STEERABLE, FOLLOW THE LEADER DEVICE

Abstract

A highly articulated robotic probe (HARP) is comprised of a first mechanism and a second mechanism, one or both of which can be steered in desired directions. Each mechanism can alternate between being rigid and limp. In limp mode the mechanism is highly flexible. When one mechanism is limp, the other is rigid. The limp mechanism is then pushed or pulled along the rigid mechanism. The limp mechanism is made rigid, thereby assuming the shape of the rigid mechanism. The rigid mechanism is made limp and the process repeats. These innovations allow the device to drive anywhere in three dimensions. The device can remember its previous configurations, and can go anywhere in a body or other structure (e.g. jet engines). When used in medical applications, once the device arrives at a desired location, the inner core mechanism can be removed and another functional device such as a scalpel, clamp or other tool slid through the rigid sleeve to perform. Because of the rules governing abstracts, this abstract should not be used to construe the claims.

Claims

1.-41. (canceled)

42. A method of operating a highly articulated probe, the method comprising: receiving, by a computing device, position data from an input device, wherein the position data is associated with a highly articulated probe that comprises a first mechanism and a second mechanism configured to surround at least a portion of the first mechanism; translating, by the computing device, the position data into coordinate system data; and causing, by the computing device, a position of one or more shafts of one or more motors to be varied based on the coordinate system data, wherein the one or more motors are configured to control a tension of one or more wires of the second mechanism.

43. The method of claim 42, further comprising: receiving, by the computing device, an instruction from the input device to lock a position of the second mechanism; causing one or more of the one or more motors to be driven in a direction opposite a pull of the one or more wires; and causing one or more motors associated with the first mechanism to be driven in a direction of a pull of one or more wires associated with the first mechanism.

44. The method of claim 42, further comprising: causing one or more motors associated with the first mechanism to advance the first mechanism a distance equal to approximately a length of one of a plurality of links of the first mechanism; and locking a position of the first mechanism.

45. The method of claim 44, further comprising: causing one or more of the motors that are configured to control one or more wires of the second mechanism to advance the second mechanism a distance equal to approximately a length of one of a plurality of links of the second mechanism; and locking a position of the second mechanism.

46. The method of claim 42, wherein: the one or more motors comprises a first motor, a second motor, and a third motor, the one or more wires comprise a first wire, a second wire, and a third wire, the first motor is configured to control a tension of the first wire, the second motor is configured to control a tension of the second wire, the third motor is configured to control a tension of the third wire.

Description

BRIEF DESCRIPTION OF THE DRAWINGS

[0025] For the present disclosure to be easily understood and readily practiced, various embodiments of the present disclosure will now be described, for purposes of illustration and not limitation, in conjunction with the following figures wherein:

[0026] FIGS. 1A-1C are graphic demonstrations of the concept of the present disclosure;

[0027] FIGS. 2A-2D illustrate various configurations assumed by a prototype of one embodiment of the disclosed device;

[0028] FIGS. 3A-3D illustrate various views of a cylinder of the outer sleeve;

[0029] FIGS. 4A and 4B illustrate end and cross-sectional views, respectively, of a cylinder of the inner core;

[0030] FIGS. 5A-5D illustrate various views of another embodiment of a cylinder of a steerable inner core;

[0031] FIGS. 6A and 6B illustrates one example of a feeder mechanism;

[0032] FIG. 7 illustrates devices for controlling the tension on the wires;

[0033] FIG. 8 illustrates devices for controlling the tension on the wires of the outer sleeve;

[0034] FIG. 9 illustrates a device for controlling the tension on the wire of the inner sleeve;

[0035] FIGS. 10A-10C illustrate a schematic of extreme cantilever configuration for a worst case configuration (A), a simplified model (B), and a free body diagram (C) for the worst case configuration;

[0036] FIG. 11 is a block diagram illustrating the components of a control system and the flow of information between those components;

[0037] FIG. 12 is a block diagram of an exemplary electrical system for the feeding mechanism;

[0038] FIGS. 13A, 13B and 13C are an electrical schematic of a PID motor controller of FIG. 12;

[0039] FIG. 14 is flow chart of a stepped advancing mode of operation;

[0040] FIGS. 15A-15C illustrate a process for moving the device of the present disclosure;

[0041] FIG. 16 is a flow chart of a process for retracting the device;

[0042] FIG. 17 illustrates another embodiment of a link which may be used in the device of the present disclosure;

[0043] FIGS. 18 and 19 illustrate an embodiment of the present disclosure having an onboard camera mounted on the end thereof and used for visualization of internal organs; and

[0044] FIG. 20 illustrates another embodiment of the device of the present disclosure.

