and sterilization procedures adapted from organ-on-a-chip platforms.⁸¹ However, the need for frequent baseline correction limits true real-time measurements by introducing multi-step data processing requirements. Despite these challenges, there remains a need for advanced chemical sensing strategies that monitor growth process in real time over longer periods with stable and specific progress mapping. Non-destructive, temporal monitoring of chemical signals from the same organoid is essential for reproducible measurements, system standardization, and sensor-based validation of biological models. For example, despite these stability challenges, real-time sensing has been demonstrated in breast cancer stem cells differentiating into spheroids, where lactate and oxygen metabolism are monitored continuously throughout growth.⁸² As chemical sensing techniques continue to mature, addressing challenges related to stability, specificity, and long-term signal drift will be essential for enabling non-destructive, longitudinal monitoring requirements of organoid research and translational applications.

IV. RECENT ORGANOID APPLICATIONS

Organoid models are highly valuable for their ability to recapitulate physiologically relevant structures and responses, especially in testing pharmaceutical agents that can facilitate cellular growth or model diseases. The 3D organization of organoids more accurately reflects in vivo behavior compared to traditional cell layer models due to enhanced cell-to-cell interactions and the presence of extracellular matrix. Here, we highlight representative recent applications that leverage advanced microfluidics or microelectronics systems. As illustrated in Fig. 9, these applications generally fall into three main categories: drug delivery, personalized medicine, and developmental studies.

FIG. 9.

Notable recent application examples in drug delivery, personalized medicine, and developmental biology. Drug delivery: (a) Schematic demonstrating induction of steatosis in liver organoids cultured in SteatoChip device using flufenamic acid (FFA) and subsequent drug testing. (b) Device layout to generate native flow patterns to liver organoids and example fluorescent stain to compare overall volume properties with proportional lipid quantity.⁸⁴ (c) Evaluation of metformin, pioglitazone, and OCA effect on reducing volume of lipid when compared to the Flow + FFA in vitro steatosis effect organoids.⁸⁴ Personalized medicine: (d) Experimental flow of generation of patient-derived organoids (PDOs) and utilization of ATP optical sensing for testing of chemotherapeutic treatments for tumors. (e) Device layout for optimized culture of multiple PDO wells and optical sensor design using activated fluorescent Cy3 upon ATP binding.⁹⁷ (f) Comparison of chemotherapeutic compound combinations for two patients, showing different optimal treatments.⁹⁷ Developmental study: (g) Summary of study for applying nicotine to developing neural organoids in a flow system to simulate nicotine consumption during pregnancy. (h) Device layout showing directional flow of nutrients, nicotine, and removal of metabolic waste products.¹⁰⁸ (i) Fluorescent imaging of neuronal outgrowth from formed organoids over 5 days of exposure at different concentrations of nicotine. Scale bars: 50 μm.¹⁰⁸ Parts (a), (d), and (g) created in BioRender.com. Parts (b) and (c) reproduced with permission from Teng et al., Biomaterials 275, 120904 (2021). Copyright 2021 Elsevier. Parts (e) and (f) reproduced with permission from Zhang et al., Anal. Chem. 96, 10092 (2024). Copyright 2024 American Chemical Society. Parts (h) and (i) reproduced with permission from Wang et al., Lab Chip. 18, 851 (2018). Copyright 2018 Royal Society of Chemistry.

A. Drug screening

Effective drug screening relies on identifying appropriate dosages and exposure times to safely move toward clinical trials. In the pharmaceutical industry, there remains a critical need for predictive models to reduce the expenditure of resources on drugs that perform well in vitro or animal studies but fail in human clinical trials with a failure rate as high as 90%, with costs of $1–2 billion per drug brought to market.⁸³ Compared to animal models, human organoids offer an ethical and more predictive platform for testing drug efficacy in human patients. For example, Teng et al. developed a microwell device known as SteatoChip, designed to simulate steatosis, a disease that causes excessive fat accumulation in the liver.⁸⁴ The condition of steatosis was reproduced in hepatic organoids by adding flufenamic acid (FFA), followed by the examination of the efficacy of several drugs [Fig. 9(a)]. By employing fluorescent imaging to compare the lipid volume with the overall volume of the organoid [Fig. 9(b)], the performance of metformin, pioglitazone, and obeticholic acid (OCA) in reducing lipid concentration was demonstrated [Fig. 9(c)]. The study found that metformin most effectively reduced steatosis in liver organoids, consistent with its known therapeutic effectiveness in the human body. The high-throughput chip format not only simulated perfusion conditions in the liver but also enabled the simultaneous culture of large numbers of spheroids, providing a rapid and robust platform for drug testing.

