Wireless Acoustic-Surface Actuators for Miniaturized Endoscopes.

Endoscopy enables minimally invasive procedures in many medical fields, such as urology. However, current endoscopes are normally cable-driven, which limits their dexterity and makes them hard to miniaturize. Indeed, current urological endoscopes have an outer diameter of about 3 mm and still only possess one bending degree-of-freedom. In this article, we report a novel wireless actuation mechanism that increases the dexterity and that permits the miniaturization of a urological endoscope. The novel actuator consists of thin active surfaces that can be readily attached to any device and are wirelessly powered by ultrasound. The surfaces consist of two-dimensional arrays of microbubbles, which oscillate under ultrasound excitation and thereby generate an acoustic streaming force. Bubbles of different sizes are addressed by their unique resonance frequency, thus multiple degrees-of-freedom can readily be incorporated. Two active miniaturized devices (with a side length of around 1 mm) are demonstrated: a miniaturized mechanical arm that realizes two degrees-of-freedom, and a flexible endoscope prototype equipped with a camera at the tip. With the flexible endoscope, an active endoscopic examination is successfully performed in a rabbit bladder. The results show the potential medical applicability of surface actuators wirelessly powered by ultrasound penetrating through biological tissues.

Introduction

Flexible endoscopy is an important diagnostic tool in many medical fields including urology. Its major applications include the detection of hematuria, the diagnosis of bladder carcinoma and other pathologies, as well as the visualization of the upper urinary tract. Bladder cancer is the most common malignancy of the urinary tract,[1] with annually 74 000 new cases in the United States and an estimated 16 000 deaths.[2] Cystoscopic controls are necessary not only for the initial diagnosis of bladder cancer but also for follow-up examinations. Patients afflicted with bladder cancer undergo mostly repeated endoscopies of the urinary bladder using flexible devices. Compared to rigid instruments, the tip of the flexible endoscope benefits the procedure due to larger coverage areas and reduced trauma. Unlike larger flexible endoscopes, such as colonoscopes, current flexible thin endoscopes (overall diameter ∼3 mm), as are for instance used in urology, only have one bending section near the tip. Apart from the overall movements of the entire endoscope (push/pull and rotate by the operator), only one bending degree-of-freedom (DoF) can be realized in the tip itself. As they are cable-driven, they are hard to be miniaturized, for example, with a sheath diameter of 8.7 Fr. (2.9 mm diameter).[3] Furthermore, two cables are required to drive one rotational DoF and adding more DoFs or more bending sections will result in an even larger device diameter, as is the case for some colonoscopes,[4] which are then so large that they will no longer be suitable for urological endoscopic applications.

These disadvantages motivate the following technical developments: (1) Miniaturization, to reduce the diameter of the endoscope; (2) Multiple DoFs, to achieve two bending degrees-of-freedom of the endoscope tip and more bending sections; (3) Wireless actuation, to remove the constraint of the cable. The overall goal is to achieve more space coverage with minimal damage to the human tissue and minimal auxiliary medical measures, e.g., avoid general anesthesia. The ultimate solution is a wireless microrobot that can actively move through a lumen or swim through biological fluids[5,6] to perform medical inspection or treatment. In the quest to realize active endoscopes, different actuation schemes have been investigated, such as shape memory alloy actuators,[7] sliding concentric tubes,[8,9] and pneumatic/hydraulic actuators.[10] These tethered approaches offer large actuation force and precise control of the position; however, their structures are often complicated and they are rigid and difficult to be miniaturized. Wireless approaches, such as magnetics, have been also applied, e.g., to steer magnetic tips of catheters[11] or untethered capsule endoscopes.[12−14] However, large magnetic fields and/or gradients might be required to achieve the forces needed for operation, which results in the need for cumbersome and expensive dedicated equipment. It is necessary to develop a new kind of actuator that has a simple structure and is powered wirelessly and that therefore can be miniaturized and integrated more efficiently.