DESCRIPTION

[0045] A highly articulated robotic probe (HARP) 10 of one embodiment of the present disclosure shown in. FIGS. 1A-1C is essentially two concentric mechanisms, an outer one and an inner one, each of which can be viewed as a steerable mechanism. FIGS. 1A-1C show the concept of how different embodiments of the HARP 10 operate. Referring to FIG. A, we call the inner mechanism a first mechanism or inner core mechanism 12. We call the outer mechanism a second mechanism or sleeve mechanism 14. Each mechanism can alternate between being rigid and limp. In the rigid mode or state, the mechanism is just that rigid. In the limp mode or state, the mechanism is highly flexible and thus either assumes the shape of its surroundings or can be reshaped. It should be noted that the term limp as used herein does not denote a structure that passively assumes a particular configuration dependent upon gravity and the shape of its environment; rather, the limp structures described in this application are capable of assuming positions and configurations that are desired by the operator of the device, and therefore are articulated and controlled rather than flaccid and passive.

[0046] With this HARP 10, one mechanism starts limp and the other starts rigid. For the sake of explanation, assume the sleeve 14 is rigid and the core 12 is limp, as seen in step 1 in FIG. 1A. Now, the core 12 is both pushed forward by a feeding mechanism 16, described below, and its head or distal end is steered, as seen in step 2 in FIG. 1 A. Now, the core 12 is made rigid and the sleeve 14 is made limp. The sleeve 14 is then pushed forward until it catches up or is coextensive with the core 12, as seen in step 3 in FIG. 1A. Now, the sleeve 14 is made rigid, the core 12 limp, and the procedure then procedure repeats. One variation of this approach is to have the sleeve 14 be steerable as well. The operation of such a device is illustrated in FIG. 1B. In FIG. 1B it is seen that each mechanism is capable of catching up to the other and then advancing one link beyond. That requires an additional camera on the sleeve 14 but would potentially allow for quicker deployment of the HARP 10. In the current rendition, the sleeve 14 is steerable and the core 12 is not. The operation of such a device is shown in FIG. 1C.

[0047] in medical applications, once the HARP 10 arrives at: a desired location, the surgeon can remove the inner core 12 and slide either a conventional device or a custom tool through the rigid sleeve 14 to perform various operations.

[0048] The HARP 10 is not limited to surgery, but can be used in engine inspection, engine repairs, and engine retrofitting. Other applications include tank inspection, spying or surveillance-applications, bomb disarming, and inspection or repairs in tightly confined spaces such as submarines or within nuclear weapons. Other applications include structural (e.g. building) inspections, hazardous waste remediation and bioterrorists sample recovery. Clearly, the device of the present disclosure has a wide variety of applications and should not be taken as being limited to any particular application.

[0049] The HARP 10 of the present disclosure device bears some similarities to Bob Sturges' patented device (U.S. Pat. No. 5,759,151) although the present disclosure incorporates several major innovations. First, the core 12 and/or sleeve 14 is steerable. Second, the sleeve 14 can be made both rigid and limp. These two innovations allow the HARP 10 to drive anywhere in three-dimensions. Sturges' device assumes that it is moving through a tubular space, such as the large intestines. Sturges' device requires that the intestines shape the device as it goes. As Sturges' device propagates out it, cannot remember its previous configuration and hence cannot keep the path it followed due to the fact that it is composed of only one element that can become both rigid and stiff The HARP 10 can remember its previous configurations and for this reason, the HARP 10 can go anywhere in a three dimensional volume such as the intracavity spaces in a body. FIGS. 2A-2D illustrate various configurations assumed by a prototype of the device 10.

