Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2026 Aug 18.
Published in final edited form as: IEEE J Sel Top Quantum Electron. 2026 Mar 10;32(4):6900211. doi: 10.1109/JSTQE.2026.3672359

Automated Optofluidic Analysis of Single Extracellular Vesicles for Organoid Culture Monitoring

Ravipa Losakul 1, Kateryna Voitiuk 1,2, Ruiting Xu 1, Spencer T Seiler 2, Viktor Yurevych 2, Thomas D Yuzvinsky 1, Andy Wang 1, Ivana Pačar 1, David Haussler 2, Sofie R Salama 2, Mircea Teodorescu 1,2, Holger Schmidt 1
PMCID: PMC13479748  NIHMSID: NIHMS2159846  PMID: 42609747

Abstract

Advances in organoid-on-chip technology are transforming culturing methods by enabling precise control over biochemical environments and sustained nutrient delivery. Conventional methods for assessing organoid health are invasive, requiring skilled manual handling and sacrificing the tissue for analysis. We present a fully automated, integrated system for noninvasive extraction, labeling, and optical analysis of extracellular vesicles (EVs) secreted by 3D organoid cultures. EVs contain important molecular biomarker cargo, such as nucleic acids and proteins, enabling real-time insight on the health of their cell of origin. Using a polydimethylsiloxane (PDMS) optofluidic biosensor, we demonstrate: (1) optical detection of single EVs derived from organoid culture; (2) on-chip sample processing and counting of single EVs; and (3) automating conditioned media delivery, on-chip sample preparation, and optical detection of EVs in a single integrated system. Together, these results showcase a noninvasive approach to monitoring organoid health through EV signatures present in conditioned media. The integrated platform based on a modular optofluidic analysis chip further offers a foundation for integrating a variety of downstream analysis techniques, advancing continuous organoid-on-a-chip monitoring.

Index Terms—: biosensors, organoids, automation, biophotonics, optofluidics

I. INTRODUCTION

Rapid progress in organoid culture and monitoring have enabled three-dimensional, stem-cell derived models that mimic the architecture and function of organs, offering a path towards biomedical advances without animal models or human subjects. Organoid-on-chip platforms integrate organoid culture with microfluidic devices to provide a physiologically realistic in vitro model within a controllable environment. These platforms address the limitations of conventional cell culturing methods, which often induce variability in organoid growth and differentiation. By offering precise control over biochemical environments, continuous nutrient and oxygen supply, and media exchanges, as well as minimally invasive monitoring of organoids, organoid-on-chip platforms recreate the physical and biochemical conditions of living organs and lead to tissue longevity and reproducibility. [1–7].

A major challenge for operating such platforms is to extract as much information as possible from these developing tissues, including ultrasensitive assessment of their molecular biomarker content. Conventional methods for assessing culture health, such as quantitative polymerase chain reaction (qPCR) or immunoassays, offer satisfactory performance but cannot readily be integrated with the culture platform and require trained technicians for biological sample handling. Furthermore, these methods are either disruptive to organoid cultures or require terminating them altogether to access genetic information [8]. Lab-on-chip approaches to monitoring cell culture can be less disruptive to organoid growth and offer better integration possibilities. Multiple approaches have been developed, each with their own advantages and challenges. Optical methods, in particular fluorescence-based approaches, are well developed and popular for quantifying and analyzing biological particles and processes. For example, evanescent wave fluorescent biosensors require target biomolecules to bind both fluorophores and to an optical fiber functionalized with aptamer capture probes. This method allows for high sensitivity but exhibits limited penetration depth, since only the fluorophores closest to the fiber can be excited [9]. Other methods combine fluorescence-based measurements with surface plasmon resonance detection or surface-enhanced Raman spectroscopy to enhance fluorescence signals using a variety of different materials like streptavidin-coated magnetic microbeads [10], gold nanorods [11], single-walled carbon nanotubes [12], or modified Au/glass substrates coated with antibodies for plasmonic imaging of biomolecules [13]. Photonic crystal (PC) surfaces have been used to visualize individual gold nanoparticles (AuNPs) through resonance coupling to quantify nucleic acids and proteins [14–20]. Colorimetric detection methods rely on specific reagent reactions to target analytes to observe and measure color changes that are directly proportional to analyte concentration but traditionally require a multi-step process for analysis [10–21]. Recent advances that reduce sample preparation in colorimetric methods have increased complexity by introducing hydrophilic magnetic nanoparticles or antibodies conjugated with specific antibodies [22–24], various functionalized aptamers [25–31], or unique materials like quantum dot-embedded silica [32] and graphitic carbon nitride nanosheets [33] to reduce sample preparation. Electrochemical-based methods require immunoassay modifications [34] or micropatterned aptamers and hybridization chain reaction for detection amplification [35]. Lastly, resistive pulse sensing measures the electrical resistance of a given particle passing through a nanopore. It offers a label-free characterization of analytes but is susceptible to system instability from membrane strain from blocked pores and low signal-to-noise ratios due to particle size and pore diameter mismatch [36].

On the tissue culture side, integrated platforms for continuous culturing and monitoring of organoids have recently been introduced. Seiler et al [37] demonstrated a modular microfluidic cell culture platform that automated and simplified conventional cell culture methods and allowed users to continuously monitor organoid state, development, and overall health. Within an Internet of Things (IoT)-based infrastructure of benchtop modules [38], programmable microfluidics are remotely controlled and enable scheduled and precise media exchanges tailored to the organoid culture itself [39]. This minimized the need for highly trained technicians for biological sample handling, reduced potential contamination, and allowed for adaptive intervention if/when organoid health was compromised. With automated media exchanges, downstream analysis of conditioned media can be used to infer the health of organoids with repeatable, noninvasive measurements of the extracellular vesicles (EVs) they produce.

EVs are lipid-bound particles, including exosomes and micro-vesicles, that are used in intercellular communication and contain functional cargo such as proteins, lipids, and nucleic acids. These cell-specific biomarkers provide information on cellular processes and can be used to infer information about the health of its cell of origin [40–42]. As organoids consume nutrients and grow, EVs are released but are commonly discarded during media exchanges. Single EV analysis enables the characterization of individual EVs and uncover cell-specific genetic insights at a given point in time. The current state of cell development is reflected in the molecular cargo of EVs, where any genetic mutations or cell divergences can signal disturbances or distress in cell environment. For example, tumor-derived EVs from cancer cells express specific protein and miRNA mutations when detected with single EV analysis techniques and allows for immediate intervention and monitoring of disease progression [42–48]. By retaining EVs, they can be detected and quantified, providing information about organoid health in real time without disturbing, damaging, or destroying the organoid itself [49].

With integrated lab-on-chip biosensors, culturing platforms can continuously screen organoids grown in vitro with minimal disruption to their environment. Such screening could potentially reduce variability in size, structural organization, functional capacity, and gene expression, in addition to enabling long-term organoid growth [3]. Optofluidic platforms provide the ability to combine manipulation of small fluid volumes with reconfigurable photonics to achieve novel or improved capabilities in a chip-scale format [50–57]. Numerous examples of using optical waveguides, photonic crystals [58–62], and other photonic elements to enable biological particle detection on a variety of substrates have been reported [63–72]. Among these, implementing optofluidic biosensors in polydimethylsiloxane (PDMS) offers several advantages, including rapid prototyping, biocompatibility tailored to specific cell cultures, optical transparency, and mechanical elasticity for microvalve-based sample processing on the same chip [73–81].

Here, we introduce a fully automated integrated system that enables noninvasive extraction and optical analysis of individual EVs from a 3D tissue culture platform. Specifically, growth media containing EVs from a cerebral organoid culture are introduced into a PDMS-based biosensor in which nanoparticles can be labeled and detected with single-particle sensitivity. First, we demonstrate the optical counting of single EVs with a PDMS-based optofluidic biosensor. Second, we demonstrate the on-chip sample processing (fluorescent labeling) and counting of single EVs in a PDMS-based optofluidic biosensor. Third, we present an integrated system that automates both the delivery of conditioned media to our PDMS-based biosensor and the sample preparation and detection of single EVs.