Ultrasound has been widely used in urology for imaging, tissue ablation, lithotripsy, and drug delivery, and it has also been proposed as a means to transfer power wirelessly.[15,16] It has recently been shown that ultrasound can also be shaped to form complex three-dimensional patterns and this provides additional opportunities for medical ultrasound.[17] However, the common approaches convert the mechanical vibrations to electricity using the piezoelectric effect and then utilize the electricity to power a device. This conversion process adds additional complexity and suffers from low efficiency. Recently, we developed an actuator that directly converts ultrasound power into mechanical work via acoustic streaming from a surface array of microbubbles.[18,19] Acoustic streaming is a steady fluid flow driven by acoustically driven oscillations, which can be achieved at the transducer surface or at the air–liquid interface.[20,21] It is well known that oscillating bubbles in fluids can cause resonant oscillations and thus acoustic streaming[22] and microfluidic pumps and mixers have been developed on the basis of this phenomenon.[23] It was shown by Dijkink et al. that oscillating bubbles in a tube can propel or power a rotary motor on the centimeter scale,[24] and microswimmers propelled by one or a few bubbles have been reported by several groups.[25,26] However, the powering of a miniaturized medical instrument based on acoustic active surfaces has not yet been reported.

In this article, we report the application of our wireless acoustic-surface actuators to a miniaturized, multi-DoF, ultrasound-powered endoscope prototype. As illustrated in the schematic of Figure , the device is actuated by acoustic streaming, which is generated from an active surface consisting of a two-dimensional (2D) array of microbubbles. Bubbles of different sizes are addressed with unique ultrasound frequency; thus, multiple DoFs are achieved. Two miniaturized devices with side length around 1 mm were developed, i.e., a two DoFs miniaturized mechanical arm (mini arm) and a flexible endoscope equipped with a camera. The endoscope is inserted into a rabbit bladder through the tool channel of a rigid cystoscope, and is successfully used for a cystoscopic examination.

Figure 1

Schematic of the wireless actuated miniaturized cystoscope. (a) The miniaturized cystoscope performs bladder imaging with two rotational DoFs and one translational DoF. (b) An enlarged view of the cystoscope tip. The active surface, which is powered wirelessly by ultrasound, rotates the first section relative to the second section. (c) An enlarged view of the active surface, which consists of an array of microbubbles. The oscillation of the interface between air and liquid causes liquid streaming, generates a propulsion force in the opposite direction and actuates the device.

Methods

Fabrication of the Active Surface

The active surface consists of a regular array of cylindrical microcavities with a carefully chosen size of the cavity. Photolithography was used to fabricate the surface as reported previously.[18] In short, negative photoresist (SU-8 2000, MicroChem, MA) was spin-coated on a silicon wafer (WS-650 spin coater, Laurel, PA), exposed on a mask aligner (MJB4, SUSS MicroTec, Garching, Germany) with a chromium mask (Delta Mask, Enschede, Netherlands), developed, and hard-baked. Then, the wafer was diced by an automatic dicing saw (DAD 321, DISCO, Tokyo, Japan) to a rectangular array.

When a dried active surface is immersed under water, air is trapped in the cavities as bubbles due to the small size of the cavity and the hydrophobicity of the SU-8 photoresist (∼90° contact angle[27]). The diameter of the bubble is the same as that of the cavity and the length is normally shorter than that of the cavity, as shown in the inset of Figure . The size of the bubble determines the resonance frequency of a single bubble. This corresponds to the frequency at which the maximal acoustic streaming of the active surface occurs, as reported in our previous study.[18] Sodium dodecyl sulfate (SDS, 2.4 g/L) is added to the solution as a surfactant to lower the surface tension of water from 73 to 38 mN/m.[28] Our previous study shows both theoretically and experimentally that reducing the surface tension increases the oscillation amplitude of the air–liquid interface, which substantially increases the acoustic streaming force (up to ∼3.5 times).[28]

Figure 2

Wirelessly actuated endoscope powered by ultrasonic active surface. (a) A picture comparing a commercial flexible urological endoscope 2.9 mm in diameter (left), with the mini arm (middle) and flexible endoscope prototype (right) developed in this work. (b) An enlarged microscopic image of the red-dashed box in (a) shows the active surface under water, which consists of an array of microbubbles with controllable sizes. The inset is a zoom-in of the bubble in the white dashed box, which is trapped in a cavity 30 μm in diameter and 120 μm in height.

