Compact laser-controlled end effector

Abstract

A compact laser manipulation end actuator comprises a frame, at least two active mirrors and a pair of actuators. The frame has a maximum dimension in a plane orthogonal to a longitudinal axis of no more than 13 mm, and the mirrors are mounted near the distal end of the frame. The actuators are mounted to the frame and are configured to respectively change the tilt of the active mirrors relative to the frame. A path is provided through the frame to pass a laser beam to the mirrors, and wherein the mirrors are positioned via the actuators and are configurable to reflect a laser beam from each mirror out to an external target.

Description

[0001] background

[0002] The discussion of the state of the art discussed below may reflect hindsight gained from the disclosed invention; and these features are not necessarily admitted to be prior art.

[0003] Pharyngeal or laryngeal cancer is notable for its detrimental effects on quality of life.

[0004] Therefore, there is a great need for curative treatments for these cancers that allow for normal organ function after treatment. Open surgical removal of part or all of the organ represents one extreme, almost guaranteeing a good oncological outcome at the expense of functional outcome. Nonsurgical approaches (including induction chemotherapy, intensity-modulated radiation therapy, and chemotherapy combined with radiation therapy) offer better functional outcomes than open surgery while achieving good oncological outcomes. This technique is currently recommended for most laryngeal cancers. However, over the past two decades, the long-term morbidity and functional impairment associated with radiation therapy have led otolaryngologists to re-examine surgical techniques. Minimally invasive approaches have particular interest in providing greater preservation of healthy tissue and associated improved postoperative organ function compared with open surgery.

[0005] Transoral laser microsurgery (TLM) is currently the most established minimally invasive surgical method. In TLM, a carbon dioxide laser is coupled to a surgical microscope and aimed through a laryngoscope, which provides a line-of-sight to the upper airway. The surgeon uses manual tools inserted through the laryngoscope to retract the lesion and uses micromanipulators on the microscope to steer the laser beam to remove the diseased tissue. In advanced systems, the micromanipulators are motorized, allowing the surgeon to guide the laser along predetermined arcs and lines at precisely controlled speeds, resulting in very high-quality incisions. The biggest limitation of TLM is its line-of-sight restriction, which limits visualization and exposure. In addition, the narrow and long laryngoscope makes it difficult to manually manipulate the tools required for retraction and suturing.

[0006] To address these shortcomings, transoral robotic surgery (TORS) was developed in the mid-2000s. In TORS, a flexible or wristed robotic manipulator is used in conjunction with an endoscopic vision system to provide superior visualization and tissue manipulability at the surgical site. The DAVINCI multi-port and DA VINCI single-port systems (from Intuitive Surgical Inc, Sunnyvale, CA) and the FLEX robotic system (from Medrobotics, Raynham, MA, USA) are currently in clinical use.

[0007] However, TORS still falls short compared to TLM in one key respect; in TORS, excision is primarily performed using electrocautery, which results in increased postoperative pain and longer recovery times than similar procedures performed with TLM. The use of lasers in TORS is expected to increase with the recent development of hollow-core fibers capable of delivering CO2 lasers through flexible instruments, which should alleviate this problem to some extent. However, fiber-based lasers are held and manipulated by robotic tools and therefore lack the spatial repeatability, precision, and speed of free-beam scanning systems. This means that incision quality suffers relative to free-beam systems used for TLM. Free-beam systems have the added benefit of leaving the surgical site clear for exposure and visualization of the margins.

[0008] Many prototype endoscopic devices have been developed for minimally invasive laser scanning tasks, each with a different approach to solving challenging design problems. In "Endoscopic laser scalpel for head and neck cancer surgery" by S. Patel et al. in Photonic Therapeutics and Diagnostics VIII, Vol. 8207, International Society for Optics and Photonics 82071S (2012), a DC motor-driven Risley prism was used as a beam steering method, which made the control scheme easy to implement, but the scanning speed was very limited (2Hz). In "A5-mm piezo-scanning fiber device for high speed ultrafast laser microsurgery" by O. Ferhanoglu et al. in Biomedical Optics Express, Vol. 5, No. 7, 2023-2036 (2014), a piezoelectric actuator was used to directly bend the optical fiber to achieve a large bandwidth (1kHz) in a thin device profile (5mm diameter); but the field of view (200×200μm) was affected. In A. Acemoglu et al., “Design and control of a magnetic laserscanner for endoscopic microsurgeries” in IEEE / ASME Transactions on Mechatronics (2019), electromagnets were used to bend the fiber, which increased the range of motion to 3 mm × 3 mm, but sacrificed bandwidth (15 Hz) in doing so. Finally, as described in R. Renevier et al., “Endoscopic lasersurgery: design, modeling and control” in IEEE / ASME Transactions on Mechatronics, Vol. 22, No. 1, 99-106 (2017), the FEMTO-ST Institute developed a system that uses off-the-shelf piezoelectric linear actuators to articulate a silicon mirror mounted on a tip-tilt stage, achieving a 20 mm × 20 mm field of view despite a lower scanning speed (17 Hz).

[0009] Overview

[0010] A compact laser manipulation end effector and a method of using the end effector, particularly for surgery, are described herein, wherein various embodiments of the device and method may include some or all of the elements, features, and steps described below.

[0011] A compact laser manipulation end actuator comprises a frame, at least two active mirrors and a pair of actuators. The frame has a maximum dimension in a plane orthogonal to the longitudinal axis of no more than 13 mm, and the mirrors are mounted near the distal end of the frame (i.e., closer to the distal end than to the proximal end of the frame). The actuators are mounted to the frame and are configured to respectively change the tilt of the active mirrors relative to the frame. A path is provided through the frame to pass a laser beam to the mirrors, and wherein the mirrors are positioned and configurable via the actuators to reflect a laser beam from each mirror toward an external target.

[0012] In a method for robotic laser manipulation, a laser beam is generated and directed from a proximal end of an end effector to an area near a distal end of the end effector. The laser beam is reflected in an area near the distal end of an end actuator having at least two active mirrors, and an actuator is used to change the tilt orientation of the at least two active mirrors to shift the laser beam through a target.

[0013] The apparatus and methods discussed herein can capture the advantages of both TLM and TORS in a laser scanning system operated at the end of a robotic manipulator. The apparatus and methods enable more precise incisions and, correspondingly, less tissue damage than existing laser or electrocautery tools. Furthermore, moving the laser fiber into the robotic manipulator can relax some of the constraints on fiber design and use, which is advantageous considering that currently available CO2-laser fibers are expensive (approximately $1000) and are disposable. Below, we describe a prototype device that implements this TLM / TORS hybrid paradigm.

[0014] Minimally invasive surgical approaches have recently emerged that allow for greater preservation of healthy tissue, but they remain limited in important ways. Here, we describe a device that combines the advantages of two of the main existing minimally invasive approaches: (a) the high-quality incisions and reduced postoperative pain achievable with transoral laser microsurgery and (b) the superior visualization and tissue manipulability afforded by transoral robotic surgery. In the specific example, the 11-mm-diameter device is connected to a fiber laser source and directs a focused laser beam across an 18-x-10-mm plane in a controllable trajectory at speeds of up to 7 m / s.

[0015] By combining the strengths of TORS and TLM, this approach holds promise for improving the quality of care, for example, in pharyngeal and laryngeal cancers. It allows for better access and exposure than TLM alone, and higher incision quality than TORS alone, with the end result being a simpler procedure for the surgeon and better functional outcomes for the patient. Among other applications, this approach could be used to perform laser-assisted cardiac ablation, laser-assisted gastrointestinal surgery, laser-assisted extra-abdominal surgery, or laser-assisted transnasal skull base surgery.

[0016] The end effectors described herein can be precisely actuated at high speeds, enabling important new ways for robots to interact with living tissue. In particular, the end effectors can be incorporated into microrobotic devices to provide dexterous control of laser energy during minimally invasive surgery. The high-bandwidth distal actuation enabled by these devices and methods can be integrated with existing surgical tools to precisely control the position of fiber-delivered lasers. The manufacturing and modular assembly methods described herein enable the creation of a device that is significantly smaller and has a higher bandwidth than the current state of the art, while achieving a similar range of motion as existing tools. For example, the device can be no larger than 6 mm in diameter and no larger than 16 mm in length, and can focus and steer a fiber-delivered laser beam over a large range (e.g., + / -10° in two axes) at high speeds (e.g., 1.2 kHz bandwidth) with good static repeatability (e.g., + / -200 μm). This approach allows for better control of energy delivery during surgery than existing rigid and cable-driven tools, as the laser speed as well as the duration of energy application can be precisely controlled over a wide range. This dexterous energy delivery method can be applied to different surgical fields where controllable energy is required in limited space. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] Figure 1 A first example of an end effector 10 of a laser scanning system for transoral robotic surgery is shown. The end effector 10 receives a fiber optic input and manipulates a focal point on a surgical plane.