II. METHODOLOGY

A. Device Design

A PDMS-based optofluidic device design containing both sample preparation and optical detection sections on a single analysis chip was used in this study as shown in Fig. 1a. The sample preparation section (automaton) is based on a previously developed device [73–75] and consists of 15 pneumatically actuated lifting-gate valves [76–80] that allow for programmable fluidic motion through a network of liquid channels within an enclosed device, as seen in Fig. 1a. The PDMS valves are lifted (opened) and lowered (closed) using negative and positive pressure, respectively, and allow for fluid to move between neighboring valves. Valve structures are 500 μm in diameter with approximately 5–7 μm channel height throughout the device. The lifting-gate valves are programmable to actuate in a predetermined sequence for on-chip sample preparation, as well as subsequent optical detection on the same device. The optical detection section, shown in Fig. 1b, consists of solid-core waveguides that perpendicularly intersect with a fluidic analyte channel. An external continuous wave (CW), 488 nm wavelength laser operating at 20 mW output power (Omicron LuxX, Model #488–60) was fiber-coupled orthogonally to the fluidic channel via an excitation waveguide and created an excitation region (red box).

Fig. 1.

Fig. 1.

Open in a new tab

PDMS Optofluidic biosensor. (a) Complete optofluidic analysis chip with integrated on-chip sample preparation (automaton) and optical detection regions. The sample preparation section (red) consists of 15 pneumatically actuated valves with a fluidic channel network connected with 6 inlet ports. The optical detection region (blue) has a fiber-coupled laser with a 488nm wavelength. Fluorescence is detected off-chip via collection waveguide, using a 60x lens with 0.85 numerical aperture (Newport Corporation, Model#M-60x, Irvine, CA), 532 nm long pass filter (Semrock, Model #BLP01-532R-25), and avalanche photodetector (APD, Excelitas Technologies). (b) Close-up of optical detection section that consists of two sets of PDMS-based solid-core waveguides that intersect with an orthogonal analyte channel. Laser is fiber-coupled via excitation waveguide and created an excitation region (red box) at its intersection with the analyte channel.

EVs were fluorescently labeled with SYBR Gold Nucleic Acid Gel Stain (ThermoFisher, Waltham, MA, USA), which targets to all forms of nucleic acids, has an excitation wavelength of 495 nm and emission wavelength of 537 nm. Fluorescently labeled EVs passing through the excitation region were excited and their fluorescence emission was collected off-chip via collection waveguide to an avalanche photodiode with a dark count rate of 100 counts/second and dead time of 24 ns (APD, Excelitas Technologies). A 532 nm long pass edge filter (Semrock, Model #BLP01–532R-25) was used block excitation wavelengths and capture the fluorescence emission of labeled particles. An image of an actual fiber-coupled optofluidic analysis chip used for measurements is shown in Fig. 2a. The dimensions of the excitation region was determined with an imaged optical mode at the edge of the excitation waveguide, shown in Fig. 2b. The width of mode image was measured (red line) to create an intensity profile (above) to determine the FWHM in pixels and converted to microns according to camera specifications (Thorlabs, Model #DCC1645C-HQ). The dimensions of the excitation region are reported with results. A top-down view of the excitation region (blue spot) in the Automaton is shown in Fig. 2c and is illuminated with a fiber-coupled laser via excitation waveguide.

Fig. 2.

Fig. 2.

Open in a new tab

(a) Illuminated optofluidic analysis chip under test with fiber-coupled laser. (b) Image of the optical mode of excitation waveguide transmitted through chip. The red line is the measured mode width and the top graph is the intensity profile used to determine the FWHM of the mode, measured in pixels. The pixel values are then converted according to the camera specifications and used to determine the excitation region. (c) Top-down view of the excitation region (blue spot) located at the intersection of the excitation waveguide and liquid channel and is illuminated with a 488 nm laser via excitation waveguide.

B. Device Fabrication

The optofluidic device was fabricated using a soft lithography process involving three layers. A detailed flowchart of all fabrication steps is included in the supplementary material. Briefly, a silicon master wafer consisting of fluidic channels and optical waveguides (Waveguide layer) was fabricated using negative photoresist SU-8 2005 (Kayaku, Round Rock, TX, USA). Another silicon master consisting of pneumatic valves and lines (Pneumatic layer) was fabricated using SU-8 2025 to produce taller features. A third blank wafer was used as a capping layer. Optical waveguides were made in the waveguide layer by capitalizing on total internal refraction and creating the refractive index difference by using two different base-to-curing agent ratios when mixing PDMS (Sylgard 184).

First, the waveguide core (refractive index 1.421) was formed with a 5:1 PDMS mixture spun onto the waveguide master at 5000 RPM for 15 min and then cured in a 60°C oven for 2 hours. Second, the waveguide cladding (refractive index 1.419) is formed using a 10:1 PDMS mixture spun on at 1,000 RPM for 5 min and then cured overnight [62]. Simultaneously, the pneumatic layer consisted of the lifting gate valves for sample preparation and was formed with the 10:1 PDMS poured over the pneumatic master as well as a blank master. All three layers were placed in a 60°C oven and cured overnight. After curing, the pneumatic layer was removed from the master and treated with oxygen plasma with the waveguide layer to activate their surfaces. With a mask aligner (OAI, Milpitas, CA, USA), the layers were aligned with the pneumatic valves placed over their appropriate fluidic channels, bonded via direct contact, and then cured overnight in a 60°C oven to strengthen the bonding. Afterwards, the waveguide master was removed, and 1.5 mm and 1 mm hole punches were made to create liquid and pneumatic ports, respectively. For final bonding, the exposed waveguide-pneumatic layer and blank layer surfaces were activated using oxygen plasma, bonded together to complete the three-layer device, and then placed on an acrylic mold that applied negative pressure to lift the pneumatic valves to prevent localized bonding.

C. Experimental Setup

Optofluidic devices were fixed on a removable glass slide stage and placed in a preset optical path towards a single photon detector (Excelitas Technologies, SPCM-AQR-14-FC) shown in Fig. 3. Their solid-core collection waveguides were aligned on-axis to a 60x collection lens with a 0.85 numerical aperture (NA) (Newport Corporation, Irvine, CA) by adjusting the device stage position and focusing the lens on the waveguide. A 532 nm long-pass fluorescence filter (Semrock Inc, West Henrietta, NY) and iris (Thorlabs, Newton, NJ) were placed along the optical path to filter out laser emission. A 40x objective lens with 0.65 NA (Newport Corporation, Irvine, CA) coupled to a multimode fiber (Thorlabs, Newton, NJ) was directly connected to the APD. When collecting data, a custom-made program with 0.1 ms time resolution captures and collects fluorescence signals in the analyte channel of the device. The time resolution is fast enough to capture individual fluorescence events in the excitation region.

Fig. 3.

Fig. 3.

Open in a new tab

Experimental setup demonstrating the optical path of a coupled analysis chip. A fiber-coupled to a 488 nm wavelength laser is aligned with the excitation waveguide of the analysis chip. Simultaneously, the optical path towards the avalanche photodetector (APD) begins with the collection waveguide coupled into a fixed 60x, 0.85 NA objective lens. The fluorescence from the excitation region then travels through a fixed iris and 532 nm long pass filter to eliminate environmental noise. Then it is collected with another 40x, 0.65 NA objective lens that is coupled with a multimode fiber connected directly to the APD.

D. Bioassay Sample Preparation and Detection

Analytes used in our experiments consisted of commercial lyophilized exosomes derived from human urine (Creative Biolabs, Shirley, New York, USA) and conditioned media derived from human dorsal cortical cerebral organoids at week 9 and week 18. SYBR Gold Nucleic Acid Gel Stain (ThermoFisher, Waltham, MA, USA) was diluted to a 10x concentration and used to validate nonspecific labeling and detection of EVs with a peak excitation of 496 nm and peak emission of 541 nm.

Week 9 organoids were grown in phenol-red free media, allowing the use of SYBR Gold stain without fluorescence interferences from phenol-red, where phenol-red is a pH indicator dye commonly used in cell culture. Due to the absence of phenol-red in the conditioned media and the availability of syringe filters, Week 9 conditioned media was prefiltered using a 0.22 μm Millex syringe filter (MillieporeSigma, Burlington, MA, USA). The culturing conditions for Week 9 included: 500 mL Brain-Phys media, 10 mL glucose supplement, 3.41 mL of 16 mg Geltrex extracellular matrix (ECM), 5 mL of 100x N-2 supplement, 5 mL lipid concentrate, 5 ml penicillin-streptomycin (Pen-Strep), 10 mL of 50x B-27 supplement, 5 mL of 250 μg/mL Amphortecirin B (Fungizone), 50 mL fetal bovine serum (FBS, Hyclone), and 50 μL heparin.