[0050] The following discussion contains details of a preferred embodiment. The reader should recognize that the present disclosure is not to be limited by the detailed information that follows. Rather, the detailed information is intended for purposes of illustration and not limitation. As can be seen in FIGS. 3A-3D and 4A and 4B the sleeve 14 and core 12, respectively, are made up in this embodiment of concentric cylinders 22, 24, respectively, although links of other shapes may be used, e.g. a dogbone configuration (not shown) as well as links of a type that are not concentric, e.g. backbone configuration (see FIG. 17), among others. The ends of the cylinders 22, 24 are not flat but instead one end 26 is an outer or convex hemisphere and the other end 28 is an inner or concave hemisphere, both with the same radius of curvature R. The cylinders 22, or links, of the outer sleeve 14 are chained back-to-back such that the concave end 28 of one mates with the convex end 26 of an adjacent cylinder. Similarly, the cylinders 24, or links, of the inner core 12 are chained hack-to-back. The result is a spherical-like joint, from a kinematic point of view. In the current embodiment, each link is able to rotate on the adjacent link's head, acting as a spherical joint with approximately 14 range of motion in any direction, although other ranges of motion are possible. The cylinders 22 have three channels 30 extending therethrough for control wires.

[0051] FIGS. 5A and 5B illustrate another embodiment in which the core 14 is steerable. The core 14 of this embodiment is comprised of cylinders 24 which have a convex 26 and concave 28 end. However, the cylinders 24 have three channels 32 for control wires.

[0052] The heads (i.e. the distal cylinders) of both the sleeve 14 and the core 12 are steerable using three cables which are attached at, for example, 120 from each other. As can be seen in FIGS. 3A-3D and FIGS. 5A-5D, there are three small cylindrical channels 30, 32, respectively, for wires to pass through. In the version of the device shown in FIG. 4A and 4B, the inner cylinder 24 has only one wire, in which case there is only one hole 34 through its center.

[0053] It will be appreciated that although the preferred embodiment described above utilizes cables or wires, alternative means of manipulating the limp elements, such as miniature pneumatic or hydraulic cylinders or other mechanical linkages situated between individual links, can be employed without falling outside the scope of this invention.

[0054] The links, and hence the HARP 10, can be made out of virtually any material, including plastic, which allows it to be used online with NMR. One current prototype of our device has an outer diameter of the outer sleeve 14 of 12 mm and an outer diameter of the inner core 12 of 6 mm. The choice of 12 mm is based on available port sizes. Each link of the outer sleeve 14 weighs 1.5 grams and each link of the inner core 12 weighs 0.5 grains. Currently, the number of links in each of the inner core 12 and outer sleeve 14 is seventeen. Therefore, the total weight of the device 10 is thirty-four grams and its total length is 300 mm. These dimensions are intended for purposes of illustration and not limitation.

[0055] As noted, the core 12 and sleeve 14 can be made rigid or limp using wires or cables. Although there are many variations, in the current prototype the outer sleeve 14 consists of a set of cylinders 22 strung on three wires. The three wires are 120 apart, making it possible to steer in any direction. This design provides a radius of curvature of approximately eight centimeters. When the wires are pulled towards the back of the sleeve 14, the cylinders 22 are pulled towards each other. When the pulling force increases, the friction force between adjacent cylinders 22 increases until the whole outer sleeve 14 stiffens (i.e. enters the rigid mode). When the pulling force is released, the outer sleeve 14 becomes limp. Thus, the wires together with their respective motors form a locking device. The motors, along with the electronics for controlling the motors, form a means for controlling the tension on the wire. When the outer sleeve 14 is positioned one cylinder in front of the inner core 12, and the inner core 12 is stiff, the distal link of the outer sleeve 14 can be oriented by pulling one or more of the three wires. The magnitude of the pulling force which is exerted on each wire can he controlled. By pulling the three wires with the same magnitude, the outer sleeve 14 becomes stiff without changing its shape.

[0056] The inner core 12, like the outer sleeve 14, consists of a set of cylinders. In contrast to the outer sleeve 14, the inner core 12 does not need (but may optionally have) a steering ability. The inner core 12 does need the ability to change from rigid mode, to limp mode, and back. Therefore, in embodiments where the inner core 12 need not be steerable, the links of the inner core 12 may be strung on a single wire, which enables a small diameter for the device 10.

[0057] One type of feeding mechanism 16, shown in FIGS. 6A and 6B, inserts and retracts the HARP 10 into and out of, respectively, a region of interest. The feeder 16 has two movable carts. A first cart 42, carried in a first fixed tray 43, drives the outer sleeve 14 while a second cart 44 carried in a second fixed tray 45 drives the inner core 12. Each cart 42, 44, and hence each of the inner core 12 and outer sleeve 14, is driven independently by separate linear actuators 46, 48 respectively. The linear actuators 46, 48 may carry shaft encoders (not shown) used for position control as is known.