[0019] Figure 2 and Figure 3 It is a long exposure photo showing the use of Figure 1 Typical cutting profiles used in laser-based tissue resection for end effector design.

[0020] Figure 4 is a perspective view of a second example of a compact laser-manipulated end effector 10 for robotic surgery.

[0021] Figure 5 yes Figure 4 An end view of a compact laser-manipulated end effector 10 for robotic surgery is shown in FIG.

[0022] Figure 6 yes Figure 4 and Figure 5 A side view of a compact laser-manipulated end effector 10 for robotic surgery is shown in FIG.

[0023] Figure 7 yes Figure 4-Figure 6 A perspective view of a compact laser manipulation end effector 10 for robotic surgery is shown in FIG. 1 , with exploded views of the various parts of the end effector 10 .

[0024] Figure 8 is a perspective view of a first example of a laser scanning end effector 10 defining the locations of key subsystems.

[0025] Fig. 9 Shows Figure 8 The optical subsystem 48 of the end effector 10. The optical subsystem 48 connects the fiber optic source to the laser scanner.

[0026] Fig.10 Shows Figure 8 The actuation subsystem 50 of the end actuator 10 includes the piezoelectric bending actuators 18 and 22 and a rigid mechanical frame (base) 40 that aligns the optical element with the rest of the device.

[0027] Fig.11 The mirror system 52 of the end effector 10 is shown, including the motion transmission linkage 38 and the planar active mirror 28 / 30 .

[0028] Fig.12 The slider crank drive motion is shown schematically representing the selected design values ​​shown in the table.

[0029] Fig.13 and Fig.14 A description of a slider crank arm on a first example of a manufactured linkage mechanism for an end effector, wherein Fig.13 A first mirror linkage mechanism 38 is shown, and Fig.14 The second mirror linkage 38 is shown. The first set of solid lines 71 represent the effective link of the crank slider, the dashed line 73 represents the rotation axis of the crank, and the second set of solid lines 75 represent the solid part that is bent back to facilitate assembly of the linkage 38. The joints are represented by black circles 77, and the slider translation and crank rotation are represented by dashed black traces 79.

[0030] Fig.15 is a diagrammatic representation of a laser scanning system defined by key symbols.

[0031] Figure 16-Figure 21 A manufacturing process of a connecting rod mechanism is shown. Fig.16Various layers of rigid material 60 , flexible material 62 , and adhesive material 64 are shown cut using laser micromachining. Fig.17 Layers 60 and 62 are shown aligned on pins 68 and laminated using heat and pressure on a laminate sheet 66. Fig.18 In FIG. 1 , additional laser micromachining demonstrates the structure of the linkage mechanism 38 of the first embodiment of the end effector. Fig.19 In FIG. 1 , after releasing the assembly tab, the linkage 38 is partially deployed. Fig. 20 In the 3D printed assembly jig 70, a 3D printed assembly jig 70 is used to bend the linkage arms to their desired positions. Fig.21 In the embodiment of the present invention, the last tab is released; and the linkage 38 is ready to be integrated with the rest of the device.

[0032] Fig. 22 The measured values ​​of the mirror angles as a function of the actuator voltage for the first mirror 28″ and the second mirror 30″ of the first instance of the end effector are plotted. The theoretical model of the first mirror 28′ and the second mirror 30′ captures the overall trend well, with mean absolute errors of 0.6° and 1.2° for the first mirror and the second mirror, respectively.

[0033] Fig.23 The frequency response of the actuator-actuator-mirror system of the first linkage actuator 18 and the second linkage actuator 22 of the first example of the end effector is plotted.

[0034] Fig.24 A first example laser spot position tracking setup of an end effector as viewed from a high speed camera is shown. The theoretical field of view 72 and the field of view 74 obtained from a standoff distance of 20 mm are shown.

[0035] Fig.25 The first example of using the end effector to draw shapes at different speeds is shown, demonstrating the open-loop trajectory tracking capability of the laser scanner. The high-speed performance of the system is demonstrated by its repeatability over a range of command control speeds.

[0036] Fig.26 Three mirror angle configurations of a three-mirror galvanometer are shown.

[0037] Fig. 27 Two mirror angle configurations of a dual-mirror galvanometer are shown.

[0038] Fig.28 Two additional mirror angle configurations of a dual-mirror galvo are shown.

[0039] Fig.29An example of a first active mirror 28 and its associated actuator 18 and passive transmission linkage 38 of a second example of an end effector is shown.

[0040] Fig.30 An example of a second active mirror 30 and its associated actuator 22 and passive drive linkage 38 of a second example of an end effector is shown.

[0041] Fig.31 Modular components of a second example of a laser steering end effector are shown.

[0042] Fig.32 and Fig.33 Graphs of frequency responses of a first mirror and a second mirror of a second example of an end effector with respect to phase and amplitude, respectively.

[0043] Fig.34 This is an image of a second example of an end effector produced by high-speed motion of a laser manipulation system.

[0044] Fig.35 It shows how the high speed image (produced at 3900 mm / s) closely matches the trajectory produced by the laser system of the second example of the end effector at a lower speed (7.8 mm / s).

[0045] In the drawings, reference characters refer to the same or similar parts throughout the different views. The drawings are not necessarily drawn to scale; instead, emphasis is placed on illustrating certain principles in the examples discussed below. For any drawings that include text (words, reference characters, and / or numbers), an alternative version of the drawing that does not include text should be understood to be part of the present disclosure; and a formal alternative drawing without such text may be substituted.

[0046] Detailed Description

[0047] The foregoing and other features and advantages of various aspects of the present invention will be apparent from the following more specific description of various concepts and specific embodiments within the broader scope of the present invention. Various aspects of the subject matter introduced above and discussed in more detail below can be implemented in any of a number of ways, as the subject matter is not limited to any particular implementation. Examples of specific implementations and applications are provided primarily for illustrative purposes.

[0048] Unless otherwise defined, used or characterized herein, the terms (including technical and scientific terms) used herein should be interpreted as having a meaning consistent with their accepted meaning in the context of the relevant art, and should not be interpreted as an idealized or overly formal meaning unless expressly so defined herein. For example, if a particular concept is referenced, the concept may be substantially (although not completely) pure, as actual and imperfect realities may apply; e.g., the potential presence of at least trace amounts of impurities (e.g., less than 1% or 2%) can be understood to be within the scope of the description. Similarly, if a particular shape is referenced, the shape will include imperfect variations from the ideal shape (e.g., due to manufacturing tolerances). Percentages or concentrations expressed herein may be expressed by weight or volume. The processes, procedures and phenomena described below may occur at ambient pressure (e.g., about 50-120 kPa; e.g., about 90-110 kPa) and temperature (e.g., -20°C to 50°C; e.g., about 10°C to 35°C), unless otherwise specified.

[0049] Although the terms, first, second, third, etc., can be used to describe various elements, these elements are not limited by these terms. These terms are only used to distinguish one element from another. Therefore, without departing from the teachings of the exemplary embodiments, the first element discussed below can be referred to as the second element.

[0050] For ease of description, spatially relative terms such as "above", "below", "left", "right", "front", "back", etc. may be used herein to describe the relationship of one element to another, as shown in the figure. It will be understood that spatially relative terms and the configurations shown are intended to include different orientations of the device in use or operation in addition to the orientations described herein and described in the figure. For example, if the device in the accompanying drawings is turned over, the elements described as "below" or "below" other elements or features will then be oriented to be "above" other elements or features. Therefore, the exemplary term "above" can include both above and below orientations. The device can be oriented in other ways (e.g., rotated 90 degrees or in other orientations), and the spatially relative descriptors used herein are interpreted accordingly. The term "approximately" can refer to within ±10% of the value. In addition, in the case of providing a range of values, each sub-range and each individual value between the upper and lower ends of the range are envisioned and therefore disclosed.