Week 18 organoids were grown in media containing phenol-red dye and without adequate filtration, would interfere with fluorescence measurements when staining EVs with SYBR Gold stain. Therefore, week 18 conditioned media was prefiltered to isolate EVs and remove phenol-red dye using a 70 nm qEV size exclusion column (SEC) from Izon Sciences (Christchurch, New Zealand) and TRIzol buffer. The culturing conditions for Week 18 included: 500 mL DMEM/F-12, 10 mL Embryonic Stem Cell (ESC) qualified FBS, 5 mL GlutaMax supplement, 5 mL N-2 supplement, 10 mL B-27 supplement, 5 mL lipid concentrate, 5 mL Pen-Strep, 5 mL Amphortecirin B (Fungizone), and 50 μL heparin with 0.12 mg/ml Geltrex ECM.

III. Results

A. Optical Counting of Single EVs

We first demonstrated the optical counting of single EVs in our PDMS-based devices via nonspecific staining of nucleic acids. EVs in conditioned media derived from organoids at week 9 of the cerebral cortex patterning protocol were prepared off-chip. The conditioned media was filtered with a 0.20 μm syringe filter. SYBR Gold was used for nonspecific staining and diluted from a 10,000x stock concentration to a 10x concentration using 1x concentration Phosphate-Buffered Saline (PBS) buffer. Next, 2 μL of commercial EVs were mixed with 2 μL 10x SYBR Gold stain and 16 μL PBS buffer, protected from light with foil, and incubated on a rotary mixer for 20 minutes. The labeled EVs were then pipetted into the fluidic analyte channel inlet and pulled through the device using negative pressure. Results of these calibration tests are provided in the supplementary material.

Next, EVs (Creative Biolabs) derived from human urine were resuspended in PBS buffer, aliquoted with 1.0×108 particles/mL concentration, and labeled off-chip using the protocol described above. Depicted in Fig. 4a, the labeled EVs were pipetted into the fluidic analyte channel inlet 3, and valves 15, 14, 13, 10, 9, and 8 were opened with negative pressure applied to their respective pneumatic ports using the control box touchscreen. Negative pressure was then applied to the fluidic analyte channel outlet, and the off-chip, prelabeled EVs were pulled through the device.

Fig. 4.

Fig. 4.

Open in a new tab

Optical Counting of Single EVs. (a) Schematic of valve and liquid channel network of the PDMS analysis chip. Off-chip labeled EVs were pipetted directly into inlet 3 of the analysis chip and pulled through as depicted in green. (b) Time-domain fluorescence time trace where detected events are indicated with red peaks. (c) Intensity histogram of all detected events. (d) A single detected event with respect to the noise background. (e) Live concentration of fluorescence trace over time.

Three technical replicate measurements were taken to ensure results were reproducible and accurate. The time domain fluorescence trace of a selected measurement is shown in Fig. 4b, where the parallel cluster wavelet analysis (PCWA) algorithm reported in [82–84] was used to distinguish detected events from the background in the analyte channel. This program used a continuous Ricker wavelet transform to match the waveform of expected events detected via single-mode excitation, recording the intensity and Δt (time domain FWHM) of each event. The concentration of detected events was determined using the total number of events (N), the excitation volume calculated from imaging the propagating mode in the waveguide and width of the analyte channel (Vexc), the total length of the time trace (T), and the average FWHM of detected events (Δtavg). The excitation volume for the analysis chips with a 0.5 mm valve diameter was determined to be 1.77×10−10 mL (mode width x mode height x channel width = 3.81 μm x 3.09 μm x 15 μm). The detected concentration was determined to be 4.04×106 particles/mL, with 211 detected fluorescence events and an average FWHM of 1.02×10−4 s in a 300 s trace. Fig. 4c shows the corresponding intensity histogram with a similar agreement between expected concentration and device sensitivity. A single detected particle is shown in Fig. 4d and demonstrates the fluorescence intensity well above the noise background. Fig. 4e shows the live concentration of the experiment which is determined within approximately 50 seconds.

B. On-Chip Labeling and Counting – Manual GUI

We next demonstrated and optimized the on-chip sample preparation and optical detection capabilities in the optofluidic analysis chips via nonspecific staining and detection of commercial EVs. To optimize this procedure, a series of interchangeable on-chip labeling and incubation times was performed on an analysis chip that was pretreated with 0.01 mg/ml BSA + 0.05% Tween20 and washed with PBS buffer immediately afterwards. Commercial EVs derived from human urine were resuspended in PBS and SYBR gold was diluted from stock concentration to 1x concentration in PBS.

With two input ports for positive and negative pressure, a custom-made pneumatic control modules consists of a compute module (Raspberry Pi 4) that allows for IoT-based messaging between other IoT-based devices and is connected to a touchscreen and a digital I/O board (National Instruments) to control a series of 16 solenoid valves. The touchscreen has a custom-made graphical user interface that outlines the pneumatic valves of the sample preparation section of the analysis chip and acts as a direct switch for positive or negative pressure to the connected solenoid valves. When a valve is switched to the ‘ON’ position, negative pressure is applied to the valve. The solenoid valves are connected to 16 individual pneumatic lines that are inserted into the pneumatic ports of the chip. Device functions were operated manually using the control box touchscreen to open/close the pneumatic valves.

Shown in Fig. 5b, 10 μL of 1x SYBR Gold stain and 10 μL of 1 × 108 particles/mL commercial EVs were pipetted into inlet 3 (blue) and 6 (yellow), respectively. PBS buffer is pipetted into the remaining inlets 1, 2, 4, and 5 (beige). 1x SYBR Gold stain was pulled into the device with a vacuum by opening valves 15, 14, 13, 10, 9, and 8 for two minutes, while valves 15 and 14 left open afterwards. Then, 1 × 108 particles/ml commercial EVs were pulled into the device with vacuum by opening valves 5, 6, 7, 10, 9, and 8 for two minutes, with valves 5 and 6 left open afterwards. Once the device was primed with both commercial EVs and stain, valves 15 and 5 were closed to push analytes to valves 14 and 6, respectively, to consolidate the analytes into one valve and create pressure for fluid movement. To begin the labeling process, valves 13 and 7 were opened, valve 14 and 6 were closed and then valve 7 was opened in succession to introduce both analytes within valves 13, 10, and 7. The analytes were then labeled in the following sequence: valve 7 closed, valve 10 closed, valve 10 opened, valve 7 opened, valve 13 closed, valve 10 closed, valve 10 opened. This L-shape sequence was repeated for durations of 2, 3, 5, or 10 minutes. Once mixing was done, valves 13 and 7 were closed to push the mixed analytes into valve 10. Then valve 9 was opened to allow for the analytes to incubate in valves 10 and 9 for durations of 2, 3, 5, or 10 minutes. For post-incubation detection, valve 10 was closed to push labeled analytes into valve 9. In succession, valve 8 was opened, valve 9 was opened, and then valve 9 was closed. Afterwards, PBS was pushed through the devices by opening valves 1, 2, 13, 10, 9, 8, and then closing valves 1, 2, 13, 10, 9, 8 in succession. This sequence was repeated for twenty to thirty iterations to ensure that all labeled EVs in the valves were pushed through the detection region. Afterwards, PBS was pulled with a vacuum through the outlet as well as valves 1, 2, 13, 10, 9, and 8 for the remaining 2 minutes.

Fig. 5.

Fig. 5.

Open in a new tab

On-Chip Labeling and Counting – Manual GUI. (a) Average concentration detected with respect to on-chip mixing and incubation times. It was determined that 2-minute mixing and 5-minute incubation resulted in the highest average concentration. (b) Schematic of sample preparation valve network of analysis chip. 1x SYBR Gold stain was inserted in inlet 3 (blue), commercial EVs were inserted in inlet 6 (yellow), and PBS was inserted in the remaining inlets (beige). (c) Time-domain fluorescence time trace of on-chip 2-minute mixing and 5-minute incubation. Detected EVs are indicated with red peaks (d) Intensity histogram of all detected events. (e) A single detected event with respect to the noise background. (f) Live concentration of fluorescence trace over time.

