I. Core Research Summary

• Aerosol ultra-thin mesh dry electrode, skin-conformal and gel-free.

By directly writing an aerosol‑based silver nanowire serpentine mesh structure, the electrode achieves a total thickness of only 67 μm and can conform tightly to the curved surfaces of the eye socket and nasal bridge without the need for conductive adhesive. It withstands 100% biaxial stretching and 180° bending with a small radius, while maintaining a negligible increase in contact impedance over one hour of continuous monitoring, with a signal‑to‑noise ratio comparable to that of medical gel electrodes.

• High-precision real-time recognition of ocular vergence signals

Optimized the three-channel nasion reference electrode configuration to distinguish near‑, mid‑, and far‑range vergence movements at 1°, 2°, and 3°, achieving a signal classification accuracy of 91%. Paired with sixth-order filtering and an integrated subspace classification algorithm, it effectively suppresses blinks and head‑motion artifacts, enabling detection of eye‑movement differences as small as 0.5°.

• VR Integrated Home-Based Rehabilitation Training System

Two standard ophthalmic training programs—Virtual Brock and Eccentric Ring—eliminate the interference caused by head tremors in physical instruments. Compared with traditional physical training devices, the VR‑based system achieves an 11% improvement in recognition accuracy and can generate real-time quantitative training data.

• Fully flexible wireless data acquisition hardware

Aerosol‑printed conductive traces are integrated with a laser‑processed PI substrate, while a miniature Bluetooth low‑energy circuit is embedded in a flexible rear‑neck substrate. A magnetically attached lithium battery enables quick and easy removal, and the entire system operates without rigid cables, making it compatible with both everyday wearables and VR headsets when worn over them.

• Large-scale, low-cost additive manufacturing solution

Aerosol‑based digital direct writing eliminates the need for photolithographic masks; silver nanoparticle ink is deposited on demand, enabling batch customization of electrode dimensions to suit various facial shapes. The entire flexible circuit is formed at low temperatures and is compatible with medical‑grade elastic transfer substrates.

II. Core Operating Principle of Aerosol Spraying and Its Unique Process Advantages

(1) Complete Preparation Process for Flexible Bioelectrodes

• PMMA sacrificial layer spin-coated onto a glass substrate, with plasma activation of the platform to enhance ink wettability;
• Ultrasonic atomization of silver nanoparticle ink, with coaxial sheath‑gas jet focusing, enables aerosol direct‑write of serpentine mesh‑like conductive patterns.
• Low-temperature sintering at 200°C removes the oleylamine binder coating on the ink, enhancing the densification of silver particles and reducing electrical impedance.
• RIE etches away the excess PI regions, acetone is used to dissolve the sacrificial layer, and the complete electrode is transferred onto a medical-grade elastic film.
• Flexible aerosol wiring, chip bonding, and elastic encapsulation are employed to assemble the wireless data acquisition module.

(II) Five Core Process Advantages (Compared with Lithography / Sputtering / Screen Printing)

• Direct Writing of Ultra-Thin Mesh Electrodes for Curved-Surface Matching

Aerosol-based non-contact deposition enables the printing of micron-scale hollow serpentine grids; the resulting thin film is exceptionally flexible and conformal, seamlessly adhering to the irregular curved surfaces of the nasal bridge and periorbital region—something rigid lithographic electrodes cannot achieve.

• Ink parameters are precisely adjustable, balancing conductivity and flexibility.

It allows for free adjustment of atomization flow rate, printing layer count, and sintering duration, enabling precise control over silver film thickness and porosity while balancing low impedance with high tensile performance. A single device can be customized to accommodate a wide range of biosensor electrodes.

• Maskless Digital Rapid Prototyping

With CAD, you can instantly adjust mesh line widths and electrode dimensions, rapidly iterate designs to accommodate both pediatric and adult facial morphologies, eliminate the need for photolithography mask fabrication, and significantly shorten the R&D cycle for small‑batch medical devices.