[0058] Each of the carts 42, 44 carries one or more motors necessary for controlling the wires of the inner core 12 and outer sleeve 14. For example, as seen in FIG. 7 and FIG. 8, the cart 42 carries motors 50, 51, 52 which control the tension on wires 54, 55, 56 of outer sleeve 14. As shown in FIG. 9, second cart 44 has a motor 58 for controlling the tension on wire 59 of the inner core 12. Each of the motors 50, 51, 52 and 58 may be provided with shaft encoders (not shown) used for position control as is known. If the inner core 12 were to also be steerable, it too would require three motors,

[0059] For the 12 mm diameter HARP 10, the feeder's 16 dimensions are 400 mm (long) by 100 mm (width/height), while the HARP is 300 mm long. The 12 mm prototype HARP 10 is inserted into a protective plastic bag to achieve sterility. However, the device can be constructed out of inexpensive ABS plastic rendering the device disposable.

[0060] We selected motors to handle the worst case configuration for the device, i.e. when the motors tensioning the wires have to exert the most torque. The worst case configuration is when the HARP 10 is stretched out in a cantilever position, the outer sleeve 14 is limp, and the inner core 12 supports its own weight as well as the weight of the outer sleeve 14.

[0061] To estimate the axial force needed to be applied by the wire 59 of the inner core 12 to support this configuration, we use a simplified model of this extreme configuration. The simplified model is shown in FIGS. 10A-10C where we approximate the system parameters as follows: outer sleeve cylinder weight as 1.5 grams, inner sleeve cylinder weight as 0.5 grams, and the number of cylinders in each is seventeen. Therefore, the total weight of the device is thirty-four grams and its total length is 300 mm. Finally, the outer diameter of the outer sleeve 14 is 12 mm and the outer diameter of the inner core 12 is 6 mm. The choice of 12 mm is based on available ports and the 6 mm follows by design.

[0062] The weight of the device is simplified to a point mass at the center of gravity of the device. The largest torque is exerted on the area between the two proximal cylinders of the HARP 10. Therefore we developed a simplified model to include only one long cylinder that is in contact with the proximal cylinder. The wire of the inner core is applied with an axial force, F, at the center of the HARP 10. A free body diagram of the simplified model is shown in FIG. 10C.

[0063] The approximated relation between the force F and the torque applied on a circular surface with radius r and friction coefficient is shown in equation (1).

[00001]$\begin{array}{cc}=.Math.F.Math.r.Math.F=\frac{}{.Math.r}=\frac{50.Math..Math.N.Math.\mathrm{mm}}{.Math.3.Math..Math.\mathrm{mm}}\frac{17.Math..Math.N}{}& \left(1\right)\end{array}$

It is clear from equation (1) that the friction coefficient is an important design criteria. When the friction between cylinders is low, the pulling force that is needed to withstand the mechanism's own weight is enormous. To find the accurate friction coefficient between cylinders some empirical tests were needed.

[0064] Three different materials were tested: Aluminum T6061-T6, Garolite G11/FR5 and Garolite G10/FR4. The aluminum and Garolite G11/FR5 had a friction coefficient of approximately 0.2-0.3, but after a few minutes of being rotated under load, the contact surface was polished and smoothed out, and the friction coefficient dropped dramatically making these materials unfit for our design. The Garolite G10/FR4, which is a high pressure laminated glass reinforced epoxy, has a very high friction coefficient (approximately 0.5) and was durable to polishing. This material is also MRI compatible.

[0065] Based on these tests, we decided to use the Garolite G10/FR4. This material enabled the use of reasonable pulling force of the wire (approximately 35-40 N) to hold the weight of the entire device in the extreme configuration described above. Furthermore, this pulling force was sufficient to withstand additional torques caused by steering the distal link of the outer sleeve.

[0066] An implication of using the Garolite G10/FR4 was the need for a non-abrasive wire. Therefore we used the Spectra polyethylene fiber wire, with 0.030 diameter, a breaking force of 150 lbf, and a low stretch (about 3%). An additional advantage of the Spectra wire is its very tight radius curvature that enabled the use of a small diameter pulley (4 mm diameter) making it possible to achieve a high pulling force per torque.