The results of the on-chip labeling and incubation optimization series are shown in Fig. 5a, where each mixing and incubation measurement is a result of three technical replicates. We found that 2-minute mixing and 5-minute incubation resulted in the highest concentration of detected events, indicating optimum labeling conditions in our device. The selected 300s time-domain fluorescence trace shown in Fig. 5c contains 499 detected events, yielding a concentration of 2.22 ×106 particles/mL, with an average FWHM of 1.18×10−4 s per event. Fig. 5d shows the corresponding intensity histogram of the experiment, Fig. 5e shows the fluorescence signal from a single detected particle, and Fig. 5f shows the cumulative detected concentration (cumulative events / elapsed time) throughout the experiment.

C. Automated Analysis of Single EVs – Remote Controlled Analysis Chip + Pump

Our final step is to demonstrate a fully integrated system capable of automated analysis of single EVs through remote control. Within an IoT-based infrastructure [38], sample delivery, on-chip labeling, and detection were coordinated and automated.

Sample delivery was driven by a custom-built syringe pump module as described in [37, 39]. On-chip labeling and detection were performed using optofluidic analysis chips equipped with the pneumatic control module described above and seen in Fig. 6f. Each device contained a single board computer module (Raspberry Pi 4) connected to a local server, and communicated via MQTT messaging protocol [38]. Device operations were initiated by executing prewritten Python scripts in response to incoming MQTT messages.

Fig. 6.

Fig. 6.

Open in a new tab

IoT-Based System Schematic. (a) Remote laptop with Experiment Dashboard webpage connected to a local network enabling IoT-based communication using MQTT protocol. (b) Wireless transmission of commands from the remote laptop to single-board compute modules (Raspberry Pi) (c) Compute module of the Pump (d) or Pneumatic control module (e). Both the pump and pneumatic control module physically interface with the analysis chip (f).

A web-based experiment dashboard, hosted on a remote server enabled real-time control and monitoring. When a user triggered a command through the dashboard, the corresponding MQTT message was transmitted to the compute module of the pump or pneumatic control module. Upon message receipt, the compute module executed the corresponding functions on the device, completing the automated workflow. The system further supports user-defined scheduling of recurring tasks, enabling unattended and continuous operation [38].

To prepare the analysis chip for an automated EV analysis assay, 60 μL of 0.01 mg/ml BSA + 0.05% Tween20 mixture was pipetted across all six inlets of an analysis chip with 0.5mm valve diameter. Then, a message request was sent from the Experiment Dashboard to the control box to actuate all the sample preparation valves to open and pull the mixture through the device for 5 minutes with vacuum. Valves 1, 4, 5, 11, 12, and 15 are closed to prevent additional BSA/Tween20 from entering the device, and the remaining mixture is incubated in the device for 15 mins. Afterwards, the BSA/Tween20 mixture was replaced with PBS buffer in all inlets, and another message request was sent to open all sample preparation valves and pull through the device for 5 mins with a vacuum.

Prior to a measurement, PBS was removed from inlet 3 and replaced with 10 μL of 10x SYBR Gold stain. PBS was also removed from inlet 6 but left empty. The fluidic line that connects to the pump was positioned over inlet 6. As shown in Fig. 6, a message request was first sent to the pump compute module. The pump pulled 500 μL of filtered conditioned media from a vial, then switched ports to the fluidic line that was positioned over inlet 6 of the analysis chip and dispensed the 500 μL of conditioned media into the line. Next, 1 mL of filtered air was pulled and dispensed into the fluidic line to push the aliquot through and prime the fluidic line. Afterwards, 200 μL of filtered air was pushed into the fluidic line and dispensed a 5 μL aliquot of conditioned media to inlet 6 of the analysis chip. Once it was visually confirmed that the aliquot was dispensed, another message request was sent to the pneumatic control module to actuate the pneumatic valves to begin the on-chip labeling and incubation sequence as described above. Immediately after incubation, a final message request was sent to start the detection sequence.

The fully automated and remote controlled sample delivery, on-chip labeling and detection was repeated for three separate measurements. Fig. 7a is the time-domain fluorescence trace for the automated analysis of single EVs and is separated into two sections. The first, denoted in red, shows the detection step of the fluorescence trace where the sample preparation section of the analysis chip was actuating valves to push PBS through the device, without the assistance of vacuum. The second section, denoted in blue, was the washing step of the fluorescence trace where PBS was pulled through the device with vacuum and allowed the remaining labeled EVs to be detected through the excitation region. Fig. 7b shows a zoomed-in view of the fluorescence trace where detected events are clearly visible and differentiated from the background, with a single detected event shown in Fig. 7e. Fig. 7c shows the intensity histogram of the trace and shows that lower intensity events are differentiated from higher intensity events. The detected concentration was determined to be 7.75 ×109 particles/mL with 6859 particles and an average FWHM of 2.02×10−4 s in a 400 s trace. The EV concentration was rapidly determined within 30 seconds in Fig. 7f and corresponds to the first region of Fig. 7b where the initial spike of detected event occurs.

Fig. 7.

Fig. 7.

Open in a new tab

On-Chip Labeling and Counting – Remoted Controlled Analysis Chip and Pump. (a) Time-domain fluorescence of detected EVs labeled on-chip with 10x SYBR Gold. Sample preparation and subsequent detection (red) and washing step were operated remotely with Python scripts. (b) A zoomed-in time-domain fluorescence time trace of detected EVs, indicated by red peaks. Separated regions of detected events indicate valve actuation of labeled EVs being pushed towards the excitation region. (c) Intensity histogram of all detected events. (d) Schematic of sample preparation valve network of analysis chip. 10x SYBR Gold stain was inserted in inlet 3 (blue), commercial EVs were inserted in inlet 6 (yellow), and PBS was inserted in the remaining inlets (beige). (e) A single detected event with respect to the noise background. (f) Live concentration of fluorescence trace over time, where concentration was determined within 30 seconds.

IV. Conclusion

We have demonstrated a system that automates the downstream optical analysis of single EVs by integrating a waveguide-based optofluidic device with a 3D organoid tissue culture platform. First, we verified the detection capabilities of our PDMS devices with the optical counting of single EVs derived from cerebral organoids. Second, we demonstrated the combined functionality of our PDMS devices with manual on-chip sample preparation (EV labeling) and optical counting of single EVs using commercial EVs. Third, we established an integrated system that automates the functionalities of our PDMS device and connects sample delivery to our devices. This fully automated, integrated system was implemented with IoT-connected modules, enabled remote control, and significantly reduced operating costs and the risk of sample mishandling and contamination.

The success of organoid models will require long-term homeostatic organoid growth with real-time results and with minimal disruption. Future work towards advancing organoid culturing technology will include fully automated, repeatable, and quantitative single EV analysis, as well as integrated live organoid culture for continuous analysis. The work presented in this paper represents an important step towards automating organoid culturing systems with EV-derived continuous RNA and protein measurement-driven, feedback-based control and analysis, and enables future research towards long-term organoid growth with real-time results and minimal disruption.

Supplementary Material

supp1-3672359

Additional information is included in Supplementary Material Color versions of one or more of the figures in this article are available online at http://ieeexplore.ieee.org

Acknowledgment

We acknowledge the support and fruitful discussions with G. G. Meena.

This work was funded by the NIH National Human Genome Research Institute, United States under Award number 1RM1HG011543 (D.H., S.R.S., M.T., and H.S), and the Eugene-Cota Robles Fellowship (R.L.). We thank the infrastructure provided by the National Research Platform (NRP), and CENIC for the 100 Gbps networks. The NRP is supported in part by the National Science Foundation (NSF) awards CNS-1730158, ACI-1540112, ACI-1541349, OAC-1826967, OAC-2112167, CNS-2100237, CNS-2120019, the University of California Office of the President, and the University of California San Diego’s California Institute for Telecommunications and Information Technology/Qualcomm Institute.

Biographies

graphic file with name nihms-2159846-b0001.gif

Ravipa Losakul received her B.S. degree in applied physics in 2016 and her M.Sc. degree in electrical engineering in 2018, from the University of California, Santa Cruz, Santa Cruz, CA, USA.

She currently is a Ph.D. candidate in electrical and computer engineering at the University of California, Santa Cruz. Her research interests revolve around the development of nanoscale and optofluidic devices for applications in biosensing, semiconductor physics, and lab-on-chip technologies that enable rapid molecular diagnostics. Previously, her research focused on graphene nanodevices for DNA sequencing, thin film semiconductor engineering, and the characterization of irradiated silicon-based particle detectors.

