Abstract
Widefield GCaMP imaging has become an essential tool in neuroscience, enabling real-time visualization of neuronal activity, particularly in ultrasound neuromodulation studies. However, traditional ultrasound transducers obstruct the light path, thereby complicating the alignment of the optical and acoustic paths. This study introduces a transparent ultrasound transducer (TUT) fabricated from lithium niobate, specifically designed to facilitate the alignment of the optical and acoustic paths. Indium tin oxide (ITO) is used as the transparent electrode of the TUT, providing transmittance higher than 70% at wavelengths critical for GCaMP imaging. This system supports simultaneous ultrasound stimulation and imaging without optical interference, effectively accommodating the needs of dual-channel widefield GCaMP imaging. By alternating between blue and violet light for stimulation, our system allows for the distinction between neural and non-neural activity.
Keywords: Transparent Ultrasound Transducer, GCaMP Imaging, In Vivo, Neuromodulationd
I. Introduction
GCaMP imaging is a powerful technique widely used in neuroscience to monitor calcium dynamics within neurons, as the intracellular concentration of calcium is closely linked to neural activity. GCaMP, a genetically encoded calcium indicator, fluoresces upon binding to calcium, providing a real-time, indirect measure of neuronal firing. This technique has revolutionized brain function studies by enabling the observation and quantification of neural activity with high spatial and temporal resolution. In widefield GCaMP imaging, large portions of the brain are illuminated, and the resulting fluorescence is captured, offering a comprehensive view of activity across different brain regions. GCaMP imaging permits the tracking of rapid changes in neural activity, making it an indispensable tool for real-time neuroimaging.
The development of imaging technologies that enable the real-time observation of neuronal activity is crucial for advancing our understanding of brain function, particularly in the context of neuromodulation. GCaMP imaging, which allows for the visualization of calcium dynamics within neurons, has been widely adopted in ultrasound neuromodulation research [1, 2]. However, integrating this technique with ultrasound stimulation posed significant challenges due to the physical obstruction caused by traditional transducers in the light path. To circumvent this issue, researchers have resorted to tilting the transducer [1, 2] or using a ring-shaped transducers with a central hole for optical imaging [3, 4]. Both approaches introduce complications in aligning the acoustic and optical paths.
In this study, we introduce a transparent ultrasound transducer (TUT) designed to overcome these limitations. Fabricated from lithium niobate [5, 6], the TUT features a transparent electrode made of indium tin oxide (ITO) [5, 6] and exhibits a resonance frequency of approximately 3.1 MHz. Additionally, the inclusion of an ITO-coated glass matching layer provides an additional resonance frequency at 1.38 MHz. The transducer’s high optical transmittance across the visible spectrum ensures that the GCaMP imaging system can function effectively without optical interference from the ultrasound transducer.
Our in vivo experiments using both blue light and violet light demonstrate the efficacy of this dual-channel system, confirming that the TUT can seamlessly integrate ultrasound stimulation with GCaMP imaging without compromising image quality. By comparing the results from blue light and violet light excitation, it is possible to distinguish between neural and non-neural activity, such as hemodynamic-related changes [7]. By enabling clear and unobstructed imaging across these spectral ranges, the TUT represents a significant step forward in neuroscience, offering enhanced capabilities for future neuromodulation research.
II. Method
A. Transparent Ultrasound Transducer Design
The TUT was fabricated using 36° Y-cut lithium niobate (LiNbO3), a well-known transparent piezoelectric material. ITO was employed as the electrode for transparent electrical patterns. After patterning ITO (250 nm) and Ti/Au (20 nm/400 nm) electrodes on a 3-inch lithium niobate wafer with a thickness of 1 mm using a lift-off process, the wafer were diced into a size of 10 mm × 10 mm. In this study, a lithium niobate piece with a 7 mm diameter ITO pattern was used. To enhance the optical transmittance and electrical performance of the ITO layer, an annealing process was employed, resulting in a sheet resistance of approximately 10.9 Ω/sq. The transducer housing was 3D printed with PLA, and electrical connections were made using conductive epoxy, wire, and an SMA connector, as shown in Fig. 1 (a).
Fig 1.
(a) Photo image, (b) transmittance, (c) electrical input impedance, (d) surface displacement per input voltage of the TUT
Typically, 1 mm thick lithium niobate has a resonance frequency above 3 MHz. To operate at lower frequencies and for the front size electrical connection of lithium niobate, 1.1 mm thick ITO coated glass was used as a matching layer. As shown in Fig. 1 (c), when 1 mm thick lithium niobate and 1.1 mm thick glass were bonded, the resonance frequencies were found to be approximately 1.35 MHz and 3.05 MHz, as shown in Fig. 1 (c), based on the electrical impedance calculated by the KLM model [8] and electrical impedance measured with an impedance analyzer (4294A, Keysight Technologies, Santa Rosa, CA, USA).