[0067] FIG. 11 is a block diagram illustrating the components of a control system and the flow of information between those components. The feeding mechanism 16 interfaces with a control computer 62 through a bus conversion module 64. In the present embodiment, the conversion module 64 converts USB to I.sup.2C and back again. Outgoing data from the feeding mechanism 16 is input to the module 64 for conversion to the USB and is then input to a USB port 66 on the computer 62. Incoming data to control software 68 may include motor current data and motor encoder data for each of the motors in the feeding mechanism 16. Joystick data (position data) may also he received from a joystick 70. A monitor 72 may be responsive to video data from a camera mounted on the distal end of the outer sleeve 12 and/or inner cure 14 to provide visual feedback to a user regarding the position of the distal end of the HARP 10. The control software 68 may output motor current limit commands and motor position commands which are input to the feeding mechanism 16,

[0068] FIG. 12 is a block diagram of an exemplary electrical system which may be used for the feeding mechanism 16. A plurality of PID motor controllers 74 are used to control the various linear actuators 46, 48 and motors 50, 51, 52, 58. An example of one of the PID motor controllers 74 is shorn in detail in FIGS. 13A, 13B and 13C. The PID motor controller 74 is built around a PIC 18F series microcontroller 95. The motor controller 74 also features a quadrature decoder chip 96, an H-bridge chip 97, and a quad op-amp chip (TLV2374-IPW) having four amplifiers 98, 99, 100 and 101. The controller 74 also features a number of discrete components in support of these integrated circuits as well as three connectors, one connector 103 for the motor, and the other two connectors 104, 105 to create a stacking bus structure.

[0069] The quadrature chip 96 decodes the two channel encoder data from a motor encoder and outputs up and down clocks which are fed into the microcontroller 95. The microcontroller 95 uses its counter circuits to count the forward and reverse movements of the motor shaft to calculate the current position of the motor shaft

[0070] The H-bridge chip 97 is used to drive the motor. This chip takes a PWM signal and a direction signal as input from the microcontroller 95 and switches the motor on and off according to these signals. This chip handles the higher currents and voltages required by the motor which are beyond the capabilities of the microcontroller 95 outputs.

[0071] The third chip is the quad op-amp. This chip is used entirely for motor current monitoring. One amplifier 98 is used in a non-inverting configuration with a gain of 74.2 with its output fed to one of the microcontroller's 95 analog to digital pins. A second amplifier 99 is configured as a non-inverting amplifier with a gain of 500 and is also connected to an analog to digital pin of the microcontroller 95. This higher gain gives a more precise current measurement at lower current values. A third amplifier 100 is used to control a current source which drives an LED with current proportional to the motor current. This, along with an LED which indicates the direction of motion, gives the operator a clear visual indication of what the controller 74 is doing.

[0072] The microcontroller (PIC) 95 is the heart of this board. This chip features an I.sup.2C; bus peripheral which is used for communication with the host computer 62. This is a two way link used to send commands from the computer 62 to the controller 74, while status information flows back from the controller 74 to the computer 62. The commands from the computer 62 are related to position goals and current limits. The status sent back to the computer 62 includes motor electrical current measurements and motor encoder values.

[0073] The primary role of the microcontroller 95 is to run a PID position control loop. The purpose of this PID loop is to minimize the error between the motor encoder count and a dynamically generated position goal. This position goal is generated by another component of the microcontroller program, the trajectory generation system.

[0074] The trajectory generation system of controller 74 creates goals based on a final goal supplied over the I.sup.2C bus, and a desired time-to-goal value also supplied over the I.sup.2C bus, as is known in the art. The trajectory generator uses these values to create a trapezoidal velocity profile to bring the motor to the desired final position in the specified amount of time. This working goal is updated at approximately which is also the frequency of the PID loop.

[0075] Simultaneously, the microcontroller 95 is also taking readings of the motor current via the on chip analog to digital converter. These values are compared against an I.sup.2C bus supplied maximum current value, and if the measured current exceeds the desired maximum, the PWM output to the H-bridge is throttled hack. This also vans at the same frequency as the PID loop.

[0076] The controller 74 is designed to allow multiple controllers 74 to be stacked by utilizing a 40 conductor board to board connector 104, 105 (FIG. 13B) on both the top and bottom of the board. Because the system uses the multi-drop I.sup.2C bus, only two connections are needed between boards in addition to power. The majority of the conductors on these stacking connectors are used for power to overcome the single pin current limitations of small connectors like these.