Miss Losakul is the recipient of the Eugene V. Cota-Robles Fellowship.

graphic file with name nihms-2159846-b0002.gif

Kateryna Voitiuk received her Ph.D. in biomolecular engineering and bioinformatics in 2024 from the University of California, Santa Cruz, Santa Cruz, CA, USA.

She is currently a Postdoctoral Scholar at the University of California, Santa Cruz, with a background bridging computer hardware, software, and biology. Her research focuses on developing integrated laboratory automation systems that combine Internet of Things (IoT) devices, microfluidics, imaging, and electrophysiology to study three-dimensional brain organoid models. She is the Chief Technology Officer and Co-Founder of Open Culture Science, Inc., where she continues to lead technology development for reproducible and scalable biological experimentation.

Dr. Voitiuk is a recipient of the University of California President’s Dissertation-Year Fellowship and has previously been recognized as a Koret Scholar, NIH NHGRI T32 Trainee, and ARCS Foundation Scholar.

graphic file with name nihms-2159846-b0003.gif

Ruiting Xu received her B.S. and M.S. in mechanical engineering from the China University of Petroleum, Beijing, China in 2015 and 2018, respectively, and the Ph.D. in mechanical engineering in 2023 from the University of Akron, Akron, Ohio, USA. She is currently a Postdoctoral Researcher in the electrical and computer engineering at the University of California, Santa Cruz, CA, USA.

Her research interests include microfluidic and nanofluidic sensor design, label-free biomolecular detection, and high-throughput biosensing.

Dr. Xu has authored/coauthored multiple peer-reviewed papers on biosensor development and microfluidic sensing technologies, covering topics such as aptamer-based molecular recognition, resistive pulse sensing, nanoparticle-assisted detection.

graphic file with name nihms-2159846-b0004.gif

Spencer Seiler received his Ph.D. in bioengineering and biomedical engineering in 2024 from the University of California, Santa Cruz, Santa Cruz, CA, USA.

He is currently a Postdoctoral Scholar at the University of California, Santa Cruz, where his research focuses on developing automated microfluidic platforms for neural tissue models that integrate fluidics, sensors, imaging, and high-density electrophysiology for longitudinal studies of brain organoids. His work bridges biotechnology, engineering, and neuroscience to improve experimental control and reproducibility in complex tissue cultures. He is also the Chief Executive Officer and Co-Founder of Open Culture Science, Inc., where he continues to advance open, scalable technologies for automated cell culture and biological experimentation.

Dr. Seiler previously held engineering and development roles at several Bay Area biotechnology startups, including Berkeley Lights, Ultima Genomics, and Miroculus, and participated in the NSF I-Corps program, conducting customer discovery to guide translational research and commercialization of academic innovations.

graphic file with name nihms-2159846-b0005.gif

Viktor Yurevych received his B.S. degree in biochemistry in 2018 from Taras Shevchenko National University of Kyiv, Kyiv, Ukraine and his M.S. degree in medical biology in 2020 from Radboud University, Nijmegen, Netherlands.

He is currently a Ph.D. Candidate at the Salama Lab, University of California, Santa Cruz (UCSC), CA, USA, where he is developing novel liquid biopsy methods for monitoring cortical development and investigating tumor-host crosstalk in glioma-organoid co-culture systems. Previously, he investigated the role of tumor translatomes and novel microproteins in pediatric malignancies at the Van Heesch Lab, Princess Máxima Center for Pediatric Oncology in the Netherlands (2020–2021). He was also a Visiting Researcher at the Klas G. Wiman Lab, Department of of Oncology-Pathology, Karolinska Institutet, Stockholm, Sweden (2019–2020), and a Research Intern at the Jack A. Schalken Lab, Department of Urology, Radboud Institute for Molecular Life Sciences (RIMLS), Netherlands (2019). His current research interests include liquid biopsy methods, extracellular RNA, tumor-host crosstalk, and pediatric malignancies.

Mr. Yurevych received the QB3 Graduate Innovators Fellowship for 2025–2026. He was also the recipient of the Radboud Scholarship Programme and the RadboudUMC Study Fund Scholarship for master's students, both from 2018 to 2020.

graphic file with name nihms-2159846-b0006.gif

Thomas D. Yuzvinsky received his Ph.D. degree in physics with designated emphasis in nanoscale science and engineering from the University of California, Berkeley, Berkeley, CA, USA in 2006. He is currently a staff scientist in the Department of Electrical and Computer Engineering at the University of California, Santa Cruz, CA, USA, where he manages a dualbeam microscopy facility and microfabrication cleanroom, working with multiple research groups to bridge the fields of physics, chemistry, biology, and engineering.

graphic file with name nihms-2159846-b0007.gif

Andy Wang is currently working toward his B.S. degree in electrical and computer engineering at the University of California, Santa Cruz, CA, USA and is an process engineering intern at Applied Materials, Santa Clara, CA, USA. His research focuses on optofluidic devices and materials characterization.

Ivana Pačar received her B.S. in Biomedical/Medical Engineering at the University of Rochester, Rochester, New York, USA. She is currently pursuing a Ph.D at the University of California, Santa Cruz, Santa Cruz, CA, USA.

graphic file with name nihms-2159846-b0008.gif

David Haussler received his Ph.D. in computer science from the University of Colorado at Boulder, CO, USA. He is currently a Distinguished Professor in Biomolecular Engineering at the University of California, Santa Cruz, Scientific Director of the UC Santa Cruz Genomics Institute.

At the intersection of mathematics, computer science, and molecular biology, his experimental research focuses on the molecular evolution of DNA, RNA, and protein sequences, with a current goal in developing a new generation of Live Cell Biotechnology.

Dr. Haussler is a member of the National Academy of Sciences, the National Academy of Engineering, the American Academy of Arts and Sciences, and a fellow of AAAS and AAAI. He has won a number of awards, including the 2015 Dan David Prize, the 2011 Weldon Memorial Prize from University of Oxford, the 2009 ASHG Curt Stern Award in Human Genetics, the 2008 Senior Scientist Accomplishment Award from the International Society for Computational Biology, the 2005 Dickson Prize for Science from Carnegie Mellon University, and the 2003 ACM/AAAI Allen Newell Award in Artificial Intelligence.

graphic file with name nihms-2159846-b0009.gif

Sofie R. Salama received her Ph.D in molecular, cell and developmental biology from the University of the California, Berkeley, Berkeley, CA, USA. She is currently a Professor at the University of California, Santa Cruz, with appointments in Molecular, Cell, and Developmental Biology and Biomolecular Engineering. Her laboratory investigates how recent human genome evolution influences brain development, function, and disease using human and ape brain organoid models. She directs the Genomics Institute’s Center for Live Cell Technologies, an NIH/NHGRI Center of Excellence in Genomic Sciences developing tools for high-content cellular analysis.

Dr. Salama is also a Principal Investigator of the Braingeneers project, which integrates engineering, genomics, and neuroscience to improve brain organoid technologies and longitudinal phenotyping.

graphic file with name nihms-2159846-b0010.gif

Mircea Teodorescu received his B.S. and M.S. degress in mechanical engineering from the Politehnica University of Bucharest, Romania, in 1995 and 1996, respectively, and his Ph.D. in mechanical engineering from Wayne State University, Detroit, MI, USA in 2003. He is currently an Associate Professor at the University of California, Santa Cruz, with appointments in Electrical and Computer Engineering and Biomolecular Engineering, and serves as Director of Bioelectrical Engineering at the UCSC Genomics Institute. His research focuses on scalable in vitro neuroscience platforms that integrate microsystems engineering, real-time biological data acquisition, and neural tissue modeling. Combining mechanical, electrical, and computer engineering principles, his lab develops tools for biomolecular engineering, biomechanics, and assistive technologies through modeling, embedded systems, rapid manufacturing, and sensing.