GCaMP imaging typically involves using blue light (480-490 nm) for excitation and detecting green light (510-520 nm) for emission. For dual-channel widefield GCaMP imaging, violet light (405 nm) is also used for excitation [7]. The TUT designed for dual-channel widefield GCaMP imaging must have high transmittance for blue, violet, and green light. The transmittance of ITO coated lithium niobate and ITO coated glass was measured using a spectrophotometer (Cary 6000i UV-Vis-NIR, Agilent, Santa Clara, CA, USA) and was found to be higher than 70% at wavelengths above 400 nm.
The surface displacement per input voltage in an air medium calculated from KLM model shows the peak values of 5.25 nm/V and 1.50 nm/V at 1.35 MHz and 3.05 MHz, respectively. The surface displacement measured by the laser Doppler vibrometer (LDV, OFV-2700, Polytec, Menlo Park, CA, USA) was approximately 4.86 nm/V and 1.00 nm/V at 1.36 MHz and 3.03 MHz, respectively (Fig. 1 (d)).
B. Dual-Channel Widefield GCaMP Imaging
For the in vivo experiments, the TUT was integrated into a dual-channel widefield GCaMP imaging system. The transducer was positioned between the imaging target and the microscope, enabling simultaneous ultrasound stimulation and GCaMP imaging. Imaging was performed with a PCO Edge 5.5 CMOS camera at 20 frames per second, alternating between blue light (470 nm) and violet light (405 nm) excitation every other frame while filtering emitted light with a band-pass filter centered at 525 nm (green). The frames were then separated into separate blue and violet light channels at 10 frames per second. The emission spectrum of GCaMP in response to violet light is independent of calcium concentration, which is useful for identifying hemodynamic and other artifacts that may be present in the blue channel. The light is activated at the start of the recording and ultrasound was applied for 300 ms starting 10 seconds after the start of the recording.
GP4.3 mice expressing GCaMP6s in excitatory neurons of the cortex were prepared for experiments by removing the scalp and thinning the skull under 4% isoflurane, then applying a thin coating of MetaBond adhesive. Mice were allowed to recover from surgery for at least 1 week before recording. For recordings mice were anesthetized with 4% isoflurane and maintained at 1% isoflurane.
III. Result
A. Ultrasound Transducer Characterization
Pressure field measurements were performed using a scanning system (AIMS III, Onda Corporation, Sunnyvale, CA, USA) and a hydrophone (HGL-0200, Onda Corporation, Sunnyvale, CA, USA) to measure the output pressure of the ultrasound transducer. The resonance frequencies were determined by measuring the pressure while varying the frequency at a location 17.5 mm away from the transducer surface. The output pressure showed peaks at 1.38 MHz, 3 MHz, and 4.3 MHz, respectively, as shown in Fig. 3 (a).
Fig 3.
Pressure measurement result of the TUT. (a) Frequency response at 17.5 mm far from the surface. 2D scan result at (b) 1.38 MHz, (c) 3 MHz, and (d) 4.3 MHz.
Since the fabricated TUT is a flat transducer, the location of the natural focus varies depending on the frequency. As shown in Fig. 3, The natural focus occurred at approximately 24 mm, 28 mm, and 35 mm away from the transducer surface at 1.38 MHz, 3.1 MHz, and 4.3 MHz, respectively. The pressure at the natural focus was 24 kPapp, 117 kPapp, and 128 kPapp at 1.38 MHz, 3.1 MHz, and 4.3 MHz, respectively.
B. In Vivo Widefield GCaMP Imaging
A widefield GCaMP fluorescence image of the mouse brain was acquired with a camera through the TUT, as shown in Fig. 4 (a). The GCaMP fluorescence image taken at t = 2 s is acquired 2 seconds after the light source was turned on, and the image taken at t = 12 s is acquired 2 seconds after the ultrasound transducer was activated. The widefield images acquired through the TUT show signals presumed to be neural activity, and the left and right cerebral cortex can be identified.
Fig 4.
(a) Widefield image of GCaMP fluorescence from blue light excitation experiment at t = 2 s and 12 s. Average of the widefield image by time for (b) blue light and (c) violet light excitation from the experiment using a frequency of 1.38 MHz.
Fig. 4. (b) and (c) shows the average results of blue light and violet light excitation across 10 trials, where the ultrasound transducer was operated at a frequency of 1.38 MHz. In this experiment, signals presumed to be neural activity were observed in the blue light excitation results, but not with violet excitation.
In the blue light excitation results of the experiment conducted at a frequency of 3 MHz, a change in GCaMP fluorescence was observed 2 seconds after the light was turned on similar to the 1.38 MHz experiment, but no change due to ultrasound was observed (Fig. 5 (a)). However, in the violet light excitation results, GCaMP fluorescence decreased immediately after the ultrasound transducer was activated.
Fig 5.
Average of the widefield image by time for (b) blue light and (c) violet light excitation from the experiment using a frequency of 3 MHz.
