Illustration of a future wearable surface patch that would use the new technique for neural imaging. [Credit: Carnegie Mellon University College of Engineering]
While endoscopic imaging has become incredibly flexible and miniaturized, inserting a camera into the human body is still invasive and comes with risk—particularly when you’re trying to image complex and delicate vital organs, such as the brain.
Researchers at Carnegie Mellon University, USA, have introduced what they bill as a non-invasive alternative to endoscopy (Light Sci. Appl., DOI: 10.1038/s41377-019-0173-7). Their technique—thus far only tested out in the lab on nonliving tissue analogs—would employ ultrasound technology to sculpt a virtual relay lens in biological tissue. That virtual lens, in turn, would focus light within the otherwise scattering tissue and allow optical imaging of organs using light sources from outside of the body.
The turbidity roadblock
Light-based imaging is a huge boon to medicine; exposure to visible, ultraviolet and infrared light, is safer, for example, than X-ray imaging. However, biological tissue is a turbid medium—a mixed bag of biological particles that scatter most light, especially in the visible range of the spectrum. That presents a hurdle for optical imaging, and has historically limited the depth and resolution of optical imaging techniques within the human body, especially as increasing the light intensity could lead to photo-thermal damage of the tissue as well as an uptick in background noise.
For deep-tissue imaging, doctors currently rely on surgically implanted camera probes, or endoscopes, to get a look at what’s really going on underneath the surface. Such micro-endoscopes depend on a graded-index (GRIN) lens to relay the image to the clinician. Since endoscopes are implantable, they can maintain the optimal intensity and resolution at the desired depth, but they have limited mobility and can even cause inflammation when repositioned in the body.
While this physical GRIN lens setup gives endoscopy a leg up for optical imaging in deep tissue, the Carnegie Mellon team—comprising Assistant Professor of Electrical and Computer Engineering Maysam Chamanzar and Ph.D. student Matteo Giuseppe Scopelliti—believes it’s found a way to create a “virtual GRIN lens” inside of the body, bypassing the invasive scope altogether. How? By turning the turbidity of biological tissue from an enemy to an ally.
Sculpting a virtual lens
Ultrasound imaging works by generating in situ images using sound waves. As those waves travel through a medium, they can compress as well as rarefy, or thin, the medium. Because light travels more slowly through compressed regions than it does in rarefied regions, the researchers realized that they could use this effect of ultrasound to control the speed of light as it travels through the body, inducing transparency in a target region and allowing more light to penetrate—essentially sculpting a virtual lens.
According to the team, this method is distinct from other acousto-optic approaches in that the light is modulated by the biological tissue itself. Thus the photons are confined and relayed from the depth of the target medium. Compared with the limited mobility of endoscopes, the virtual lens developed by Chamanzar's lab can be ultrasonically sculpted and dynamically reconfigured to focus light within the tissue and image different regions at different depths non-invasively.
In the setup, the team launches ultrasonic waves from a piezoelectric transducer, generating pressure waves into the medium being imaged. As the pressure waves propagate through the medium, they modulate its density and thus its local refractive index, creating a refractive-index profile that varies along the radial direction. The high-pressure, high-density regions are compressed and the negative-pressure, low-density areas are rarefied, creating a graded refractive-index profile. This process in turn builds a modulated phase front of light, focusing the beam through the ultrasound-created virtual GRIN lens.
Phantom tissue demonstrations
In demonstrations, Chamanzar and Scopelliti built a setup in which a 650-nm laser modulated at an ultrasonic frequency (832 kHz) passed through a cylindrical piezoelectric transducer in a “phantom tissue” medium (optical thickness = 5.7 MFP). The researchers immersed a fluorescent target 29-mm-deep into the medium. Driving the ultrasound at 832 kHz with an amplitude of 46.3 V, the pressure waves generated the desired refractive index profile in the ultrasonic cavity, focusing the beam of light in the medium and effectively relaying the image through the medium and resolving it with an external microscope.
Chamanzar and Scopelliti were able to tune the quality of the image as well as the depth of penetration in these experiments by varying the ultrasound intensity, frequency and pattern.
In biomedicine, the researchers envision applying this technique to handheld ultrasound devices or wearable patches on the skin. They believe that the acousto-optic imaging technique could be especially helpful for monitoring neural activity and diagnosing skin disease and malignant tumors. The technique could also be a boon to clinical research—by using the ultrasonic device to monitor brain activity in rats, for example, researchers could potentially gain valuable insights into diseases such as Parkinson’s or epilepsy.
Beyond biomedical applications, the technique could also be used for optical imaging in machine vision, metrology and various other industrial applications.