Additional studies comparing 2D and 3D models of various organs, including the breast,⁸⁵ colon,⁸⁶ and liver,⁸⁷ have shown that 3D models yield more accurate predictions of drug responses and patient survival. Among available platforms, microwell arrays are particularly effective for drug screening, as they enable the controlled generation of large numbers of spheroids and allow for precise exposure to drug concentrations. Drug evaluation typically involves assessing organoid size, viability, and specificity to the target organ, and often incorporates targeted assays for markers of cell death or metabolic activity. In this context, electrochemical sensors have emerged as powerful tools for real-time monitoring. Rapid formation and timely evaluation of organoids are essential to identifying effective treatment within clinically relevant timeframes, both for general drug screening and for personalized medicine applications.

B. Personalized medicine

Capturing the drug response of organoids is highly valuable not only for evaluating the therapeutic potential but also for predicting patient-specific outcomes. Patient-derived organoids represent a significant step toward personalized medicine, especially for optimizing the efficacy of chemotherapy treatments for cancer patients.^88,89 Patient-derived organoid models have the benefit of reproducing tumor properties including protein expression or structural variations, enabling a detailed examination and simulation of patient responses to chemotherapeutics or combination therapies. These models have proven effectiveness in modeling cancers of the liver,^90,91 pancreas,^92–94 breast,⁹⁵ and glioblastoma,⁹⁶ among others. However, there remains a critical need to develop faster and more reproducible culture systems to shorten the time from patient biopsy to robust analysis for clinical decision-making and, ultimately, to the application of the therapeutic agent back to the individual patient.

A recent study by Zhang et al. demonstrated the feasibility of rapidly generated, personalized organoids to monitor patient-specific metabolic changes due to combination chemotherapy using an optical adenosine triphosphate (ATP) sensor [Fig. 9(d)].⁹⁷ The sensor incorporated Cy3, which fluoresces upon binding to ATP-targeted aptamer strands, producing an intensity signal proportional to ATP presence. The sensor molecules were injected into the microfluidic device that enabled both culturing and imaging organoids in the same integrated platform [Fig. 9(e)]. The results showed a reduction in metabolism across all treatments, with the combination 1P3C being more effective for patient 1, while 3P1C was more effective for patient 2 [Fig. 9(f)]. This study highlights the potential of personalized treatment modeling within 15 days of biopsy sampling, a timeframe that aligns with clinical intervention needs, by combining microfluidics with optical sensing for streamlined monitoring.

C. Developmental studies

Fundamental questions remain about organ development during fetal growth and the cellular processes underlying tissue regeneration and remodeling in the adult body. In vitro biological models have been used to identify remodeling processes, such as the hematopoietic function of bone marrow⁹⁸ and immunological relationships, such as the role of macrophages in the development of fetal brain tissue.⁹⁹ Organoid models derived from differentiated stem cells offer a window into organogenesis, enabling the study of the formation of substructures that are critical to organoid function. Notable examples include the formation of wrinkling patterns in brain organoids¹⁰⁰ and the formation of islet cell groups in pancreatic organoids.¹⁰¹ As illustrated in Fig. 9(g), a microfluidic platform was developed to investigate the impact of nicotine on neural organoid development over five days. In this flow chip system, nicotine and nutrients were perfused into preformed embryoid bodies (EBs), while metabolic waste was simultaneously removed [Fig. 9(h)]. Subsequent fluorescent imaging revealed reduced neural outgrowths from nicotine-exposed organoids, indicating that even short-term exposure to high concentrations of nicotine is highly disruptive to neural development [Fig. 9(i)]. The in vitro visualization of the developmental disruption provides a tangible and ethical approach to studying the effects of nicotine exposure during gestation.

A key consideration in designing organ development models is achieving long-term culture and sufficient perfusion to deeper layers of the organoid to support maturation. Longevity is particularly important, as human organ development spans at least nine months, and tissue remodeling occurs throughout the human lifespan. Generating suitable microfluidics to be incorporated within organoids remains challenging, especially in creating appropriately sized structures for synthetic vascularization to ensure effective perfusion. Organoid development is typically assessed by evaluating the identity, structure, and function of the organoid, often using immunohistology of relevant proteins. However, these methods are disruptive, requiring fixation or sectioning of organoids for imaging. Sensor-based alternatives that target biomarkers offer a promising noninvasive approach to monitor developmental markers. Current organoid models are currently limited to a few months of growth, halting tissue maturation at an early stage. Future research will need to focus on developing long-term, noninvasive culture and monitoring methods to more effectively recapitulate organ development.