IV. Conclusion
This study introduced the TUT designed to integrate with widefield GCaMP imaging, addressing the limitations of traditional transducers. The TUT, made from lithium niobate with a transparent ITO electrode, demonstrated high optical transmittance and effective resonance frequencies, enabling simultaneous ultrasound stimulation and GCaMP imaging. While in vivo widefield GCaMP imaging experiments indicated compatibility of the TUT with dual-channel blue and violet light excitation, further investigation is needed to fully interpret the potential differentiation between neural and non-neural activity observed in these experiments. Nevertheless, these preliminary findings suggest that the TUT could enhance the precision and effectiveness of GCaMP imaging in neuroscience research, particularly in the context of ultrasound neuromodulation.
Fig 2.
Experiment setup for in-vivo dual-channel widefield GCaMP Imaging
TABLE I.
Acoustic Properties for KLM Model
| Lithium Niobate |
Glass | |
|---|---|---|
| Density [kg/m3] | 4640 | 2300 |
| Longitudinal Wave Velocity [m/s] | 7390 | 5500 |
| Relative Dielectric Constant | 58 | – |
| Piezoelectric Coupling Coefficient | 0.4232 | – |
Acknowledgment
This research was supported by the National Institutes of Health Brain Research through Advancing Innovative Neurotechnologies (BRAIN) initiative R01 NS112152. Part of this work was performed at the Stanford Nano Shared Facilities (SNSF) and Stanford Nanofabrication Facilities (SNF), supported by the National Science Foundation under award ECCS-2026822
Contributor Information
Young Hun Kim, E. L. Ginzton Lab, Department of Electrical Engineering, Stanford University, Stanford, CA, USA.
Martin Loynaz Prieto, Department of Molecular and Cellular Physiology, Stanford University, Stanford, CA, USA.
Chunfu Lin, E. L. Ginzton Lab, Department of Electrical Engineering, Stanford University, Stanford, CA, USA.
Yichi Zhang, E. L. Ginzton Lab, Department of Electrical Engineering, Stanford University, Stanford, CA, USA.
Kamyar Firouzi, E. L. Ginzton Lab, Department of Electrical Engineering, Stanford University, Stanford, CA, USA.
Merritt Maduke, Department of Molecular and Cellular Physiology, Stanford University, Stanford, CA, USA.
Kim Butts Pauly, Department of Radiology, Stanford, University, Stanford, CA, USA.
Pierre Khuri-Yakub, E. L. Ginzton Lab, Department of Electrical Engineering, Stanford University, Stanford, CA, USA.
References
- [1].Yoo S, Mittelstein DR, Hurt RC, Lacroix J, and Shapiro MG, “Focused ultrasound excites cortical neurons via mechanosensitive calcium accumulation and ion channel amplification,” Nature Communications, vol. 13, no. 1, p. 493, 2022, doi: 10.1038/s41467-022-28040-1. [DOI] [Google Scholar]
- [2].Guo H. et al. , “Effects of focused ultrasound in a ‘clean’ mouse model of ultrasonic neuromodulation,” iScience, vol. 26, no. 12, p. 108372, 2023, doi: 10.1016/j.isci.2023.108372. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [3].Estrada H. et al. , “High-resolution fluorescence-guided transcranial ultrasound mapping in the live mouse brain,” Science Advances, vol. 7, no. 50, p. eabi5464, 2021, doi: 10.1126/sciadv.abi5464. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [4].Murphy KR et al. , “A tool for monitoring cell type–specific focused ultrasound neuromodulation and control of chronic epilepsy,” Proceedings of the National Academy of Sciences, vol. 119, no. 46, p. e2206828119, Nov. 2022, doi: 10.1073/pnas.2206828119. [DOI] [Google Scholar]
- [5].Dangi A, Agrawal S, and Kothapalli S-R, “Lithium niobate-based transparent ultrasound transducers for photoacoustic imaging,” Optics Letters, vol. 44, no. 21, pp. 5326–5329, 2019, doi: 10.1364/OL.44.005326. [DOI] [PubMed] [Google Scholar]
- [6].Cho S. et al. , “An ultrasensitive and broadband transparent ultrasound transducer for ultrasound and photoacoustic imaging in-vivo,” Nature Communications, vol. 15, no. 1, p. 1444, 2024, doi: 10.1038/s41467-024-45273-4. [DOI] [Google Scholar]
- [7].Couto J. et al. , “Chronic, cortex-wide imaging of specific cell populations during behavior,” Nature Protocols, vol. 16, no. 7, pp. 3241–3263, 2021, doi: 10.1038/s41596-021-00527-z. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [8].Krimholtz R, Leedom DA, and Matthaei GL, “New equivalent circuits for elementary piezoelectric transducers,” Electronics Letters, vol. 6, no. 13, pp. 398–399(1), Jun. 1970, [Online]. Available: https://digital-library.theiet.org/content/journals/10.1049/el_19700280 [Google Scholar]