[0051] In addition, in the present disclosure, when an element is referred to as being "on another element," "connected to" another element, "coupled to" another element, "in contact with another element," etc., it can be directly on another element, directly connected to another element, directly coupled to another element, or directly in contact with another element, or intervening elements may be present unless otherwise specified.

[0052] The terms used herein are only for the purpose of describing specific embodiments and are not intended to limit the exemplary embodiments. As used herein, singular forms, such as those introduced with the articles "a" and "an", are intended to include plural forms unless the context indicates otherwise. In addition, the terms "include", "including", "comprise" and "comprising" specify the presence of the elements or steps, but do not exclude the presence or addition of one or more other elements or steps.

[0053] In addition, the various components identified herein may be provided in an assembled and completed form; or some or all of the components may be packaged together and sold as a kit with instructions (e.g., in written, video, or audio form) for a customer to assemble and / or modify to produce a finished product.

[0054] In minimally invasive surgery, internal anatomy is accessed through natural orifices or small external incisions. It includes a variety of different practices such as catheter-delivered stents (angioplasty), flexible endoscopy of the gastrointestinal tract, laparoscopy for abdominal disease, and transnasal surgery for neurological diseases. Common to these procedures is the need for wrist-like articulation at the distal end of the tools to manipulate tissue and visualize the surgical site. The need for this articulation was an important motivation for the introduction of robotics into minimally invasive surgery, which remains a vibrant area of ​​research and development today.

[0055] Cable-driven methods of proximal actuation are the primary approach to achieving wrist articulation because of the significant size constraints imposed by the size of the access method and the anatomy. However, the drawbacks of cable actuation are well known: the output force is limited by the allowable cable tension, and the bandwidth is limited by the mass and elasticity of the series elastic and recoil-inhibiting preload components. These limitations have prompted significant research efforts to create new distal dexterous mechanisms actuated by other means, such as shape memory actuators, pneumatic fluids, or flexible rods, to achieve dexterity without the limitations of cable actuation.

[0056] Here, we focus on a subset of tools used in minimally invasive surgery—those used for energy delivery. Common surgical energy sources include monopolar and bipolar radiofrequency current (electrosurgery), thermal cautery (direct current heating), ultrasonic vibrations, argon beam coagulation (argon assists the conduction of radiofrequency current), and lasers. These tools are critical to the surgical workflow because they enable cutting, coagulation, desiccation, and carbonization of tissue deep within the body. The different energy sources ultimately have the same effect—denaturing proteins by heating the tissue. Furthermore, these energy sources are currently used in similar ways. Specifically, the energy is brought close (e.g., within a few millimeters) to the tissue, and the energy is delivered directly from the electrode or fiber to the anatomical structure being treated.

[0057] However, it is possible to adopt a significantly more sophisticated approach when delivering laser energy. Although current laser-based tools are used in a static, close-contact manner, a dynamic approach can be adopted to manipulate the laser energy using a micro-laser galvo as a wrist. Because the laser can be focused and steered using low-inertia optics, high-bandwidth distal actuation can be used to control the position of the laser. This yields the same benefits that a robotic wrist confers to a mechanical end effector (i.e., the ability to avoid corners and maneuver around obstacles) while enabling the additional ability to precisely control the speed of the laser on the tissue over a wide range. This is important because laser speed strongly affects the duration of the laser shot, which is a key factor in determining the quality of the laser / tissue interaction with respect to incision depth, spread of thermal damage, and hemostatic efficacy.

[0058] The challenge in manufacturing such robotic devices is the opto-electro-mechanical complexity required in a small package. Device diameter is a key limiting factor, depending on the type of surgical tool used; a sampling of commonly used tools shows a range of different sizes: colonoscopes from 9.7 mm to 14.8 mm (e.g., PCF-PH190I / L and CF-FH260AZI / L colonoscopes from Olympus Medical Systems); laparoscopic tools are 5 mm or 8 mm (e.g., daVinci SP from Intuitive Surgical); rhinolaryngoscopes for transnasal access range from 2.6 mm to 4.9 mm (e.g., ENF-V3 and ENF-VT3 rhinolaryngoscopes from Olympus Medicalsystems); and cardiac catheters range from 2.67 mm to 3.33 mm, (e.g., AcuNav catheter from Biosense Webster).

[0059] Leveraging breakthroughs in millimeter-scale design and fabrication, we show how to integrate microcrystal laser beams that can be manipulated with a subset of these tools. Our approach builds on recent advances at the intersection of robotics and medicine that leverage millimeter-scale actuators and sensors to more precisely control tools and manipulate tissue. An example of our laser manipulation solution is 6mm in diameter and can therefore be seamlessly integrated into existing workflows in flexible gastroenterology and single-port surgery, settings where traditional tools are difficult to fit and lasers are particularly important.

[0060] The end effector with microrobotic laser beam manipulation can be integrated with surgical tools such as a colonoscope (Olympus CF-100L). The end effector can also be attached externally to the surgical tool, or because it is small enough, it can be integrated internally into a flexible scope for special purposes. The device can be 16mm in length and can have the ability to manipulate a focused laser beam through + / -10° in two orthogonal axes with a mechanical bandwidth of 1.2kHz.

[0061] The performance we achieve represents a substantial improvement over existing technologies, which tend to be bulky and slow. These solutions are larger than one centimeter in diameter and larger than two centimeters in length. While these devices achieve a similar range of motion to ours, speed is sacrificed due to the actuators and mechanisms they employ. In contrast, our approach employs piezoelectric bending actuators that act through a flexure-based transmission to produce rotational motion of a low-inertia mirror. In this way, we achieve a mechanical bandwidth that is one to two orders of magnitude greater than existing technologies and excellent repeatability (i.e., precision).

[0062] In this paper we describe substantial advances in design, fabrication, and control that allow us to surpass the performance of state-of-the-art devices. These include general principles for building micro-laser galvanometers, as well as specific design details of our device. In the case of our approach, we separate the device into modules—lenses, actuators, mirrors, etc.—that can be individually verified and then mounted on a disc and assembled onto a tracked superstructure that serves as a common mechanical ground, thereby minimizing the number of assembly features and facilitating alignment of critical components. For fabrication, we can use lamination techniques to create micro-compliant mechanisms, allowing us to produce large angles of mirror rotation without sacrificing miniaturization. For control, a low-dimensional hysteresis compensation scheme can be used to correct for hysteresis in the piezoelectric bending actuators and intermediate mechanisms.

[0063] Figure 1A laser scanning system for transoral robotic surgery is shown in the figure, which includes a compact laser manipulation end effector 10. The device receives laser input from an optical fiber 12 and manipulates the focus point on the surgical plane. The device can include two lasers, including a lower power navigation laser for illuminating the internal cavity to assist in visualization and navigating the end effector to the desired position, and a higher power surgical laser for cutting tissue. The laser light passes from the optical fiber 12 through a collimator 14 to a mechanical structure 16, which includes a first linkage actuator 18 and a second linkage actuator 22, the first linkage actuator 18 is configured to displace a first linkage and mirror assembly 20, and the second linkage actuator 22 is configured to displace a second linkage and mirror assembly 24. Figure 2 and Figure 3 is a long exposure photograph showing a typical cutting profile used in laser-based tissue resection.

[0064] like Figure 4-Figure 11 As shown, recently developed microfabrication techniques [such as "Monolithic fabrication of millimeter-scale machines" by P.S. Sreetharan et al., Journal of Micromechanics and Microengineering, Vol. 22, No. 5, 055027 (2012)] can be used to fabricate a micromechanical actuator linkage 38 that converts the quasi-linear motion of high-bandwidth piezoelectric bending actuators 18 and 22 into rotational motion for determining the orientation of mirrors 28 and 30 that receive the laser beam after reflecting the beam from a fixed mirror 42. An example of our device includes two such actuator-actuator-mirror combinations that are orthogonally placed so that the two actuation inputs from actuators 18 and 22 correspond to ablation in orthogonal directions on the target tissue. The light is transmitted to the laser scanner using an optical fiber in a laser injection catheter 36, where the output is fed as a free beam (no longer contained in the optical fiber) into a gradient refractive index collimating lens 32 and a micro plano-convex focusing lens 34 assembly. The free focus beam is then steered by rotating the two active mirrors 28 and 30 using a crank slider transmission to direct the beam from the actuation subsystem 50 ( Fig.10 ) to the desired rotational motion. Scan frame 40 (comprising a rigid disk-shaped platform 92 slidably mounted on rigid rods 94) provides mechanical grounding for actuators 18 and 22 and transmission 38, while also ensuring that optical subsystem 48 ( Fig. 9 ) and mirror system 52( Fig.11 ) to ensure precise alignment between them.