Assembly of the Miniaturized Endoscope

The mini arm has a dimension of 1 mm × 1 mm in cross section and is 5 mm long. It contains three sections with two hinges, which are mounted orthogonal to each other (Figure a middle) such that the arm can achieve two DoF. The sections and the pins were made out of aluminum and brass, respectively. After assembly of the mini arm, two active surfaces (1 mm × 4 mm) with different microcavity sizes, i.e., actuator 1: an array of 9600 “small” cavities, each 55 μm in depth, 10 μm in diameter, 10 μm in spacing; actuator 2: an array of 1520 “large” cavities, each 120 μm in depth, 30 μm in diameter, 20 μm in spacing, were attached to two orthogonal surfaces of the first section using cyanoacrylate (Superflex Gel, UHU, Bühl, Germany).

The flexible arm was assembled and integrated with a Naneye 2D CMOS sensor (Awaiba, Yverdon-Les-Bains, Switzerland), which is 1 mm × 1 mm in cross section and 1.85 mm long. The flat wire connected to the Naneye is a flexible ribbon of four wires with a total dimension of 0.72 mm width and 0.188 mm height, which also serves as an elastic sheet to provide the restoring force in the flexible arm design. An active surface (1 mm × 9 mm, with “large” cavities) is attached to one side of the flat wire using cyanoacrylate.

Ultrasound Excitation Setup

A customized ultrasound excitation setup was made to test the endoscope. It is a glass tank of 50 × 50 × 50 mm3, with one piezoelectric ceramic disk (Ø50 mm × 3 mm, PIC 181, PI Ceramic, Lederhose, Germany) attached to one side wall by epoxy glue. To increase the propulsive force and stabilize the bubbles, 2.4 g/L SDS surfactant was added in all fluids.[28] The driving signal was a sinusoidal wave generated from a function generator (33220A, Agilent technologies, CA), and amplified 30 times with a customized power amplifier and transformer. Two frequencies, i.e., 51.3 and 115.3 kHz, were used to address the arrays of “large” and “small” bubbles, respectively. For the wireless actuation of the endoscope in the rabbit bladder, an ultrasonic bath (2510E-DTH, Bransonic, CT) was used for higher power ultrasound of 100 W at 42 kHz ± 6%.

Endoscope Actuation in the Rabbit Bladder

The cystoscopy with the ultrasound-powered endoscope is demonstrated in an isolated rabbit bladder (Oryctolagus cuniculusforma domestica). The organs were obtained from a local slaughterhouse. The careful representation of the urinary bladder in the retroperitonal convolute was followed by the preparation including ligation of both ureters and the removal of the bladder including the proximal urethra. To observe the movement of the endoscope inside the bladder, a rigid 8 Fr. cystoscope (Richard Wolf, Knittlingen, Germany) was used. The bladder was fixed to the cystoscope by a 4-0 Vicryl suture (Ethicon, Somerville) and filled with ∼100 mL SDS aqueous solution. The developed flexible endoscope was introduced by a 3 Fr. flexible gripper (Richard Wolf) through the tool channel of the cystoscope into the bladder, as shown in detail in Figure S-1. The illumination and visualization were achieved by a 5132 Auto LP HLight Module and a 5520 1CCD Endocam Module (Richard Wolf) connected to the rigid cystoscope. The video from the cystoscope and the flexible miniaturized endoscope are recorded and shown as Videos in the Supporting Information.