[0077] Motors are connected to their respective controllers 74 through a 25 100 mil spacing right angle header 103. This is a versatile connector which allows ribbon cable or discrete wiring to the motors. The motors used on this system were supplied with ribbon cable IDC connectors.

[0078] Each board also features 5 LEDs, one of which is shown in FIG. 13A and the others shown in FIG. 13C, with the others connected to the microcontroller 95 as shown in FIG. 13A. There is a red LED (closest to the stacking connector) which is controlled by the motor current. Next is a yellow LED indicating the sign of the error from the PID loop, which indicates the direction the motor's shaft is being commanded to turn. The middle LED is currently unused and is available for future use. The green LED is used as a power/boot indicator and is under software control. Finally, another red LED (near the motor connector) is toggled each time an error is detected on the I.sup.2C bus to visually indicate bus reliability.

[0079] Before the system can be used, the first mechanism 12 and the second mechanism 14 must he homed; that is, their relative positions must be determined. This is done by retracting both mechanisms until the linear actuators 46, 48 have reached the end of their range of travel, a condition detected by the system as an increase in actuator current. The control software 68 uses encoder information to record the positions of the first mechanism 12 and second mechanism 14, and homing is complete.

[0080] The control software then puts the system in stepped advancing mode which is illustrated in FIG. 14. First, the first mechanism 12 is made rigid at 76. See also FIG. 15A. That is done by driving its tensioning motor 58 in the direction opposite the tension wire's 59 pull until the motor 58 stalls at its current limit. The second mechanism 14 is made limp at 78 by driving its tensioning motors 50, 51, 52 in the direction of the tension wires' pull by a fixed number of rotations, so that the wires become slack. The second mechanism 14 is advanced at 80 (See also FIG. 15B) so that it is positioned with its distal end one link's length beyond the first mechanism's 12 end. An encoder on the linear actuator 46 that is pushing the second mechanism 14 counts how far the second mechanism has moved.

[0081] At this point, the software 68 is ready for user input. The software 68 monitors the position of the joystick 70 at 82, translating the two-axis data from the joystick into the three-axis coordinate system of the second mechanism 14. The positions of the shafts of the motors 50, 51, 52 controlling wires 54 55, 56, respectively, are varied at 84 according to the translated joystick position. Once the user has steered the protruding link of the second mechanism 14 to the desired angle, the user presses a button on the joystick 70 to lock that angle in place as shown at 86. The three tensioning motors 50, 51, 52 are driven in the direction opposite the tension wires' 54, 55, 56 pull, until their current limits are reached and the motors stall at 88. This begins with the wire closest to the inside of the angle being formed, progressing after a small fixed delay to the next wire and finally the wire on the outermost part of the angle; tensioning the wires in this order preserves the user-selected angle more accurately than tensioning them simultaneously.

[0082] Once each of the three motors 50, 51, 52 has reached its current limit, the second mechanism 14 is rigid and the first mechanism 12 can safely be made limp as shown by 90 by driving its tensioning motor 58 a fixed number of rotations in the direction of the tension wire's 59 pull. The first mechanism 12 is then advanced by one link's length, so that its distal end is even or coextensive with the second mechanism's 14 distal end as shown by 92 and FIG. 15C. An encoder on the linear actuator 48 that pushes the first mechanism 12 is used to count how far the first mechanism 12 advances. Once in this position, the distal links of both the first mechanism 12 and the second mechanism 14 are at the same angle. The first mechanism 12 is then made rigid to preserve this angle, and the second mechanism 14 is made limp as shown by 76 and 78. In this way the process continues, steering the distal link, locking its position in place, and advancing in single-link steps until the first mechanism 12 and second mechanism 14 have been advanced to the desired length and path shape.

[0083] The description of motion in conjunction with FIG. 14 and FIGS. 15A-15C assumes that a camera, not shown, is positioned on the distal link of the second mechanism 14. Thus, the second mechanism 14 does the steering while the first mechanism 12 simply follows along. The process could of course be reversed with the camera being positioned on the first mechanism 12 such that the first mechanism 12 does the steering, while the second mechanism 14 does the following. In another embodiment, cameras could be provided on the distal ends of both the first mechanism 12 and the second mechanism 14 such that the two mechanisms leap frog one another. That is, rather than one of the mechanisms merely catching up with the other, after it is caught up, it can advance by one link because of its capability of being steered. The other mechanism then catches up and advances by one link's length because it too has the capability of steering. It is anticipated that with such a device, a desired position for the end of the HARP 10 may be obtained more quickly.