Dr. Teodorescu is a Principal Investigator of the multi-institutional Braingeneers project, advancing technologies for brain organoid models and longitudinal phenotyping.

graphic file with name nihms-2159846-b0011.gif

Holger Schmidt (Fellow, IEEE) received his Ph.D. degree in Electrical and Computer Engineering from the University of California, Santa Barbara, Santa Barbara, CA, USA. He was a Postdoctoral Fellow with MIT, Cambridge, MA, USA. He is currently the Narinder Kapany Chair of Optoelectronics and Distinguished Professor of Electrical and Computer Engineering at the University of California Santa Cruz, Santa Cruz, CA, USA. His research interests include integrated photonics, optofluidics, single molecule analysis, nanopore sensors and chip-scale spectrometry. He has authored or coauthored more than 500 publications and is a Fellow of Optica and the National Academy of Inventors. He received the Engineering Achievement Award from the IEEE Photonics Society in 2019 and the Innovator of the Year Award at UCSC in 2024.

References

  • [1].Andrews MG, Kriegstein AR, “Challenges of Organoid Research,” Annu Rev Neurosci, vol. 45, pp. 23–39, Jul 2022, doi: 10.1146/annurev-neuro-111020-090812. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [2].Fan X, Hou K, Lui G, Shi R, Wang W, Liang G, “Strategies to overcome the limitations of current organoid technology – engineered organoids,” J. Tissue Eng, vol. 15, no. 16, Apr. 2025, doi: 10.1177/20417314251319475. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [3].Park SE, Georgescu A, Huh D, “Organoids-on-a-chip,” Science, vol. 364, no. 6444 pp. 960–965, Jun 2019, doi: 10.1126/science.aaw7894. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [4].Wang Y and Qin J. “Advances in human organoids-on-chips in biomedical research,” Life Med. vol. 2, Feb 2023, doi: doi: 10.1093/lifemedi/lnad007 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [5].Wang H et al. “Human organoids-on-chips for biomedical research and applications,” Theranostics, vol. 14 (2), pp. 788–818. Jan 2024, doi: 10.7150/thno.90492. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [6].Wang H et al. “Engineering organoids-on-chip for drug testing and evaluation,” Metabolism-Clinical and Experimental, vol. 162, pp. 156065, Nov 2024, 10.1016/j.metabol.2024.156065 [DOI] [PubMed] [Google Scholar]
  • [7].Zhao Z et al. “Organoids,” Nat Rev Methods Primers. vol. 2, pp. 94, Dec 2022, doi: 10.1038/s43586-022-00174-y [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [8].Yang S et al. “Organoids: The current status and biomedical applications.” MedComm, vol. 4, No. 3, pp. e274, May 2023, doi: 10.1002/mco2.274. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [9].Li S, Zhu L, Zhu L, Mei X, Xu W, “A sandwich-based evanescent wave fluorescent biosensor for simple, real-time exosome detection(dagger),” Biosens Bioelectron, vol. 200, no. 113902, Dec 2021, doi: 10.1016/j.bios.2021.113902. [DOI] [PubMed] [Google Scholar]
  • [10].Li J et al. “Highly Sensitive Exosome Detection for Early Diagnosis of Pancreatic Cancer Using Immunoassay Based on Hierarchical Surface-Enhanced Raman Scattering Substrate,” Small Methods, vol. 6, pp. e2200154, Apr. 2022, doi: 10.1002/smtd.202200154. [DOI] [PubMed] [Google Scholar]
  • [11].Ullah Khan N et al. “Ultrasensitive Detection of Exosome Using Biofunctionalized Gold Nanorods on a Silver-Island Film,” Nano Lett, vol. 21, no. 13, pp. 5532–5539, Jun. 2021, doi: 10.1021/acs.nanolett.1c00830. [DOI] [PubMed] [Google Scholar]
  • [12].Liu H et al. “Construction of a sensitive SWCNTs integrated SPR biosensor for detecting PD-L1(+) exosomes based on Fe(3)O(4)@TiO(2) specific enrichment and signal amplification,” Biosensors & Bioelectronics, vol. 262, no. 11, pp. 65–27, Jun. 2024, doi: 10.1016/j.bios.2024.116527. [DOI] [PubMed] [Google Scholar]
  • [13].Yang Y, Zhai C, Zeng Q, Khan AL, Yu H, “Multifunctional Detection of Extracellular Vesicles with Surface Plasmon Resonance Microscopy,” Anal Chem, vol. 92, no. 7, pp. 4884–4890, Mar. 2020, doi: 10.1021/acs.analchem.9b04622. [DOI] [PubMed] [Google Scholar]
  • [14].Wang X et al. “A Target Recycling Amplification Process for the Digital Detection of Exosomal MicroRNAs through Photonic Resonator Absorption Microscopy,” Angew. Chem. Int. Ed, vol. 62, pp. e202217932, Apr 2023, 10.1002/anie.202217932. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [15].Zhuo Y et al. “Single nanoparticle detection using photonic crystal enhanced microscopy,” Analyst, vol. 139, pp. 1007–1015, Jun 2014, 10.1039/C3AN02295A. [DOI] [PubMed] [Google Scholar]
  • [16].Zhao B et al. “Digital-resolution and highly sensitive detection of multiple exosomal small RNAs by DNA toehold probe-base photonic resonator absorption microscopy,” Talanta, vol. 241, pp. 123256, 2022, 10.1016/j.talanta.2022.123256. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [17].Zhao B et al. “Single-step, wash-free digital immunoassay for rapid quantitative analysis of serological antibody against SARS-CoV-2 by photonic resonator absorption microscopy,” Talanta, vol. 225, pp. 122004, 2021, 10.1016/j.talanta.2020.122004. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [18].Canady TD et al. “Digital-resolution detection of microRNA with single-base selectivity by photonic resonator absorption microscopy,” PNAS, vol. 116(39), pp. 19362–19367, Sept 2019, 10.1073/pnas.1904770116. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [19].Ghosh S et al. “A compact photonic resonator absorption microscope for point of care digital resolution nucleic acid molecular diagnostics,” Biomed. Opt. Express, vol. 12(8), pp. 4637–4650, 2021, 10.1364/BOE.427475. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [20].Che C et al. “Accelerated Digital Biodetection Using Magneto-plasmonic Nanoparticle-Coupled Photonic Resonator Absorption Microscopy,” ACS Nano, vol. 16(2), pp. 2345–2354, Jan 2022, 10.1021/acsnano.1c08569. [DOI] [PubMed] [Google Scholar]
  • [21].Cunningham BT, Li P, Lin B, Pepper J, “Colormetric resonant reflection as a direct biochemical assay technique,” Sensors and Actuators B: Chem, vol. 81, no. 2–3, pp. 316–328, Nov. 2001, doi: 10.1016/S0925-4005(01)00976-5 [DOI] [Google Scholar]
  • [22].Lin B et al. “A label-free optical technique for detecting small molecule interactions,” Biosensors and Bioelectronics, vol. 17, no. 9, pp. 827–834, Jun. 2002, doi: 10.1016/S0956-5663(02)00077-5. [DOI] [PubMed] [Google Scholar]
  • [23].Kim R et al. “Colorimetric Detection of HER2-Overexpressing-Cancer-Derived Exosomes in Mouse Urine Using Magnetic-Polydiacetylene Nanoparticles,” Small, vol. 20, no. 13, pp. e2307262, Mar. 2024, doi: 10.1002/smll.202307262. [DOI] [PubMed] [Google Scholar]
  • [24].Li N et al. “Immunofluorescent analysis of exosomes using a microchip filled with transparent antibody-conjugated beads for breast cancer liquid biopsy,” Anal Chim Acta, vol. 1345, pp. 343743, Feb. 2025, doi: 10.1016/j.aca.2025.343743. [DOI] [PubMed] [Google Scholar]
  • [25].Zhao W et al. “Microsphere mediated exosome isolation and ultrasensitive detection on a dielectrophoresis integrated microfluidic device,” Analyst, vol. 146, no. 19, pp. 5962–5972, Sept. 2021, doi: 10.1039/d1an01061a. [DOI] [PubMed] [Google Scholar]
