Content uploaded by Anuj Baskota
Author content
All content in this area was uploaded by Anuj Baskota on Aug 29, 2023
Content may be subject to copyright.
Compact GHz Ultrasonic Micro-Imager for Cells and Tissues
Anuj Baskota
1,
*, Justin Kuo
1
, Serhan Ardanuç
1
, and Amit Lal
1
1
Geegah Inc., Ithaca, NY, United States
*Corresponding author: anuj@geegah.com
Optical microscopy is a powerful tool for studying cells and tissues, and it is arguably the most well- known type of microscopy
due to its ubiquity in labs and schools. Ultrasonic imaging, while not as common as optical microscopy, has also been a valuable
imaging modality for visualizing tissues and organs in real-time and in a non-invasive way. The majority of medical diagnostic
processes use ultrasonic pulses in the 3–10 MHz range, which typically have a lateral resolution of 0.15 to 5 mm [1,2].
However, these devices are bulky systems that limit their mobility and are extremely expensive. Here, we report on a compact,
portable ultrasonic imager that can operate in the GHz frequency range with a spatial resolution of 50 μm and can show real-time
images at a frame rate of up to 20 fps. We demonstrate the scanning of chicken tissues, showing contrast between fat and muscle
cells, as well as the visualization of individual onion cells. Using ultrasound instead of light also allows for imaging in the dark and
provides additional information about the mechanical properties of tissues, in addition to the visualization aspect.
The imaging system, which, to our knowledge, is the first “ultrasonic camera”of its kind, consists of an array of 128x128,
50x50μm Aluminum Nitride (AlN) transducers. The necessary circuitry for transmit and receive for each transducer is integrated
within the pixel using a 130 nm CMOS process. The overall packaging of the GHz ultrasonic imager is compact and handheld,
with a silicon surface as the sensing region (Figure 1A). More details on the circuitry and fabrication of the GHz imager, with
another application of imaging nematodes, are described in our previous works [3,4]. Ultrasonic pulses generated by the trans-
mitting pixels are transmitted and reflected off the backside of the silicon. The silicon layer opposite to the CMOS/transducer side
becomes the imaging surface where samples can be placed (Figure 1B).
Thin slices of approximately 100 to 200 μm thick tissues, consisting of fat and muscles, were extracted from chicken thighs
(commercially purchased from a supermarket) using a biopsy needle. The sample was placed on the silicon surface for imaging.
To demonstrate the ability of the imager to visualize individual cells, a thin membrane of onion cells was also imaged. A linear
stage was used to gently press the layer for better contact of the cells with the chip. The transducer was operated at a frequency of
1.853 GHz, and the imaging was performed at an acquisition rate of 6.5 fps. The background air echo was subtracted from the
acquired image to reduce baseline noise, resulting in each scan being faster than 160ms.
The ultrasonic image of the biopsy slice shows areas with higher reflected ultrasonic echoes separating fat tissues from muscles
(Figure 2A). As fat cells have lower acoustic impedance than muscles, they absorb less ultrasound, causing higher reflected echo
magnitude to be received by the transducers [5]. The ultrasonic images of onion cells also reveal clear contrast between the cell
walls and individual cells, which appear as compartments (Figure 2B). Similarly, the length and width of individual cells extracted
from ultrasonic images (n=18) were 338 ±33 μm and 85 ±12 μm, respectively. For the same cells, the dimensions obtained from
optical images were 324 ±21 μm and 74 ±6μm (paired t-test p-value >0.05 for optical vs. ultrasonic), which shows minimal
variation in spatial data measurements between the two imaging modalities.
Fig. 1. A) The imager unit with the silicon surface as the sensing region. B) Schematic showing pressing of thin sample (tissue slice, cell layers) against the
silicon surface for ultrasonic imaging.
Microscopy and Microanalysis,29 (Suppl 1), 2023, 1116–1117
https://doi.org/10.1093/micmic/ozad067.572
Proceedings
Downloaded from https://academic.oup.com/mam/article/29/Supplement_1/1116/7228434 by Cornell University Library user on 29 August 2023
Fig. 2. A) Ultrasonic image of tissue slice consisting of fat and muscle layers. The fat layers have greater reflected echo compared to muscle cells due to
lower ultrasound absorption. B) Ultrasonic and optical image of individual onion cells.
References
1. A Carovac, F Smajlovic and D Junuzovic, Acta Informatica Medica 19(3) (2011), p. 168.
2. E Fabiszewska et al.,Polish Journal of Radiology 82 (2017), p. 773. https://doi.org/10.12659/PJR.904135
3. A Baskota, J Kuo and A Lal, Microscopy and Microanalysis 28(S1) (2022), p. 1594. https://doi:10.1017/S1431927622006389
4. J Kuo et al., 2021 IEEE International Ultrasonics Symposium (IUS). https://doi.org/10.1109/ius52206.2021.9593762
5. DR Wagner, Journal of Obesity (2013), p. 1. https://doi.org/10.1155/2013/280713
Microscopy and Microanalysis, 29 (Suppl 1), 2023 1117
Downloaded from https://academic.oup.com/mam/article/29/Supplement_1/1116/7228434 by Cornell University Library user on 29 August 2023