Results and Discussion

Wireless Actuation of the Mini Arm

As shown in Figure , the mini arm consists of three sections connected in series and can achieve two rotational DoF. Two active surfaces with different microcavity sizes are attached to two orthogonal surfaces on the first section. The mini arm is mechanically held under water by tweezers. The mini arm contains no electrical components or connections. Instead, the mini arm is powered completely wirelessly by ultrasound. The ultrasound that is used to excite the arm is generated by a piezoelectric transducer, which can also be seen in Figure , behind the arm at a distance of ∼30 mm.

Figure 3

Wireless actuation of the mini arm. (a) Sketch of the miniaturized arm made of aluminum and assembled through brass hinges. Two actuators 1 and 2 are attached orthogonally to the first section, which actuates the arm with two rotational DoF. (b) Sequential snapshots of a video. At 9–11 s, the first section is actuated with ultrasound at 115.3 kHz, and at 20–24 s, the second section is actuated with ultrasound at 51.3 kHz. See also Video S-1 in the Supporting Information.

Once the mini arm is put under water, air is trapped in the microcavities of the active surfaces as individual bubbles due to the surface hydrophobicity and small diameter of the cavities (see Methods). The size of the microbubbles is determined by the cavity size, which has a unique different acoustic resonant frequency. A detailed physical study of the frequency response of the bubbles is reported in a previous work.[18] In short, the resonant frequencies of micron-sized bubbles lie in the range of 50–200 kHz and the smaller the bubble size the higher is its resonant frequency. For the actuation of the mini arm, an ultrasonic field of 115.3 kHz was first applied and it only excited the “small” bubbles on the actuator 1, which caused ultrasonic streaming and thus propelled the first section into rotating to the left (image at 11 s in Figure b). When the ultrasound was switched off, the arm returned to the original position due to gravity. Similarly, when an ultrasound of 51.3 kHz was applied, only actuator 2 was activated; thus, the second section of the mini arm rotated out of the plane (image at 24 s in Figure b). See also Video S-1 in the Supporting Information for the real-time movement of the mini arm. The reproducibility is demonstrated by switching each frequency on and off two times (see Video).

Increasing the DoF in the tendon-driven endoscope requires adding at least two more driving wires; thus, the endoscope structure is more rigid and complicated and the diameter is larger. The wireless-driven active surface reported here solves this problem. The bandwidth response of each kind of actuator is about 10 kHz, thus in principle 18 independent actuators can be achieved within the 20–200 kHz frequency range. Whereas for cystoscopy three DoFs (two rotational DoFs ω1, ω2, and one translational DoF, see Figure ) are normally enough; for more complicated intraluminal anatomy, such as the upper urinary tract, more DoFs will be beneficial. If the hinges are arranged in orthogonal directions, three arm sections will realize three independent rotational DoFs. Because the back and forth motions of one DoF requires two actuators, six actuators in total will be needed. The redundant DoFs could be useful for applications other than bending the endoscope, for example, to open and close a gripper. More importantly, as all actuators are driven wirelessly, no additional wire is required for more DoFs, and the overall diameter of the endoscope can remain very small. It is noteworthy that each active surface can be 100–200 μm thin; therefore, equipping the endoscope tip with two DoFs (four active surfaces) will increase its diameter by much less than half a millimeter. Moreover, even more active surfaces can be added at different positions along the articulated tip without making the diameter any larger. Notably, the streaming actuation force is independent of the direction of the incoming ultrasound, which can be applied from the most convenient position. It is also interesting to point out that the adoption of active surfaces is a scalable approach. The mini arm is designed to be 1 mm × 1 mm in cross section, because the smallest commercial camera has a cross section of 1 mm × 1 mm. In other words, 1 mm is far from the size limit for the active surface and the principle illustrated here is still valid for even smaller devices in the micrometer range.