  • [26].Li C et al. “Molybdenum Disulfide-Integrated Iron Organic Framework Hybrid Nanozyme-Based Aptasensor for Colorimetric Detection of Exosomes,” Biosensors (Basel), vol. 13, no. 8, pp. 800, Sept. 2025, doi: 10.3390/bios13080800. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [27].Xu L, Chopdat R, Li D, Al-Jamal KT, “Development of a simple, sensitive and selective colorimetric aptasensor for the detection of cancer-derived exosomes,” Biosens Bioelectron, vol. 169, pp.112576, Sept. 2020, doi: 10.1016/j.bios.2020.112576. [DOI] [PubMed] [Google Scholar]
  • [28].Jung YK and Son MH, “Polydiacetylene-based aptasensors for rapid and specific colorimetric detection of malignant exosomes,” Talanta, vol. 268, pp.125342, Oct. 2023, doi: 10.1016/j.talanta.2023.125342. [DOI] [PubMed] [Google Scholar]
  • [29].Huang Z et al. “Terminal deoxynucleotidyl transferase based signal amplification for enzyme-linked aptamer-sorbent assay of colorectal cancer exosomes,” Talanta, vol. 218, pp. 121089, May 2020, doi: 10.1016/j.talanta.2020.121089. [DOI] [PubMed] [Google Scholar]
  • [30].Kuang J et al. “A colorimetric aptasensor based on a hemin/EpCAM aptamer DNAzyme for sensitive exosome detection,” Analyst, vol. 147, no. 22, pp. 5054–5061, Oct. 2022, doi: 10.1039/d2an01410f. [DOI] [PubMed] [Google Scholar]
  • [31].Kuang J et al. “The integration platform for exosome capture and colorimetric detection: Site occupying effect-modulated MOF-aptamer interaction and aptamer-Au NPs-dopamine interaction,” Anal Chim Acta, vol. 1329, pp.343234, Sept. 2024, doi: 10.1016/j.aca.2024.343234. [DOI] [PubMed] [Google Scholar]
  • [32].Kim HM et al. “Multi-Quantum Dots-Embedded Silica-Encapsulated Nanoparticle-Based Lateral Flow Assay for Highly Sensitive Exosome Detection,” Nanomaterials (Basel), vol. 11, no. 3, pp. 768, Mar. 2021, doi: 10.3390/nano11030768. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [33].Wang YM et al. “Enhancement of the Intrinsic Peroxidase-Like Activity of Graphitic Carbon Nitride Nanosheets by ssDNAs and Its Application for Detection of Exosomes,” Anal Chem, vol. 89, no. 22, pp. 12327–12333, Oct. 2017, doi: 1021/acs.analchem.7b03335. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [34].Kasetsirikul S et al. “Low-cost electrochemical paper-based device for exosome detection,” Analyst, vol. 147, no. 16, pp. 3732–3740, Jul. 2022, doi: 10.1039/d2an00875k. [DOI] [PubMed] [Google Scholar]
  • [35].Zhang W et al. “Electrochemical micro-aptasensors for exosome detection based on hybridization chain reaction amplification,” Microsyst. Nanoeng, vol. 7, no. 63, Aug. 2021, doi: 10.1038/s41378-021-00293-8. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [36].Anderson W, Lane R, Korbie D, and Trau M, “Observations of Tunable Resistive Pulse Sensing for Exosome Analysis: Improving System Sensitivity and Stability,” Langmuir, vol. 31, no. 23, pp. 6577–6587, May 2015, doi: 10.1021/acs.langmuir.5b01402. [DOI] [PubMed] [Google Scholar]
  • [37].Seiler ST, Mantalas GL, Selberg J et al. “Modular automated microfluidic cell culture platform reduces glycolytic stress in cerebral cortex organoids,” Sci Rep, vol. 12, no. 1, pp. 20173, Nov. 2022, doi: 10.1038/s41598-022-20096-9. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [38].Park DF, Voitiuk K et al. “IoT cloud laboratory: Internet of Things architecture for cellular biology,” Internet Things (Amst), vol. 20, pp. 100618, Oct. 2022, doi: 10.1016/j.iot.2022.100618. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [39].Voitiuk K et al. “A feedback-driven brain organoid platform enables automated maintenance and high-resolution neural activity monitoring,” Internet of Things, vol. 33, pp. 10167, Jun. 2025, doi: 10.1016/j.iot.2025.101671. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [40].Doyle LM, Wang MZ, “Overview of Extracellular Vesicles, Their Origin, Composition, Purpose, and Methods for Exosome Isolation and Analysis,” Cells, vol. 8, no.7, pp. 727, Jul. 2019, doi: 10.3390/cells8070727. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [41].Zaborowski MP, Balaj L, Breakefield XO, Lai CP, “Extracellular Vesicles: Composition, Biological Relevance, and Methods of Study,” Bioscience, vol. 65, no. 8, pp. 783–797, Aug. 2015, doi: 10.1093/biosci/biv084. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [42].Arjmandi H et al. “Extracellular Vesicle-Mediated Communication Nanonetworks: Opportunities and Challenges,” in IEEE Communications Magazine, vol. 59, no. 5, pp. 68–73, May 2021, doi: 10.1109/MCOM.001.2000994 [DOI] [Google Scholar]
  • [43].Ferguson S et al. “Single-EV analysis (sEVA) of mutated proteins allows detection of stage 1 pancreatic cancer,” Sci Adv. vol. 8(16), 2022, doi: 10.1126/sciadv.abm3453 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [44].Ferguson S et al. “Single extracellular vesicle analysis for early cancer detection,” Trends Mol. Med. vol. 28(8), pp. 681–692, 2022, doi: 10.1016/j.molmed.2022.05.003. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [45].Tain F et al. “Protein analysis of extracellular vesicles to monitor and predict therapeutic response in metastatic breast cancer,” Nature Commun, vol. 12(1), pp.2536, May 2021, doi: 10.1038/s41467-021-22913-7. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [46].Kalluri R and McAndrews KM, “The Role of Extracelluar Vesicles in Cancer,” Cell, vol. 186(8), pp. 1610–1626, 2023, doi: 10.1016/j.cell.2023.03.010. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [47].Imanbekova M et al. “On the dilemma of using single EVs analysis for liquid biopsy: the challenge of low abundance of tumor EVs in blood,” Theranostics, vol. 15(16), ppl 8031–8048, July 2025, doi: 10.7150/thno.115131. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [48].Sinawang PD et al. “Plasma Preparation Strategies for Extracellular Vesicle-Based Biomarkers in Metastatic Castration -Resistant Prostate Cancer,” Journal of Extracellular Biology, vol. 4(8), pp. e70071, Jul 2025, doi: 10.1002/jex2.70071. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [49].Silver B et al. “Characterization of extracellular vesicles and miRINA released by cerebral organoids,” Curr Res Toxicol. vol. 8, pp. 100229, Mar 2025, doi: 10.1016/j.crtox.2025.100229. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [50].Schmidt H and Hawkins AR, “The Photonic Integration of Non-Solid Media Using Optofluidics,” Nat. Photonics, vol. 2011, no. 5, pp. 598–604, Aug. 2011, doi: 10.1038/nphoton.2011.163. [DOI] [Google Scholar]
  • [51].Fan X and White IM, “Optofluidic Microsystems for Chemical and Biological Analysis, Nature Photonics, vol. 5, no. 10, pp. 591–597, Sept. 2011, doi: 10.1038/nphoton.2011.206. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [52].Hawkins AR and Schmidt H, Handbook of Optofluidics, Boca Raton, FL, USA: CRC Press, 2010, doi: 10.1201/9781420093551 [DOI] [Google Scholar]
  • [53].Psaltis D, Quake SR, Yang C, “Developing Optofluidic Technology through the Fusion of Microfluidics and Optics,” Nature, vol. 442, no. 7101, pp. 381–386, Jul. 2006, doi: 10.1038/nature05060. [DOI] [PubMed] [Google Scholar]
  • [54].Monat C, Domachuk P, and Eggleton B. “Integrated optofluidics: A new river of light.,” Nature Photonics, vol. 1, no. 2, pp.106–114, Feb. 2007, doi: 10.1038/NPHOTON.2006.96. [DOI] [Google Scholar]
  • [55].Fainman Y, Lee L, Psaltis D, and Yang C, Optofluidics: Fundamentals, Devices, and Applications, Columbus, Ohio, USA, McGraw-Hill, Inc., 2009. [Google Scholar]
  • [56].Mandal S, Goddard JM, and Erickson D, “A multiplexed optofluidic biomolecular sensor for low mass detection,” Lab Chip, vol. 9, no. 20, pp. 2924–2932, Jul. 2009, doi: 10.1039/B907826F. [DOI] [PubMed] [Google Scholar]