[0065] like Figure 4-Figure 7 As shown, the end actuator 10 may include two actuators 18 and 22 in the form of piezoelectric bimorph cantilevers 18 and 22, as described in US Pat. No. 9,038,942 B2 (Harvard).

[0066] The piezoelectric bimorph cantilevers 18 and 22 can be manufactured by intelligent composite microstructure manufacturing techniques, such as described in "The First Takeoff of a Biologically Inspired At-Scale Robotic Insect" by R. Wood in IEEE Transactions on Robotics, 24(2), 341(2008). First, individual thin sheets [such as PbZrTiO3 (PZT)-5H piezoelectric layers and M60J carbon fiber / cyanate resin prepreg layers] can be laser micromachined into the desired link shape. These layers can then be stacked and aligned and subjected to a controlled curing cycle to adjust the temperature, pressure, and time of curing. Each of the resulting actuators 18 and 22 is secured at a proximal end to a frame 40 (comprising a plurality of rigid plates 92 mounted along a plurality of rigid rods 94) and at a distal end to a passive drive linkage 38, wherein the passive linkage 38 provides a drive structure for converting the bending displacement of the cantilever actuator 18 / 22 to adjust the tilt of one of the active mirrors 28 / 30 to steer the reflection of the laser toward a desired target. A voltage source is also coupled to the actuator 18 / 22, wherein applying a voltage to either piezoelectric layer will cause the actuator 18 / 22 to bend.

[0067] exist Fig.11 1 shows one of the channel drive linkages 38 coupled to one of the active mirrors 28 / 30. The mirror 28 / 30 is mounted to pivot about a rotation pin 44 mounted in a rigid mount extending from a mechanical ground plate 54 in the frame, such that rotation of the rotation pin 44 in the rigid mount via displacement of the passive linkage 38 produces a change in the tilt angle of the active mirror 28 / 30.

[0068] Design requirements were derived from the anticipated surgical environment. We selected a target field of view of 10 × 20 mm based on the nominal dimensions of the vocal cords. An upper diameter limit of 13 mm was chosen for the device to allow space for retraction tools, visualization, and illumination. A standoff distance of 20 mm was chosen to balance the need for visualization and tissue manipulation while minimizing the field of view of the scanning system. Finally, we selected a target spot size of 250 μm and a target surface speed of 100 mm / s, both consistent with the scanning system used in TLM.

[0069] Many different lasers are used in laryngeal surgery, including CO2, potassium titanyl phosphate (KTP), and diode lasers. While CO2 lasers are preferred in many cases, fibers that deliver this wavelength at high power tend to have large core diameters, which means that the spot size would exceed our design requirements. Therefore, we designed an example device for use with a diode laser that can be delivered at high power (60W) through a 35-μm core PM780-HP fiber from Nufern. However, for this prototype, we verified its mechanical properties with a lower-power fiber inspection laser. We selected a collimator with a focal length of 6.17mm and a 38.1mm focusing lens, which should produce a spot diameter of 210μm when combined with a fiber with a core diameter of 35μm and a numerical aperture of 0.12. Both lenses are aspheric plano-convex to reduce aberrations. The mirror sizes we selected were 2mm×2mm and 2mm×5mm, with an initial angle of 40°, selected based on the kinematic analysis described below. We chose to use an off-the-shelf factory-aligned fiber-coupled collimator (F110 FC-633 from Thorlab) to simplify integration at the expense of increasing the size of the device.

[0070] The actuator lengths were chosen to be 10.6 mm and 12.1 mm to achieve appropriate displacement without significantly lengthening the device. We assumed that the actuator stiffness is approximately equal to the actuator output stiffness, which means that the actuator's in-situ deflection is reduced by a factor of two relative to the free deflection. Under this assumption, at 200 V (the upper limit of the operating voltage to avoid damage to the piezoelectric ceramic actuator material), the actuator deflections are ±200 μm and ±220 μm, respectively.

[0071] Using the optical model described above, and considering the positions of the mirrors to minimize the device diameter, we determined that the favorable range of motion for the first mirror is ±10° and the favorable range of motion for the second mirror is ±15°. To generate these motions, we chose Fig.12 The crank slider connecting rod lengths shown in (where connecting rod (a) has a length of 2 mm for both mirrors, where the length of connecting rod (d) is 0.9 mm for the first mirror and 0.6 mm for the second mirror):

[0072]

[0073] Where θ0 is the initial angle of the crank slider, Δ xis the actuator deflection and x0 is the initial distance between the slider and the pivot. The initial angle, θ0, is 70° for the first mirror and 75° for the second mirror. Of the two linkages, the length of the crank arm has the greatest impact on the output motion - the shorter the crank arm, the greater the output motion and therefore the higher the transmission ratio. Also of note is the initial angle of the mechanism, which trades off symmetry and output amplitude. The linkage configuration we chose achieves the desired range of motion while minimizing asymmetry. Fig.13 and Fig.14 The assembled linkage is shown.

[0074] Given the mirror angle determined by the crank-slider relationship, expressed as Equation 1, the position of the laser point on the target plane can be found using the vector formula for mirror reflection. Fig.15 The symbols and geometrical definitions used are given in the schematic diagram of the system shown, which shows the trajectory of the laser beam 46 reflected from the first mirror 28 and the second mirror 30 to the target plane 58. th j of the coordinate system th The basis vectors are given by Relative to i th Coordinate system, the unit vector from point Pj to point Pk is given by Indicates. th The coordinate system is relative to i th Coordinate system orientation. Preliminary rotation angle θ around axis i. H y is the Householder transform about the y-axis.

[0075] The orientation of the mirror in 3D space is given by the following equation:

[0076] and (2)

[0077]

[0078] where θ 1,0 is a design variable representing the initial direction of the first mirror. At this time, the light beam 46 reflected from the first mirror 28 has the following direction,

[0079]

[0080] The light beam 46 then travels a distance d 12 After intersecting with the second mirror 30, it is expressed as follows:

[0081]

[0082] Therefore, the position p2 of the laser beam 46 on the second mirror 30 is given by the following equation:

[0083]

[0084] Similarly, the reflected beam 46 from the second mirror 30 has a direction,

[0085]

[0086] and at a distance d 2t Then intersect with target plane 58:

[0087]

[0088] Finally, the position p of the laser point on the target surface t Given by the following equation:

[0089]

[0090] The manufacturing process used to manufacture the transmission linkage is based on the method developed in "Monolithic fabrication of millimeter-scale machines" by P.S. Sreetharan et al., Journal of Micromechanics and Microengineering, Vol. 22, No. 5, 055027 (2012), and is Figure 16-Figure 21 The layers of rigid material 60 (e.g., steel), more flexible material 62 (e.g., polyimide), and adhesive 64 (e.g., acrylic adhesive) are cut using a 7-W, 355-nm DPSS laser micromachining system (Oxford Laser E Series), and then aligned and placed in a heated press with a lower plate 66 at Fig.17 ) to bond these layers 60, 62 and 64 together. The bonded laminate is removed from the press and a preliminary release cut is made, such as Fig.18 Then, as Fig.19 As shown, the laminate is bent into its desired configuration; the first mirror drive involves the use of an assembly fixture 70, such as Fig. 20 As shown, it is bent to the correct angle, while the second mirror utilizes alignment pins for stacking. The laminate is then locked in its aligned configuration with cyanoacrylate (CA) glue before a final cut is made to release the desired degrees of freedom. Each linkage 38 is set up under a confocal microscope and its functional dimensions are measured to validate the process. The fabrication of the linkage 38 is a very repeatable process with a standard deviation of less than 4% of the expected value across all 12 linkages built, which corresponds to a 5% variation in the linkage transmission ratio. The laminate transmission linkage 38 will bend at the longitudinal gap between the rigid linkages (approximately a pivot) where the flexible material is exposed.