Wireless Actuation of the Flexible Endoscope

Although the actuation scheme is wireless, there is currently no wireless camera small enough to fit through the tool channel of a cystoscope. Therefore a tethered camera such as the Naneye 2D sensor was used. The design of the flexible arm takes advantage of the connecting wires for the camera and adopts it as an elastic sheet to provide the restoring force when the ultrasonic transducer is turned off. Figure a,b show the movement of the flexible endoscope, which consists of three parts: a miniaturized camera at the tip pointing forward, a flat wire connecting the camera and also acting as a recoil spring, and an active surface. The effective bending length of the wire is 22.5 mm, and the streaming active surface is attached next to the camera causing the deflection in one direction. The deflection is driven by an ultrasound transducer ∼30 mm away, and is controlled by the driving voltage which shows a linear relationship to the output power in this range (Figure c). The total length of the flexible arm is about 33 mm, which is realistic for cystoscopy in a human bladder. A normal human bladder capacity is ∼400 mL,[29] so considered as a sphere, a filled human bladder has an average radius of approximately 46 mm. This also means the miniaturized camera can be quite close to the bladder inner surface, thus providing a clearer image of the bladder wall. When the wire is thinner or softer, or the ultrasound power is increased, it is possible to realize even larger deflections.

Figure 4

Wireless actuation of the flexible endoscope prototype. (a) Sketch of the flexible arm. The active surface is attached to the flat wire, which serves as an elastic sheet to provide the restoring force. (b) A superimposed image showing the deflection of the flexible arm is controlled by the driving voltage of the ultrasonic transducer. See also Video S-2 in the Supporting Information. (c) The deflection exhibits a linear relationship with the driving voltage, and reaches a maximum of 60° in the current setup.

The flexible arm has a simpler design compared with the above mentioned mini arm, i.e., no mechanical hinges are required and it can be fabricated more easily. There are ∼3400 bubbles on an area of 1 mm × 9 mm surface, each generating a force up to ∼120 nN (bubbles 120 μm in depth, 30 μm in diameter).[30] Therefore, the estimated output force generated by the active surface is ∼0.4 mN. The force is smaller than those typically generated by tethered approaches, but large enough to effectively steer a miniaturized camera, as demonstrated in the experiments. As the force is small and the flexible endoscope prototype is compliant, the actuator–camera setup should be safe for the urinary tract wall inspection and cause no damage to the tissue.[31] Furthermore, such systems might be potentially introduced via urologic catheters such as ureteric catheters replacing conventional endoscopic inspection of the urinary tract by rigid, semirigid and flexible devices.

Safety and Stability of the Ultrasonic Streaming Actuation

Safety is one of the most important concerns of medical devices, especially for invasive procedures. The safety of the ultrasound-powered acoustic streaming is discussed here. It is difficult to compare the values directly with current ultrasonic medical devices because in diagnostic ultrasound, a higher frequency (2 MHz) coupled with pulse mode operation is used. The ultrasound used here is continuous wave with a frequency between 50 and 200 kHz. Two parameters are often used to represent the ultrasound intensity, i.e., the spatial peak-temporal average intensity ISPTA and the spatial peak-pulse average intensity ISPPA. A typical pressure amplitude in a B-scan is 1.68 MPa,[32] and the recommended intensity limits set by the U.S. Food and Drug Administration (FDA) for use in a peripheral vessel are 720 mW/cm2 for ISPTA and 190 W/cm2 for ISPPA.[33] A continuous wave means the duty cycle is 100%; therefore, ISPTA = ISPPA = ISP in this case. Currently, there is no safety guideline for continuous ultrasound, however, a previous study shows that the tissue injury intensity threshold of animal tissues for continuous ultrasound is 470 W/cm2.[34]