V. FUTURE CONSIDERATIONS

While significant advances have improved temporal and spatial control in organoid engineering, current technologies still cannot fully replicate the phenomena observed in human tissues. A major limitation is the inability to grow and evaluate organoids at levels of maturity and complexity that accurately reflect human biology. Brain development, for example, unfolds over decades, and much remains to be discovered about ways to quantify this progression in terms of both structural architecture and neural connectivity.¹⁰² Model longevity is particularly critical for studying neurodegenerative diseases such as Alzheimer’s, dementia, and general neurodegeneration—conditions that manifest after prolonged periods of brain development and remodeling. Longevity has been approached biologically utilizing clustered regularly interspaced short palindromic repeats (CRISPRs) for genetic reprogramming¹⁰³ and through model regeneration using organoid passaging and expansion.¹⁰⁴ Achieving culture longevity will require engineered systems that provide stable long-term perfusion and sensing, alongside careful selection of biological models, fabrication methods, and material systems capable of maintaining micrometer-scale spatial resolution. Integrating these biological model advances with microfluidic and microelectronic systems may further enable organoids to reach higher levels of structural and functional complexity. For example, future microfluidics platforms may incorporate spatially controlled delivery of growth factors to generate complex organoid phenotypes, while automated control systems could streamline handling and differentiation. In parallel, the development of multimodal sensors capable of probing the 3D microenvironment may improve long-term organoid stability, particularly when coupled with machine learning approaches to process large data sets and correlate sensor outputs with disease-relevant states.

As organoid platforms become increasingly instrumented, the growing volume and complexity of sensor-derived data introduce significant analytical challenges. In particular, correlating functional measurements with data from established analytical techniques, such as polymerase chain reaction (PCR) and imaging, is necessary to identify structural, molecular, or genetic landmarks. Recent work has demonstrated the integration of transcriptomic data with spatially mapped electrophysiological measurements to define niche states within organoids and uncover key functional-spatial correlations.⁵³ However, such approaches often require extensive data processing, limiting their accessibility and scalability in healthcare settings. Furthermore, for clinical translation, reproducible performance under standardized organoid culture protocols, together with simplified analytical workflows, will be essential for integration within existing medical infrastructure. In this context, automation of data handling and analysis will be critical to accommodate the number of patients and biological replicates necessary for clinical confidence. To date, artificial intelligence (AI)-based approaches have been applied to the analysis of organoid morphology, histology, and omics data to identify cell lineages, classify phenotypes, and predict developmental trajectories.^105,106 Extending these approaches to bridge corresponding structural and functional evaluations is an important next step toward healthcare-oriented organoid platforms.

Looking ahead, interdisciplinary efforts that leverage materials science, micromanufacturing, and biomedical engineering will be crucial in advancing healthcare understanding and treatment options using organoids. Emerging material systems, including stimuli-responsive materials that can change their shapes or properties in response to external cues, such as temperature, pH, light, or electrical and magnetic fields, offer new opportunities for controlled mechanical actuation.¹⁰⁷ Together, these advances hold promise for improving predictive drug screening, enhancing disease modeling, and accelerating the translation of organoid-based platforms toward personalized medicine and therapeutic applications.

NOMENCLATURE

4-AP

4-aminopyridine

AFM

Atomic force microscopy

AI

Artificial intelligence

ATP

Adenosine triphosphate

CK-MB

Creatine kinase-MB

Cr

Chromium

CRISPR

Clustered regularly interspaced short palindromic repeats

DOX

Doxorubicin

EB

Embryoid bodies

ECM

Extracellular matrix

FDA

Food and Drug Administration

FFA

Flufenamic acid

GluOx

Glutamate oxidase

HIF1-α

Hypoxia-inducible factor 1-alpha

HUVEC

Human umbilical vein endothelial cell

iPSC

Induced pluripotent stem cell

MTT

3-(4,5-di methyl thiazol-2-yl)-2,5-diphenyltetrazolium bromide

OCA

Obeticholic acid

PDMS

Polydimethylsiloxane

PEDOT

Poly(3,4-ethylenedioxythiophene)

PEGDA

Poly(ethylene glycol) diacrylate

PMMA

Polymethyl methacrylate

Pt

Platinum

RF

Radio frequency

RNA

Ribonucleic acid

TGF-β

Transforming growth factor-beta

TTX

Tetrodotoxin

Wnt

Wingless and Int-1 signaling pathway

DATA AVAILABILITY

Data sharing is not applicable to this article as no new data were created or analyzed in this study.