[0091] The small connecting rod length of the slider crank makes it challenging to assemble, but it can be easily assembled by bending its physical arm back to create a virtual connecting rod (such as Fig.13 and Fig.14 As shown), its operation becomes very easy. In addition, with this method, the bending angle of the flexure is zero when it is stationary, which makes it easier to ensure that they are in the elastic deformation zone.

[0092] The ground connection of the linkage is attached to a 7075 aluminum alloy base structure fabricated with a five-axis CNC machine (Bridgeport, Hardinge, Inc). The mirror is fabricated by laser cutting an aluminum sputtered 100 μm thick fused quartz wafer (from Denton Vacuum Desktop Pro). A protective gel layer is added to the mirror to protect the reflective coating during assembly. The linkage is then combined with its corresponding actuator and held in place with alignment pins and press-fit plates before the linkage is inserted into the mechanical structure. The actuator base and transmission assembly ground linkage are then bonded to the mechanical structure with cyanoacrylate glue. Finally, fabrication is completed by carefully removing the protective gel layer from the mirror.

[0093] To verify the kinematic model of the slider crank (Eq. 1), the mirror angle and actuator displacement were measured as a function of the actuator voltage. The mirror angle was measured with a high zoom inspection camera (PIXELINK PL-B741F from Pixelink) at a static voltage input, and the actuator displacement was measured with a laser Doppler vibrometer (PSV-500 vibrometer from Polytec GMBH) at a 1 Hz quasi-static cyclic input. Fig. 22 In Figure 1, the measured 28” / 30” and predicted 28' / 30' mirror angles are plotted as a function of the actuator input, with the theoretical value derived from the measured actuator displacement at a given voltage. The results show that the model captures the generalized behavior of the linkage well. The mean absolute errors for the first mirror 28” and the second mirror 30” are 0.6° and 1.2°, respectively.

[0094] To find a safe limit on the input drive frequency, we performed a frequency analysis of the motion actuator-actuator-mirror system. The data was collected using the same laser Doppler vibrometer with a low voltage, white noise input. The results are Fig.23 As expected, both subsystems behave very similarly to second-order linear systems. The resonant frequencies of 850 Hz and 750 Hz are consistent with physical intuition—the free-beam resonance of the actuator (1.6 kHz) has been reduced due to the increased mass of the actuator.

[0095] To demonstrate the beam steering capabilities of the device, we created a benchtop scanning arena consisting of a calibrated high-speed camera (PHANTOM V710 from Vision Research) with a 200mm macro lens and two non-flickering floodlights, all mounted on an optical table. Fig.24 The camera view and the grid used for image registration are shown in 75. One pixel corresponds to 50 μm on the image plane. We fixed our device on an optical table with a standoff distance of 20 mm determined by the distance to the target.

[0096] We then calibrate the scanner by scanning voltage space and tracking the laser spot position. The overall achieved field of view74 matches well with the model prediction72, as Fig.24 However, unmodeled compliance and assembly misalignment in the system caused slight distortions in the laser task space. Rather than adding additional degrees of freedom to our model and calibrating it, we chose to use a lookup table approach to control the position of the laser spot.

[0097] This approach produces reasonable results for task space trajectory tracking, e.g. Fig.25 , where the desired trajectory is represented by graph 76. Three shapes, an "H-shape" (left), a "star-shape" (middle) and a Lissajous figure (right), are plotted over a range of velocities (at 100 mm / s, shown by graph 78; at 1 m / s, shown by graph 80; at 2 m / s, shown by graph 82; at 3 m / s, shown by graph 84; and at 7 m / s, shown by graph 86). Each shape is shown at a base velocity 78 of 100 mm / s, which is the highest speed that can be reproduced before oscillations in the transmission erode the tracking performance of the device. This oscillation only appears in the "H-shape" trajectory at 2 m / s (graph 82). Of the three shapes, the "H-shape" is the most challenging because its trajectory undergoes rapid, abrupt changes. The Lissajous figure is the simplest because the drive voltage on the actuator is smooth and almost sinusoidal, so there is little loss in performance up to 7 m / s (graph 86).

[0098] The prototype device met most of the design requirements: field of view (slightly less than 10mm x 20mm), scanning speed (up to 7m / s), and scanning device size (11mm). The combination of assembly misalignment and unmodeled off-axis compliance resulted in some errors in our laser spot position model, such as Fig.24 This can be improved by minimizing the amount of manual labor required for assembly and by using other flexure materials or geometries with more suitable compliance ratios.

[0099] While calibration and feed-forward control allow qualitative trajectory tracking, large deviations from the expected trajectory remain, likely because the calibration is performed statically and does not capture dynamic, nonlinear, or hysteresis effects. Feedback control may therefore be necessary to accurately control the laser spot position. To this end, embodiments of the device may incorporate angle sensing on the mirror output, or directly feed back the laser position to the controller.

[0100] Another important next step is the integration of high-power optics. For this purpose, higher quality mirror coatings can be incorporated. Among the available mirror coatings, gold is a favorable choice because of its high reflectivity (>97%) at favorable wavelengths of 800nm ​​to 1000nm. Dielectric coatings can also be considered, as they can reach a reflectivity of 99.8%, but they are more selective to wavelength.

[0101] The somewhat fragile mechanics of the scope and linkage can be enclosed to protect them, and the scope can be shielded from fluid flow at the surgical site. This device can be used with existing transoral robotic tools for visualization and tissue retrieval, so the tasks can be coordinated with those tools.

[0102] In various examples, a gradient index collimating lens collects light from a cannulated fiber and directs it to a micro plano-convex focusing lens. The light is reflected by a mirror at a 45° incident angle into a micro dual-mirror galvanometer. A flexible linkage converts the collimated linear motion of the piezoelectric bending actuator into rotational motion of the galvanometer. The cylindrical profile facilitates integration with existing surgical tools.

[0103] Below, we present the design, fabrication, and control insights of our laser manipulation solution. We first describe the general design considerations that enabled miniaturization of the galvanometer portion of the device. This is the least challenging component to miniaturize in terms of design and fabrication. Next, we provide a detailed description of the design and construction of each component and its assembly. Third, we report the methods used for static position control and characterization. Finally, we characterize the dynamic characteristics of the device, demonstrate the capability for high-bandwidth operation, and interface the device to a commercial colonoscope.

[0104] The galvo mirrors are the most complex components in a device designed for miniaturization, but there are some general insights into the design space that can guide our efforts. The goal is to minimize the distance between mirrors given the size of the beam, the size of the mirrors, and the required range of motion, while avoiding collisions between the mirrors and between the reflected beam and the first few mirrors in the optical path (see Figure 26-Figure 28 ). An important high-level design consideration is the number of mirrors used. Somewhat counter-intuitively, a three-mirror design can actually be made smaller than a two-mirror design if the same range of motion is required for each.

[0105] First, consider Fig.26 The three-mirror design shown in , where the incident light beam enters from the left and is then reflected in sequence by the fixed mirror 42, the first active mirror 28, and the second active mirror 30 (circumferential diameter of 4 mm). In the image on the left, the first active mirror 28 and the second active mirror 30 are rotated by +10°. In the middle image, the first active mirror 28 is rotated by +10°, while the second active mirror 30 is rotated by -10°. In the image on the right, the first active mirror 28 is rotated by -10°, while the second active mirror 30 is rotated by +10°. If the distance between the fixed mirror 42 and the first active mirror 28 is too small, the reflected light beam from the second active mirror 30 will intersect with the fixed mirror 42. If the distance between the first active mirror 28 and the second active mirror 30 is too small, the mirrors will collide. However, increasing the distance between the mirrors not only increases the size of the device, but also means that a larger area of ​​mirrors is required to collect the reflected light with the same range of motion. The three-mirror design shown balances all of these considerations, producing a device with a + / -10° range of motion on each active mirror within a 4mm diameter footprint. Note that the use of chamfered corners on the mirrors increases the range of motion by preventing collisions in critical locations.