[0084] Although it is anticipated that steering may be achieved through the use of cameras, other mechanisms may be used. For example, the distal links of each of the first mechanism 12 and second mechanism 14 may be comprised of material which is visible through the use of x-rays, NMRI, or other such devices such that the HARP 10 may be steered by tracking the advance of the HARP 10 with such a device. Alternatively, a small amount of radioactive material may be placed on the distal end of each of the first mechanism 12 and second mechanism 14 such that the progress of the HARP 10 can be tacked. Control of the path and configuration of the HARP 10 may he accomplished through the use of intelligent algorithms. The present invention is not to be limited by the type of mechanism used to provide information for steering and/or guidance of the HARP 10.

[0085] The process of retracting the mechanisms takes place in the same way as advancing does, but in reverse order, and without steering input from the user as shown in FIG. 16. Once the user has selected Refracting mode using the joystick 70, the second mechanism 14 is retracted so that its distal end is even with the distal end of the first mechanism 12. The second mechanism 14 is made rigid, and the first mechanism 12 is made limp. Then the first mechanism 12 is retracted by one link's length, and the first mechanism 12 is made rigid. The second mechanism 14 is made limp, and again refracted until the distal end of the second mechanism is even with the distal end of the first mechanism 12. That cycle continues until the user presses a button to stop it, or until the first mechanism 12 and second mechanism 14 reach their home positions.

[0086] Turning now to experimental results, an off-the-shelf fiber optic based camera and video camera, such as the Olympus PF14 inset don tube 1.4 mm was introduced through the open central portion of the links of an inner core 12 designed to have a 6 mm diameter opening. Another option would be to integrate the video camera into the walls of the links of the outer sleeve 14, which are currently 3 mm thick. With such a device, and as shown in FIGS. 18 and 19, we were able to navigate the pericardial space of a porcine.

[0087] The normal pericardium is a double-layered, flask-shaped sac consisting of an outer fibrous envelope and in inner serous sac that is invaginated by the heart. The pericardial cavity or sac is a continuous virtual space that lies between the two opposite layers of serous pericardium. At the pericardial reflections and at the posterior wall between the great vessels, the pericardial space is apportioned into a contiguous network of recesses and sinuses; all pericardial reflections are located basally in relation to the great vessels. Thus, there are no obstacles during intrapericardial navigation along the anterior ventricular surface of the heart. There are three sinuses in the pericardial space: the superior sinus (also referred to as superior aortic recess), the transverse sinus contains several recesses between the major vessels (superior aortic, inferior aortic, right pulmonary and left pulmonary recess). The inferior aortic recess allows access to the epicardial aspect of the noncoronary and right coronary aortic cusps. The oblique sinus extends behind the atria, particularly the left atrium, in the region between the four pulmonary veins. There are five recesses of the pericardial cavity: superior aortic (SAR), inferior aortic (AR), postcaval (PCR), left pulmonary (LPVR) and right pulmonary (RPVR).

[0088] As shown, in FIGS. 18 and 19, we were able to explore the oblique sinus and visualize the posterior atrium, explore the transverse sinus with entry from the right and from the left with visualization of the left atrial appendage and explore all five recesses (SAR, IAR, PCR, LPVR and RPVR). Thus, the device and method of the present invention allow for various types of endoscopy or visualization of internal organs (or parts) or intracavity spaces from an on-board camera (or lens and light pipe) mounted on an articulated probe. It is the real time images provided by the on-board camera that allow teleoperation, i.e. remote controlled, 3D guidance in real time.

[0089] While the present invention has been described in connection with preferred embodiments thereof, those of ordinary skill in the art will recognize that many modifications, variations and substitutions are possible. For example, as shown in FIG. 20, the first mechanism 12 and the second mechanism 14 are shown in a spaced relationship, i.e. not one within the other. In such an embodiment, one or both of the mechanisms may have members for maintaining the spaced relationship as the device 10 moves. Accordingly, the present invention is not intended to be limited by the disclosed embodiments but rather is limited only by the scope of the following claims.