Measured by a needle hydrophone, the pressure amplitudes of the ultrasound used to power the mini arm and the flexible endoscope are both around 100–200 kPa, which is much lower than 1.68 MPa in the B-scan. The corresponding ultrasound intensity can be calculated by the following equation[35]where ISP is the spatial-peak ultrasound intensity, P is the pressure amplitude, ρ is the density of water (1000 kg/m3), and c is the speed of sound in water, which is ∼1520 m/s at 37 °C. If we take the highest pressure P = 200 kPa, then ISP = 2.6 W/cm2. This value is higher than the FDA recommended value for ISPTA 720 mW/cm2, which may lead to some thermal effects;[32] but 2 orders of magnitude lower than the FDA recommended value for ISPPA 190 W/cm2 and the reported tissue injury intensity threshold 470 W/cm2, so it is low enough that it is unlikely that the ultrasound used here will cause lasting mechanical damage or induce cavitation.[32] However, further studies on animal experiments are required to establish more carefully if and when one may expect thermal effects of continuous wave ultrasound on tissue.

Another issue is the stability of the air bubbles. It is observed in the experiments that the air bubbles trapped in the active surface are stable under ultrasonic excitation for at least 30 min, and the streaming actuation can be excited repeatedly during this period. Under the microscope, we observed that the air bubbles will gradually dissolve in water and the bubble size will become smaller, thus the driving frequency needs to be adjusted and be increased accordingly. But the dissolution process is relatively slow compared to the operating time of a cystoscope (depending on the examiner normally 5–15 min for an examination). The dimension of the bubble only changes by about 0.8%/min (see Figure S-3 in the Supporting Information). The dimension change is slow enough that no frequency adjustments are needed during the typical operation time of the active surface (∼5–15 min for a cystoscopy examination). Changing air to other inert gases that are insoluble in water, such as SF6, a commonly used contrast agent for ultrasound imaging,[36] will result in even longer operating time. SDS is added to the solution as a surfactant to lower the surface tension, increase the oscillation amplitude of the air–liquid interface, and thus increase the acoustic streaming force (up to ∼3.5 times).[28] SDS was chosen as it is readily available in the laboratory, but it is not supposed to be used in clinical applications. Many biocompatible surfactants, such as amino acid-based surfactants,[37] can instead be used as safe substitutes with comparable effects on the acoustic actuation. Particular attention to safety and biocompatibility will be given in the future translational studies.

Miniaturized Endoscope for Cystoscopy in a Rabbit Bladder

The flexible endoscope was tested in a rabbit bladder ex vivo. The rigid cystoscope is used to hold and refill the bladder (Figure ), and visualize the movement of the flexible endoscope inside the bladder. However, in a real application the cystoscope is not necessary, as the flexible endoscope can be inserted into the bladder through the urethra alone (as illustrated in Figure ) or through a passive catheter. In clinical practice, cystoscopy often starts with an empty bladder without urine; then, to obtain a clear view of the bladder wall, the operator fills the bladder with saline solution through the irrigation channel of the cystoscope. As there is no irrigation channel in our miniaturized endoscope prototype due to its small size, another thin catheter could be used to fill the bladder with saline solution before the examination by the acoustically actuated endoscope, which requires an aqueous environment to operate.

Figure 5

Setup for cystoscopy in rabbit bladder. (a) A schematic shows the flexible arm entering the bladder through the tool channel of a rigid cystoscope, which is fixed to urethra. The bending of the flexible arm is powered by the streaming of the active surface, and its movement is viewed via the cystoscope. (b) A picture of the bladder fixed with the endoscope. (c) An enlarged picture of the flexible arm coming out of the tool channel of a cystoscope.