[0106] The first dual-mirror design (in Fig. 27 (two configurations are shown in Figure 2) have the same diameter footprint as the three-mirror design, but their range of motion about the first mirror axis is halved. In the image on the left, the first active mirror 28 is rotated through an angle of +5°, while the second active mirror 30, which has a circumferential diameter of 4 mm, is at a neutral angle of 0°. In the image on the right, the first active mirror 28 is rotated through an angle of -5°, while the second active mirror 30 is rotated through a neutral angle of 0°. A series of trade-offs lead to this result; if the mirrors 28 and 30 (both active) are too close together, then the light beam reflected from the second active mirror 30 will intersect the first active mirror 28 for large positive rotation angles of the first mirror 28. On the other hand, if the distance between the two mirrors 28 and 30 is increased, then the second mirror 30 must be enlarged to accept incident light for large negative rotation angles of the first mirror 28. This is the second two-mirror design ( Fig.28 ); it achieves the same range of motion as the three-mirror design at the expense of a 50% increase in diameter (the circumscribed diameter of the second mirror 30 is 6 mm). In the left image, the first active mirror 28 is rotated by an angle of +10°, while the second active mirror 30 is rotated by a neutral angle of 0°. In the right image, the first active mirror 28 is rotated by -10°, while the second active mirror 30 is rotated by +8°.

[0107] Therefore, the three-mirror galvanometer ( Fig.26 ) than the same size dual-lens design ( Fig. 27 ) achieves a greater range of motion, whereas to achieve the same range of motion as a three-mirror design, a two-mirror design with a 50% larger diameter is required ( Fig.28 ). The range of motion is primarily determined by the ability of the mirrors to adequately capture and reflect incoming light, but is limited by the need to avoid collisions between mirrors and between reflected laser light and previous mirrors in the light path. These sample configurations show the limits of the range of motion where these collisions are likely to occur.

[0108] Throughout this analysis, we have assumed a fixed beam diameter of 1 mm, which is slightly larger than the low-power pointing lasers used here for device validation, but a reasonable upper limit for collimated, high-power beams. For simplicity, we have also chosen the neutral position of each mirror to be at a 45° angle of incidence to the incident beam, allowing for it to deviate slightly from 45° may produce slightly different results but does not change the structure of the design trade-offs. We have also assumed that a symmetrical range of motion is desired. Again, if this is not the case, then the resulting design may change slightly, but the nature of the design space does not need to change. Finally, it is important to emphasize that we have been considering the case where the outgoing light is parallel to the incoming fiber (forward viewing). If we wanted the outgoing light to be perpendicular to the incoming fiber (side viewing), there would be no reason to use a three-mirror design; a two-mirror design would be perfectly acceptable in terms of miniaturization and range of motion. However, since most energy delivery tools used in surgery are forward-looking, we chose to use that configuration for these examples of our device, although other examples of this device could easily be adapted for side-viewing devices.

[0109] With these general insights, we proceeded to the detailed design and fabrication of the device. For construction, we used off-the-shelf components and custom parts, which were primarily fabricated by laser micromachining (using an OxfordLasers E-series with a Coherent Avia 355-7 laser and a Canon galvanometer). The kinematic structure of the piezoelectric bending actuator motion transmission assembly of the three-mirror galvanometer is shown in Figure 2. Fig.29 and Fig.30 As shown, it includes bending actuators 18 and 22, drive linkage 38, and mirrors 28 and 30. These components convert the quasi-linear (parabolic) input motion into rotational motion of the mirrors.

[0110] The components of the three-mirror galvanometer before assembly are as follows Fig.318 and 9. As shown (in assembly order, first row from left to right, followed by second row from left to right). The components shown include a retaining clip 88, a fixed mirror 42 assembly, a second active mirror (y-axis) 30 assembly, a first active mirror (x-axis) 28 assembly, a focusing lens 34 assembly, a piezoelectric actuator 18 and 22 assembly, a collimating lens 32 assembly, and a supporting superstructure 90. The retaining clip; the mirrors 42, 28 and 32; and the actuators 18 and 32 are each mounted on a disc-shaped platform 92 having an aperture, and the platform 92 is slidably mounted to a rod 94 of the supporting superstructure 90 through the aperture, thereby providing a detachable modular structure. Each of these components can be formed via UV laser cutting. The assembly of the active mirrors 28 and 30 and the assembly of the piezoelectric actuators 18 and 22 may also involve precision lamination. Sputtering deposition can be used to form the mirror 28, 30 and 42 assembly. The lenses 32 and 34 can be off-the-shelf components. Finally, the focusing lens 34 may be shaped by diamond grinding.

[0111] Specific details of component design, manufacture, and functionality are listed below.

[0112] Steel rod superstructure 90: This part consists of two 500 μm diameter and one 300 μm diameter stainless steel rods 94 (Misumi USA) onto which the remaining parts are assembled. An alignment fixture is used to orthogonally position the rods 94 into the FR4 disk 92. Spring steel preload springs are attached to the disk 92 to compress and hold the assembled parts in place. All parts are laser micromachined.

[0113] Ferrule-terminated fiber and collimating lens 32 : An off-the-shelf ferrule-terminated fiber (SMPF0106-FC, Thor Laboratories) was assembled with a gradient index collimator 32 (GRIN2306A, Thor Laboratories) and attached to a FR4 support plate 92 .

[0114] Piezoelectric bending actuators 18 and 22: Actuators 18 and 22 were sized using the process and materials described in “Design and manufacturing rules for maximizing the performance of polycrystalline piezoelectric bending actuators” by NT Jafferis, M J Smith, and R J Wood, Smart Materials and Structures, Vol. 24, No. 6, p. 065023 (2015). We selected the actuator dimensions based on the available space near the optical components: active length 7 mm, tip length 1.8 mm, bridge length 0.5 mm, base width 1.4 mm, tip width 0.4 mm.

[0115] The actuators were driven in a unipolar configuration with a fixed bias voltage, so each actuator was controlled by a single time-varying input signal from zero to the bias voltage. In order to achieve appreciable output displacement while ensuring that the tensile strain of the piezoelectric ceramic is well below the failure limit, a fixed bias voltage of 200 V was chosen. Under these driving conditions, the actuators achieved a free displacement of + / -200 μm with a first resonance frequency of 2.6 kHz.

[0116] Focusing lens 34: Focusing lens 34 was obtained from an off-the-shelf part (#89-003, Edmund Optics) and ground to size using an alignment fixture and a diamond cut-off wheel. Registration marks laser pre-engraved into lens 34 allow for optical center alignment after grinding.

[0117] Articulated mirrors 28 and 30 and motion transmission: These are complex assemblies of rigid and flexible components manufactured using lamination techniques. They may include a four-bar crank slider linkage formed from stainless steel, polyimide and a heat curable acrylic adhesive (PYRALUX adhesive from Dupont Inc). The mirrors 28 and 30 are located on cranks, and the sliders each interface with one of the piezoelectric actuators 18 / 20. The x-axis (first) transmission additionally contains a linearized linkage to compensate for out-of-plane motion of the bending actuator tip. These are in Fig.29 and Fig.30 Mirrors 28 and 30 were made from sputtered aluminum (Denton Desktop Pro PVD system from Denton Vacuum LLC, Moorestown, NJ, USA) on a 100 μm fused silica substrate and singulated using a UV laser.

[0118] The detailed design of the transmission follows the properties of the actuator. We want each mirror to achieve + / -10° of motion (driven by the bending of the actuator in either direction), consistent with state-of-the-art devices. To achieve this motion, there are two important design considerations: (1) the transmission ratio of the linkage, and (2) the stiffness of the linkage relative to the actuator. Based on experience with dimensioning similar mechanical components, we choose a target transmission ratio of 0.1° / μm and a target stiffness that is the same as that of the actuator. In this device, we obtain transmission ratios of 0.13° / μm and 0.12° / μm (due to small manufacturing errors) and stiffnesses that are 60% and 90% of the actuator stiffness.

[0119] Note that the stiffness of the transmission is parallel to the actuator stiffness and results in a reduced free displacement of the actuator. Fabricated in this way, we obtain a range of motion of -12.5° to 20° and -11.2° to 13.7° for the dual mirrors, respectively. This asymmetry corresponds to small assembly misalignments and slightly nonlinear transmission kinematics. The neutral (zero) mirror angle and actuator position corresponds to a configuration where the output beam is aligned along the longitudinal axis of the device.

[0120] Fixed mirror 42: This aluminum sputtered fused silica mirror 42 is fixed at 45° relative to the incident light using two alignment blocks, which in turn are connected to the FR4 support plate 92.