• Low-temperature molding compatible with medical-grade flexible substrates

The maximum sintering temperature is 200°C, which does not damage the elastic transfer film or medical-grade PDMS. After printing, the structure can be peeled off intact from the skin‑adhesive substrate without leaving a rigid backing.

• Material utilization has improved significantly.

Localized silver nanoparticle ink deposition on demand, as opposed to full‑area sputter coating, significantly reduces precious‑metal consumable usage, making it ideal for cost‑effective mass production of medical consumables.

III. Five Core Technological Innovations of This Study

Innovation 1: A snake‑shaped, mesh‑structured bioelectrode specifically designed for aerosols

Innovation 2: Low-Temperature Silver Nanosintering Process System

Innovative aerosol electrode–compatible three-channel nasal bridge reference sampling architecture

Innovative 4‑aerosol routing + flexible hybrid circuit integration solution

Innovation 5: Printed Electrodes and a VR-Based Rehabilitation Closed-Loop System

IV. Existing Process Challenges and Corresponding Solutions for Overcoming Them

Challenge 1: Traditional solid conductive films fracture under tensile strain, and skin deformation generates substantial motion‑induced noise.

Challenge 2: Incomplete sintering of silver nanoparticle ink results in high electrode contact resistance and a poor signal-to-noise ratio.

Challenge 3: It is difficult to achieve uniform deposition of conductive traces on the curved surfaces around the eyes and along the nasal bridge.

Challenge 4: Conventional printed electrodes have fixed dimensions, making it impossible to accommodate individuals with different facial shapes.

Challenge 5: Multi-layer film transfer is prone to damage, resulting in a low electrode integrity rate.

V. Future Diversified Industrial Application Scenarios

1. Home-use ophthalmic vision rehabilitation device

2. Wearable devices for monitoring and managing myopia in children

3. Clinical Ophthalmic Diagnostic Support Devices

4. Virtual reality human-computer interaction peripherals

5. Neurological Disease Screening Tool

Figure 1. Schematic diagram of the overall architecture of a soft, wireless periorbital electronic system integrated with a virtual reality environment, designed for home-based treatment of ocular disorders. (A) A portable wearable device that combines skin‑like electrodes, flexible circuits, and associated therapeutic equipment with a VR system to assess convergence movements; an inset illustrates the magnetically integrated battery‑powered module. (B) A VR therapeutic environment simulating the continuous motion of multiple objects at three distinct depths—near, intermediate, and far—corresponding to eye movement ranges of 1°, 2°, and 3°, respectively. (C) A close-up view of the skin‑like electrode making conformal contact with the nasal region. (D) A highly flexible, bendable electronic circuit mounted on the posterior neck. (E, F) X‑ray images showing the enlarged mesh‑like interconnect structure (E) and the circuit’s curved configuration with a small bending radius (F). (G–I) Photographs of ultra‑thin mesh‑structured electrodes positioned near the eyebrow area. (J, K) Two types of ocular misalignment detected by this wireless ocular wearable device: esotropia (inward deviation of the eyes) (J) and exotropia (outward deviation of the eyes) (K). Image source: Saswat Mishra; imaging institution: Georgia Institute of Technology.