The miniaturized endoscope entered the bladder through the working channel of the endoscope; at first it was stationary pointing downward (image at 0 s in Figure a). When the ultrasound was applied, streaming from the active surface occurred and the flexible arm or the endoscope is bent to the right (3 s in Figure a). The streamline is weakly visualized in the image, but clearer in the Video S-3 in the Supporting Information. When the ultrasound was turned off, the flexible arm returned to the original position due to the restoring force of the flat wire (6 s in Figure a). We found that the bubbles on the active surface were stable during the whole period of examination (∼10 min) and contact with the bladder wall did not release the bubbles. From the miniaturized camera on the tip of the endoscope, another video was recorded (snapshots in Figure b). When the flexible arm was pointing downwards, the structure of the fat tissue was seen as a marker and can be directly compared to the cystoscopic images in Figure a. When the flexible arm is bent, a blood vessel on the bladder side wall can be visualized. This vessel can be seen from the outside of the bladder in Figure b, but because of its position it could not be visualized with the cystoscope in Figure a. In the Video S-3 in the Supporting Information, the two videos from the cystoscope (left) and miniaturized camera (right) are synchronized and to demonstrate the reproducibility of the imaging procedure the ultrasound is switched on and off two times. In the video, a certain level of shake of the camera is observed, which can be a problem in endoscopy. The reason is that the ultrasound power is not very stable in the experiment and the device is operated in an open-loop control. The image stability can be improved by correcting for fluctuations in ultrasound power, e.g., with feedback from an in situ pressure sensor, or by sensing the bending angle of the endoscope. The miniaturized camera has a high frame rate over 60 frames/s, which generates redundant frames; thus, the video can be digitally processed with antishaking algorithms to minimize the shaking effect. The current image quality is still not comparable to the current endoscope, but with the fast development of microelectronic devices, we believe, the imaging quality of the miniaturized camera can be further improved and the miniaturized wireless device will be the trend in the future endoscopy.

Figure 6

Cystoscopy in a rabbit bladder. (a) Sequential snapshots of a video recorded by the rigid cystoscope. At 0 s, the flexible arm is stationary and pointing downward. At 3 s, ultrasound is turned on and the arm is bent to the right due to the streaming of the active surface. At 6 s, ultrasound is turned off, the flexible arm returns to the original position. (b) Sequential snapshots of a video recorded by the Naneye camera on the flexible arm. At 0 s, the flexible arm is pointing downwards, and the shape of the fat tissue is seen as a marker comparing with the cystoscopic images above. At 3 s, when the flexible arm bends, a blood vessel on the bladder side wall is visualized, which can be seen from outside the bladder in Figure b but cannot be seen from the cystoscopic images in (a). See also Video S-3 in the Supporting Information.

Although because of smaller dimensions, a rabbit’s bladder was studied as a proof-of-concept, this work reports a general method to actuate a miniaturized endoscope in low viscosity bodily fluids. The miniaturization of the instrument leads to a significant reduction of surgical trauma and thus reducing unwanted complications for the patient. It not only can be used as a cystoscope, but may also be applied to the upper urinary tract and the kidney. Alternatively, the flexible endoscope can be tested in a realistic phantom of the kidney and urinary tract.[38] It can potentially simplify preparatory steps of surgery procedures; for example, a double-J catheter might be required to dilate the ureter for a period of time prior to the surgical manipulation on the upper urinary tract via an endoscope. With the actuation scheme developed herein, an instrument with less than 1 mm diameter could potentially avoid unnecessary interventions. Consequently the severeness of urological endoscopic surgeries could be decreased, resulting in a shortened hospital stay of the patients and thereby contribute to a cost reduction.

Conclusions

In this article, we report the first thin-layer surface actuators that permit the wireless power and control of an endoscope. The wireless actuation is based on two-dimensional arrays of microbubbles that resonate with an external ultrasound, giving rise to frequency-selective, directional acoustic streaming. The endoscope has an overall side dimension of 1 mm and fits through the tool channel of a cystoscope. A complete cystoscopy was successfully performed by this endoscope in a rabbit bladder. The novel actuation scheme is general and can therefore be further miniaturized. The method demonstrated herein thus holds potential as a powerful actuator to steer miniaturized devices for minimally invasive in vivo inspections.