[0121] Retaining Clip 88: This spring steel component fits into a groove carved into the stainless steel rod 94 of the superstructure 90. It works with the preloaded spring located at the base of the superstructure to axially constrain the assembly.

[0122] Spacer tubes and discs: These pieces are laser micromachined to ensure proper spacing and alignment between components. Their length (or thickness) is their critical dimension.

[0123] We begin characterizing the position control of the end effector by measuring open-loop repeatability, an important metric because it describes the fundamental physical limits of a device to reproduce identical motions. Statics, plasticity, and other phenomena mean that identical inputs do not produce identical outputs. Unidirectional repeatability is a measure of repeatability when the measurement points are approached from only one direction. The maximum 2σ standard distance between the sample points is 200μm, meaning that one can be 95% confident that any series of identical motions will fall within a 200μm radius of dispersion about the mean trajectory.

[0124] Position control is complicated by the presence of hysteresis. Hysteresis is a bidirectional effect that is primarily caused by domain reorientation within the piezoelectric ceramic actuator; this means that the position of the laser depends on the time history of the input. This dependence is obviously undesirable as it complicates control and makes the use of the device unintuitive. To minimize the effects of hysteresis, we implemented a feedforward compensation scheme. To validate our approach, we controlled star tracks for both uncompensated and compensated inputs. The original inputs clearly show the effects of path dependency, while the corrected inputs show significantly improved tracking; essentially, this is a measure of bidirectional repeatability. Quantitatively, we found that the maximum 2σ standard distance around the set point was 2.14 mm without compensation and 0.72 mm with compensation, which represents a reasonable improvement with feedforward compensation; further improvements can be achieved with feedback control.

[0125] Due to the low dimensionality of the workspace and input space and the lack of sensor information, we chose to control the device using a direct model-free mapping between actuator input and laser spot position, cascaded with a feed-forward compensation of hysteresis. We fit three- and two-degree polynomial surfaces to the open-loop repeatability measurement data of the first and second mirrors, respectively. These fits were centered at 73V and 93V, respectively, corresponding to the neutral angles of the mirrors.

[0126] The system can be dynamically controlled and characterized. The bandwidth of the system is limited by the high frequency resonant oscillation of the mirrors. The main resonant frequencies of the two mirrors are 1.8kHz and 1.9kHz, as shown in Fig.32 (This figure plots the frequency response of the first active mirror to a low voltage white noise input) and Fig.33 (This figure plots the frequency response of the second active mirror) is shown in the Bode plot shown. In addition, there is a lower frequency mode at 1.2kHz on the first mirror (probably due to distortion or another off-axis mode). To avoid exciting these modes, we use a limited-squish curve scheme (also called an S-shape or S-curve).

[0127] from Fig.34 and Fig.35 As can be seen in Figure 2, the system shows only a small deviation from the static trajectory 26 at high speeds. At low speeds (7.8 mm / s - Fig.35 solid line in the figure) and high speed (3,900 mm / s - Fig.35 There is only a 5% deviation between the trajectories followed by the dotted line in the figure). In addition, due to the dual-axis control, the system can track complex planar trajectories. The large bandwidth in the system can also be used to generate multi-modal profiles. There are some trajectories in which high-frequency inputs are superimposed on low-frequency inputs, which allows the user to instantly change the effective area of ​​energy application, which is particularly beneficial for situations where large areas need to be covered in a controllable manner, such as large-area hemostasis.

[0128] Due to the low profile and low mass (e.g. 700 mg) of the laser manipulation end effector, it can be easily connected to existing surgical tools. To demonstrate this, we attached it to the end of a colonoscope (Olympus CF-100L) and performed a simulated polyp removal task on a tabletop surgical simulator. We demonstrated high-speed control of the laser along a registered incision trajectory using both tele-control and robotics using a standard input device (Phantom Omni).

[0129] This work points the way forward for new methods of controlled energy delivery in minimally invasive surgery. It can simultaneously address three important challenges in surgical technology: (1) giving the surgeon greater control over the delivery of energy to soft tissue, (2) improving the maneuverability and dexterity of surgical instruments, and (3) reducing the body load on the operator and materials. Our contribution here is that by giving the surgeon control over the laser velocity as it interacts with tissue, the quality of the laser / tissue interaction can be more extensively tuned and optimized than with existing static wielding tools. Using the wrist joint, the surgeon will also gain access to previously inaccessible lesions. Our design, fabrication, and modular assembly approach is an additional contribution that enabled the creation of this device, but will also be applicable to other complex microrobotic systems and medical devices.

[0130] To achieve an accuracy consistent with the repeatability of the device, sensory feedback can be incorporated. A simple approach is to embed strain gauge sensors inside the piezoelectric bending actuator to estimate the actuator position and use current sensing to estimate the actuator velocity. The mirror position can then be calculated from the transmission motion and the laser position estimated from a model of the mirror reflection. Directly measuring the mirror position will lead to better estimates of the laser position, but the path to suitable microsensing methods is not as straightforward as sensing the motion of the actuator. Alternatively, visual feedback can also be used for high-quality estimates, although its sampling rate may be lower than what can be achieved with electromechanical sensors. In practice, some sensor fusion that combines these different information may be the best approach.

[0131] Additionally, to use the device with surgical lasers, appropriate high-power fibers, lenses, and mirrors may be incorporated. The choice of components depends on the wavelength and power of the laser being used. For example, if a CO2 surgical laser is being used, a gold-sputtered aluminum mirror and a zinc-selenide lens provide appropriate reflectivity and transmittance, respectively. Of particular interest is the "laser-induced damage threshold," a measure of the laser power an optical component can experience before degradation.

[0132] Regardless of the laser modality used, the device may be advantageously packaged to protect the device from fluids and debris in the surgical environment and to allow for sterilization. Because the tip of the laser fiber may become damaged after prolonged use, it may be advantageous to incorporate a method of decoupling the laser fiber from the device. This decoupling capability would allow the laser fiber to be cleaned and cut independently, significantly extending its life.

[0133] Because the laser position is controlled electronically rather than actuated by a cable, energy control can be decoupled from the overall motion of the surgical tool that deploys the laser. This independence is particularly important in flexible endoscopy, which often requires cumbersome manual actions, including equipment handoffs, putting down the endoscope, moving one's own hands multiple times to different parts of the instrument, and having another user operate the end effector. Therefore, one can easily imagine how a remote column, like a joystick, that is not dependent on position due to the spatial constraints of the cable could be ergonomically mounted somewhere on the endoscope or used with a second operator remote from the endoscope would improve the efficient use of time in the operating room that is spent on patient care rather than equipment manipulation.

[0134] Our laser manipulation method also enables new approaches to endoscopic visualization and visual biopsy. In optical coherence tomography and confocal microendoscopy, laser / tissue interactions are used to visualize subsurface structures, and scanning allows large areas of tissue to be viewed at once. Optical manipulation can also be used to increase the effective field of view of standard white-light imaging tools by stitching together a set of images acquired through rapid scanning. The scanning systems and modular device assembly methods described in this article can be adapted to fabricate millimeter-scale versions of these systems. Smaller versions of these systems were built using a laminate manufacturing technique that uses electrostatic and electrothermal actuators to excite the resonant scanning elements. Our method has advantages in terms of simplicity of construction and the ability to achieve a large range of quasi-static motion.

[0135] We also anticipate that this technology could be applicable to other micro-robotic systems, especially micro-aircraft and satellites where size and weight are important. This technology could enable the fabrication of miniature light detection and ranging (LIDAR) sensors for mapping and navigation, and laser scanners for wide-area atmospheric pollution sensing.

[0136] Further examples according to the present technology are set forth in the following numbered clauses:

[0137] 1. A compact laser manipulation end effector, comprising:

[0138] a frame having a proximal end and a distal end, the frame having a maximum dimension in a plane orthogonal to the longitudinal axis not exceeding 13 mm;

[0139] at least two mirrors mounted proximate the distal end of the frame, the mirrors comprising a first active mirror and a second active mirror;

[0140] a first actuator mounted to the frame and configured to change a tilt of the first active mirror relative to the frame; and

[0141] a second actuator mounted to the frame and configured to change a tilt of the second active mirror relative to the frame,

[0142] wherein a path is provided through the frame to deliver a laser beam to the mirrors, and wherein the mirrors are positioned and configurable via the actuators to reflect a laser beam away from each mirror toward an external target.