Figure 2. AJP parameter study based on characterization of deposited nanoparticle inks. (A) Close-up of the AJP architecture and deposition process within the deposition head. (B) The atomization process begins inside an ultrasonic vial, where a carrier gas (N₂) transports ink droplets through a delivery tube, a diffuser, and the deposition head; a sheath gas then focuses the particles into a narrow jet approximately 5 μm in diameter. (C) A heated platen ensures complete solvent evaporation, but surface treatment with a plasma cleaner yields sharp lines (top image); otherwise, traces tend to form bubbles over larger areas. (D) A high focusing ratio combined with a low width-to-thickness ratio reduces resistivity. (E) Resistivity values were measured using a four-point probe system at various sintering temperatures and durations. (F) Electrode resistance must remain low; thus, two-point probe measurements reveal the resistance distribution across series‑connected fractal electrodes. (G) Silver nanoparticles capped with oleylamine fatty acids decompose after both low- and high-temperature sintering: for small nanoparticles, grain size increases as temperature decreases, whereas for large nanoparticles, grain size grows with rising temperature. (H) SEM images show agglomeration and densification of binder materials following sintering. (I) X-ray diffraction analysis indicates that sintering leads to enlarged crystalline peaks and recrystallization along the (111), (200), and (311) planes; a prominent peak between the (220) and (311) planes corresponds to the silicon substrate. (J) Raman spectroscopy reveals that, at around 2900 cm⁻¹, dissociation of the binder results in the disappearance of carbon and hydrogen peaks. (K) Cyclic tensile testing of the electrode (strain of 50%). (L) Cyclic bending of the electrode (maximum angle of 180°).

Figure 3. Eye‑vergence studies based on sensor placement and analysis of classification accuracy. (A) Eye‑vergence and discrete saccadic movements, along with their corresponding EOG signals, recorded using three distinct targets positioned at distances of 40, 60, and 400 cm. (B and C) Sensor configurations simulating the conventional setup (two recording channels and one ground reference). (D and E) An optimized sensor arrangement that significantly improves eye‑vergence detection accuracy to 87%. (F and G) The final sensor configuration (Eye‑Vergence Mode 2), comprising three recording channels and one ground reference, achieves the highest possible accuracy in eye‑vergence detection. Photo credit: Saswat Mishra; photography by Georgia Institute of Technology.

Figure 4. Optimization of vergence analysis through signal processing and feature extraction. (A) Raw vergence signals acquired in real time by a wireless eye‑wearable device (top) and their corresponding derivative signals (bottom). a.u. denotes arbitrary units. (B) Data preprocessed with a bandpass filter, along with its second derivative. (C) Data further processed using a 500‑point median filter, together with its sixth derivative. To amplify the amplitudes of the second‑ and sixth‑order differential filters, scaling factors were applied. (D) Comparison of the signal‑to‑noise ratios of the second‑ and sixth‑order differential filters, showing that the sixth‑order data exhibit a broader detection range; this data was used for thresholding the vergence signals to enable real‑time classification of the dataset. (E) The sliding‑window–processed data were fed into an ensemble subspace classifier, represented by the two‑dimensional decision boundary of the feature set.

Figure 5. Comparison of eye‑movement classification accuracy between virtual reality devices and physical apparatuses. (A and B) Examples of normalized vergence saccade data recorded using a virtual reality system (A) and a physical device (B); the red line indicates the mean, and the gray shading represents the standard deviation. The virtual reality device employs a higher signal gain than the physical device. (C and D) Normalized peak velocities derived from vergence and disconjugate normalized positions in the virtual reality system (C) and the physical device (D). (E and F) Summary comparison of average classification accuracies, based on cross‑validation results, for a wearable soft‑eye‑tracking electronic device equipped with virtual reality technology (accuracy: 91%) (E) versus a physical device (accuracy: 80%) (F).

Figure 6. A convergence therapy system based on virtual reality technology. (A) An example of an ocular‑therapy protocol within this VR system: the Brock String (top image shows a 100‑cm‑long cord with three beads; bottom graph displays the measured ocular convergence signal); (B) a protocol called “Eccentric Circles,” which employs two cards each bearing concentric circle patterns (top image) along with the corresponding EOG signals (bottom graph); (C) after continuous use of this VR headset‑based program, all three participants—without strabismus or near‑point convergence difficulties—exhibited improvements in their ocular convergence ability; (D) a photograph of exotropia, illustrating a participant’s difficulty maintaining near‑point convergence during pencil push‑ups; (E) the corresponding EOG signals during convergence and divergence in the pencil‑push‑up task: the left eye exhibits a slower response during convergence, followed by an outward deviation upon blinking into the near‑distance position; Photo source: Saswat Mishra; Photography institution: Georgia Institute of Technology.