  • [57].Li Q, van de Groep J, White AK et al. “Metasurface optofluidics for dynamic control of light fields,” Nat. Nanotechnol, vol. 17, pp. 1097–1103, Sept. 2022, doi: 10.1038/s41565-022-01197-y. [DOI] [PubMed] [Google Scholar]
  • [58].Inan H et al. “Photonic crystals: emerging biosensors and their promise for point-of-car applications,” Chem. Soc. Rev, vol. 46, no. 2, pp. 366–388, Nov. 2016, doi: 10.1039/C6CS00206D. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [59].Block ID, Chan LL, Cunningham BT, “Photonic crystal optical biosensor incorporating structure low-index porous dielectric,” Sensors and Actuators B: Chem, vol. 120, no. 1, pp. 187–193, Dec. 2006, doi: 10.1016/j.snb.2006.02.006. [DOI] [Google Scholar]
  • [60].Xiong Y, Huang Q, Canady TD et al. “Photonic crystal enhanced fluorescence emission and blinking suppression for single quantum dot digital resolution biosensing,” Nat. Commun, vol. 13, no. 4647, Aug. 2022, doi: 10.1038/s41467-022-32387-w [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [61].Bhaskar S, Liu W, Tibbs J, Cunningham BT, “Photonic crystal-coupled enhanced steering emission: A prism-free, objective-free, and metal-free loss-less approach for biosensing,” Appl. Phys. Lett, vol: 124, no. 16, pp. 161102, Apr. 2024, doi: 10.1063/5.0203999 [DOI] [Google Scholar]
  • [62].Bhaskar S et al. “Photonic Crystal Grating Resonance and Interfaces for Health Diagnostic Technologies,” Chem. Rev, vol. 125, no. 14, pp. 6435–6540, Jul. 2025, doi: 10.1021/acs.chemrev.4c00653 [DOI] [PubMed] [Google Scholar]
  • [63].Cai H et al. “Optofluidic analysis system for amplification-free, direct detection of Ebola infection”, Sci Rep, vol. 5, no. 14494, Sept. 2015, doi: 10.1038/srep14494. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [64].Ozcelik D et al. “Optofluidic wavelength division multiplexing for single-virus detection,” Proc. Natl. Acad. Sci. U. S. A, vol. 112, no. 42, pp. 12933–12937, Oct. 2015, doi: 10.1073/pnas.1511921112. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [65].Stambaugh A, Parks JW, Stott MA, Meena GG, Hawkins AR, and Schmidt H, “Optofluidic multiplex detection of single SARS-CoV-2 and influenza A antigens using a novel bright fluorescent probe assay,” Proc. Natl. Acad. Sci. U.S.A, vol. 118, no. 20, pp. e2103480118, May 2018, doi: 10.1073/pnas.2103480118 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [66].Sampad MJN et al. “Label-free and amplification-free viral RNA quantification from primate biofluids using a trapping-assisted optofluidic nanopore platform,” Proc. Natl. Acad. Sci. U. S. A, vol. 21, no. 16, pp. e2400203121, Apr. 2024, doi: 10.1073/pnas.2400203121 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [67].Kim S, Streets AM, Lin RR, Quake SR, Weiss S, and Majumdar DS, “High-throughput single-molecule optofluidic analysis,” Nat Methods, vol. 8, pp. 242–245, Feb. 2011, doi: 10.1038/nmeth.1569. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [68].Parks JW, Cai H, Zempoaltecatl L, Yuzvinsky TD, Leake K, Hawkins AR, and Schmidt H, “Hybrid Optofluidic Integration” Lab on a Chip, vol. 13, no. 20 pp. 4118–4123, Aug. 2013, doi: 10.1039/c3lc50818h. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [69].Parks JW and Schmidt H, “Flexible Optofluidic Waveguide Platform with Multi-Dimensional Reconfigurability,” Sci. Rep, vol. 6, pp. 33008, Sept. 2016, doi: 10.1038/srep33008. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [70].Du K et al. “Multiplexed Efficient On-Chip Sample Preparation and Sensitive Amplification-Free Detection of Ebola Virus”, Biosensors Bioelectronics, vol. 91, pp. 489–496, May 2017, doi: 10.1016/j.bios.2016.12.071. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [71].Agaoglu S, Robles MC, Smith CD et al. “The effect of pre-polymer/cross-linker storage on the elasticity and reliability of PDMS microfluidic devices,” Microfluid. Nanofluid, vol. 21, no. 117, Jun. 2017, doi: 10.1007/s10404-017-1953-0 [DOI] [Google Scholar]
  • [72].Du K et al. “Microfluidic System for Detection of Viral RNA in Blood Using a Barcode Fluorescence Reporter and a Photocleavable Capture Probe”, Analytical Chemistry, vol. 89, no. 17, pp. 12433–12440, Oct. 2017, doi: 10.1021/acs.analchem.7b03527. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [73].Meena GG, Jain A, Parks A,JW, Stambaugh A, Patterson JL, Hawkins AR, Schmidt H, “Integration of Sample Preparation and Analysis into an Optofluidic Chip for Multi-Target Disease Detection,” Lab Chip, vol. 18, pp. 3678–3686, Oct. 2018, doi: 10.1039/c8lc00966j. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [74].Sano T, Zhang H, Losakul R, and Schmidt H, “All-in-One Optofluidic Chip for Molecular Biosensing Assays”, Biosensors, vol.12, no. 7, pp.501, Jul. 2022, doi: 10.3390/bios12070501 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [75].Sano T, Losakul R, and Schmidt H, “Dual optofluidic distributed feedback dye lasers for multiplexed biosensing applications,” Sci Rep, vol. 13, pp. 16824, Oct 2023, doi: 10.1038/s41598-023-42671-4. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [76].Hong J and Quake S. “Integrated nanoliter systems,” Nat Biotechnol, vol. 21, pp. 1179–1183, Sept. 2003, doi: 10.1038/nbt871 [DOI] [PubMed] [Google Scholar]
  • [77].Jensen EC, Bhat BP, and Mathies RA, “A digital microfluidic platform for the automation of quantitative biomolecular assays,” Lab Chip, vol. 10, no. 6, pp. 685–691, Jan. 2010, doi: 10.1039/b920124f. [DOI] [PubMed] [Google Scholar]
  • [78].Kim J, Kang M, Jensen EC, and Mathies RA, “Lifting gate PDMS microvalves and pumps for microfluidic control,” Anal. Chern, vol. 84, no. 4, pp. 2067–2071, Jan. 2012, doi: 10.1021/ac202934x. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [79].Jensen EC, Stockton A, Chisel T, Kim J, and Mathies RA, “Digital Microfluidic automaton for multi-scale combinatorial mixing and sample processing,” Lab Chip, vol. 13, no. 2, pp. 288–96, Oct. 2012, doi: 10.1039/c2lc40861a. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [80].Kim J, Jensen EC, Stockton A, and Mathies RA, “Universal Microfluidic Automaton for Autonomous Sample Processing: Application to the Mars Organic Analyzer”, Anal. Chem, vol. 85, no. 16, pp. 7682–7688, May 2013, doi: 10.1021/ac303767m. [DOI] [PubMed] [Google Scholar]
  • [81].Cai Z, Qiu W, Shao G, Wang W, “A new fabrication methods for all-PDMS waveguides,” Sensors and Actuators A Phys, vol. 204, pp. 44–47, Dec. 2013, doi: 10.1016/j.sna.2013.09.019. [DOI] [Google Scholar]
  • [82].Ganjalizadeh V, Meena GG, Wall TA, Stott MA, Hawkins AR, and Schmidt H, “Fast custom wavelet analysis technique for single molecule detection and identification,” Nat Commun, vol. 13, no. 1, 1035, Feb. 2022, doi: 10.1038/s41467-022-28703-z. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [83].Ganjalizadeh V, Hawkins AR, and Schmidt H, “Adaptive time modulation technique for multiplexed on-chip particle detection across scales,” Optica, vol. 10, no. 7, pp. 812–818, Jun. 2023, doi: 10.1364/optica.489068. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [84].Ganjalizadeh V, G Meena G, Stott MA, Schmidt H, and Hawkins AR, “Single Particle Detection Enhancement with Wavelet-based Signal Processing Technique,” Conference on Lasers and Electro-Optics, San Jose, CA, USA, 2019, doi: 10.1364/CLEO_SI.2019.STu3H.4 [DOI] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

supp1-3672359
NIHMS2159846-supplement-supp1-3672359.pdf (538.2KB, pdf)

RESOURCES