[0143] 2. The compact laser manipulation end effector of clause 1, further comprising a third mirror positioned and configured to reflect the laser beam together with the first active mirror and the second active mirror.

[0144] 3. A compact laser manipulation end effector according to clause 2, wherein the third mirror is fixedly mounted.

[0145] 4. A compact laser manipulation end effector according to any one of clauses 1 to 3, wherein the first actuator and the second actuator are cantilevers.

[0146] 5. The compact laser manipulation end effector of clause 4, wherein the cantilever is coupled to the mirror via a passive linkage mechanism.

[0147] 6. A compact laser manipulation end effector according to clause 4 or 5, wherein the cantilever is a bimorph piezoelectric cantilever.

[0148] 7. The compact laser manipulation end effector of any of clauses 1-6, further comprising a surgical laser positioned to direct a laser beam through the compact laser manipulation end effector to the mirror for manipulation via reflection from the mirror.

[0149] 8. The compact laser steering end effector of clause 7, further comprising a navigation laser positioned to direct a laser beam through the compact laser steering end effector to the mirror for steering via reflection from the mirror.

[0150] 8.1 A compact laser manipulation end effector according to clause 1, wherein the frame comprises a plurality of rods and a plurality of platforms mounted on the rods, wherein components of the compact laser manipulation end effector are mounted on the platforms.

[0151] 9. A method for robotic laser manipulation, comprising:

[0152] generating a laser beam;

[0153] directing a laser beam from a proximal end of an end effector to an area proximal to a distal end of the end effector;

[0154] reflecting a laser beam in a region proximate the distal end of the end effector with at least two active mirrors;

[0155] An actuator is used to change the tilt orientation of the at least two active mirrors to shift the laser beam through the target.

[0156] 10. The method of clause 9, further comprising displacing the laser beam using a third mirror in combination with the at least two active mirrors.

[0157] 11. The method according to clause 10, wherein the third mirror is fixedly mounted.

[0158] 12. A method according to any of clauses 9-11, wherein the actuator is a cantilever.

[0159] 13. The method of clause 12, wherein bending of the cantilever displaces a passive linkage coupled to the mirror to change the tilt of the mirror.

[0160] 14. A method according to any one of clauses 9 to 13, wherein the target is a living organism.

[0161] 15. The method according to clause 14, wherein the living organism is a human.

[0162] 16. The method according to clause 15 further includes removing benign or cancerous lesions, for example, benign or cancerous lesions in at least one of the pharynx and the larynx, by shifting the laser beam by the mirror, or performing laser-assisted cardiac ablation, laser-assisted gastrointestinal surgery, laser-assisted extra-abdominal surgery, or laser-assisted transnasal skull base surgery in situ in the human body.

[0163] When describing embodiments of the present invention, specific terms are used for the sake of clarity. For the purpose of illustration, specific terms are intended to include at least technical and functional equivalents that operate in a similar manner to achieve similar results. In addition, in some instances where specific embodiments of the present invention include multiple system elements or method steps, these elements or steps can be replaced with single elements or steps. Similarly, single elements or steps can be replaced with multiple elements or steps that serve the same purpose. In addition, various parameters or other values ​​of characteristics are specified for embodiments of the present invention herein, and these parameters or values ​​may be adjusted up or down by 1 / 100, 1 / 50, 1 / 20, 1 / 10, 1 / 5, 1 / 3, 1 / 2, 2 / 3, 3 / 4, 4 / 5, 9 / 10, 19 / 20, 49 / 50, 99 / 100, etc. (or increased by 1, 2, 3, 4, 5, 6, 8, 10, 20, 50, 100 times, etc.), or rounded approximations thereof, or adjusted up or down to any variation specified above within the range of the specified parameter (e.g., when the specified parameter is 100 and the variation is 1 / 100, the value of the parameter may be in the range of 0.99 to 1.01), unless otherwise specified. In addition, although the present invention is shown and described with reference to its specific embodiments, it should be understood by those skilled in the art that various substitutions and changes in form and detail may be made thereto without departing from the scope of the present invention. In addition, other aspects, functions and advantages are also within the scope of the present invention; and all embodiments of the present invention do not necessarily achieve all of the advantages described above or have all of the features described above. In addition, the steps, elements and features discussed herein in conjunction with one embodiment may also be used in conjunction with other embodiments. The contents of the references cited herein, including reference texts, journal articles, patents, patent applications, etc., are hereby incorporated by reference in their entirety for all purposes; and all appropriate embodiments, features, characteristics and method combinations from these references and the present disclosure may be included in the embodiments of the present invention. In addition, the components and steps identified in the background section are integral to the present disclosure and may be used or replaced with components and steps described elsewhere in the present disclosure within the scope of the present invention. In a method claim (or in the case of a method described elsewhere), if the stages are described in a particular order - with or without a leading character for ease of reference - then, unless otherwise specified or implied by the terms and wording, the stages should not be interpreted as being limited in time to the order in which they are described.

Claims

1. A compact laser manipulation end effector, comprising: a frame having a proximal end and a distal end, the frame having a maximum dimension in a plane orthogonal to the longitudinal axis not exceeding 13 mm; at least two mirrors mounted proximate the distal end of the frame, the mirrors comprising a first active mirror and a second active mirror; a first actuator mounted to the frame and configured to change a tilt of the first active mirror relative to the frame; and a second actuator mounted to the frame and configured to change a tilt of the second active mirror relative to the frame, wherein the compact laser manipulation end effector receives fiber optic input and can be integrated with a surgical tool to receive laser light from the fiber optic and focus and manipulate the laser light for dexterous control of laser energy in minimally invasive surgery, and wherein a path is provided through the frame to deliver a laser beam to the mirrors, and wherein the mirrors are positioned and configurable via the actuators to reflect a laser beam away from each mirror toward an external target.

2. The compact laser manipulation end effector of claim 1, further comprising a third mirror positioned and configured to reflect the laser beam together with the first active mirror and the second active mirror.

3. The compact laser manipulation end effector according to claim 2, wherein: The third mirror is fixedly mounted.

4. The compact laser manipulation end effector according to claim 1, wherein: The first and second actuators are cantilevers.

5. The compact laser manipulation end effector according to claim 4, wherein: The cantilever is coupled to the mirror via a passive linkage configured to linearize out-of-plane motion of the cantilever.

6. The compact laser manipulation end effector according to claim 4, wherein: The cantilever is a bimorph piezoelectric cantilever.

7. The compact laser manipulation end effector of claim 1 further comprising a surgical laser positioned to direct a laser beam through the compact laser manipulation end effector to the mirror for manipulation via reflection from the mirror.

8. The compact laser steering end effector of claim 7, further comprising a navigation laser positioned to direct a laser beam through the compact laser steering end effector to the mirror for steering via reflection from the mirror.

9. The compact laser manipulation end effector according to claim 7, wherein: The compact laser manipulation end effector is integrated with a surgical tool.

10. The compact laser manipulation end effector according to claim 1, wherein: The frame includes a plurality of rods and a plurality of platforms mounted on the rods, wherein components of the compact laser manipulation end effector are mounted on the platforms.

11. A method for robotic laser manipulation, comprising: generating a laser beam; directing a laser beam via an optical fiber to a proximal end of an end effector to a region proximal to a distal end of the end effector, the end effector being incorporated into a microrobotic device, wherein the end effector comprises a frame having a maximum dimension in a plane orthogonal to the longitudinal axis not exceeding 13 mm; reflecting a laser beam in a region proximate the distal end of the end effector with at least two active mirrors mounted near the distal end of the frame; using an actuator mounted to the frame to change the tilt orientation of the at least two active mirrors to shift the laser beam through the target, Among other things, the method is applicable to any of the following micro-robotic systems: micro-aircraft and satellites; micro-light detection and ranging (LIDAR) sensors for mapping and navigation; and laser scanners for wide-area atmospheric pollution sensing.

12. The method of claim 11, further comprising shifting the laser beam using a third mirror in combination with the at least two active mirrors.

13. The method according to claim 12, wherein: The third mirror is fixedly mounted.

14. The method according to claim 11, wherein: The actuator is a cantilever.

15. The method according to claim 14, wherein: Bending of the cantilever displaces a passive linkage coupled to the mirror to change the tilt of the mirror.

Images