Ex vivo validation of magnetically actuated intravascular untethered robots in a clinical setting

Abstract

Intravascular surgical instruments require precise navigation within narrow vessels, necessitating maximum flexibility, minimal diameter, and high degrees of freedom. Existing tools often lack control during insertion due to undesirable bending, limiting vessel accessibility and risking tissue damage. Next-generation instruments aim to develop hemocompatible untethered devices controlled by external magnetic forces. Achieving this goal remains complex due to testing and implementation challenges in clinical environments. Here we assess the operational effectiveness of hemocompatible untethered magnetic robots using an ex vivo porcine aorta model. The results demonstrate a linear decrease in the swimming speed of untethered magnetic robots as arterial blood flow increases, with the capability to navigate against a maximum arterial flow rate of 67 mL/min. The untethered magnetic robots effectively demonstrate locomotion in a difficult-to-access target site, navigating through the abdominal aorta and reaching the distal end of the renal artery.

Fig. 1. The untethered magnetic robot (UMR) can swim inside the natural pathways of a porcine aorta model under controlled conditions, enabling interventions and retrieval with minimal incisions.

A The wireless actuation and non-invasive localization of UMRs are achieved through a robotic platform, consisting out of an external rotating magnetic field and a C-arm imaging system. UMRs navigate both with and against the blood flow for various interventions. B, C A 9-mm-long UMR moves both with and against the blood flow inside the abdominal aorta and is then guided to swim within the left renal artery. The UMR’s location is highlighted by squares in the cone-beam computed tomography scans, and its trajectory is depicted by the yellow arrow (Movie S1).

Fig. 2. Ex vivo trials are conducted using a porcine aorta model.

A The ex vivo organs are harvested and connected to a circulation pump prior to each motion control session. B Placement of an intra-aortic 3D-printed filter allows the untethered magnetic robot (UMR) to remain inside the aorta irrespective of the ongoing blood circulation. C–F Wireless magnetic actuation and C-arm imaging systems enable control and localization, respectively. G, H Optical examination of the internal wall lining of the vessels shows no risk of damage by the UMR. The black squares indicate the UMR. I–L Hemocompatibility tests included I protein fouling, J biofilm formation, and K, L coagulation evaluation.

Fig. 3. UMRs are employed to navigate vascular pathways for interventions.

A X-ray fluoroscopy images are gathered online to detect the UMR, the rotating permanent magnet (RPM) actuator, and the physical surroundings using clinically relevant radiation settings. B Our robotic platform consists of a C-arm imaging and permanent magnet robotic systems. C To show the shape of the UMR, a radiocontrast agent is injected into an in vitro model. D, E The process of steering and maneuvering the UMR within the left renal artery is achieved by manipulating the rotation axis of the RPM about the z axis.

Fig. 4. The movement of the untethered magnetic robot (UMR) is influenced by several factors, including the constraints imposed by the ex vivo model’s confinement, the viscosity of the blood, and the external control inputs applied.

A, B The actuation of the UMR is evaluated by utilizing ultrasound images to identify an optimal gap between the rotating permanent magnet (RPM) and the UMR. This gap is determined to achieve enough RPM clearance while minimizing contact with the inner wall of the lumen. C Frequency response of the UMR is characterized in blood. The shaded region represents the standard deviation. D Prediction of swimming speed of a UMR in blood and through a clot. The speed of the UMR, U, is normalized with swimming speed, U_N, in a Newtonian fluid (water). De and β are the Deborah number and ratio of serum to blood viscosity, respectively. E The UMR’s swimming speed is influenced by the normalized wavenumber ν = ν^*R_cyl and diameter of the surrounding vessel (2R_ves). The white and blue markers indicate the small and large UMRs used in our study, which share the same normalized helical pitch (2π/ν) but differ in their ratios of R_cyl/R_ves.

Fig. 5. The untethered magnetic robot (UMR) is directed in a controlled manner both against and with the direction of arterial flow, maintaining movement below its step-out frequency.

The UMR is actuated against (A) and with (B) arterial flow at actuation frequency of 9 Hz. C The UMR exhibits greater lateral displacement when swimming against the flow (Movie S2).

Fig. 6. The swimming velocity of the untethered magnetic robots (UMRs) is assessed within the range of blood flow rates from 15 to 67 mL/min.

A, B To achieve consecutive straight runs at an actuation frequency of 9 Hz for each flow rate, the UMR is moved under controlled maneuver. The average speeds are determined based on data collected from five separate trials. C The robotic platform effectively maintains the UMR’s position against the highest blood flow rate of 67 mL/min (Movie S2).

Fig. 7. The untethered magnetic robot (UMR) is remotely operated to navigate through the abdominal aorta, engaging in a series of straight runs before executing a turning maneuver within the left renal artery.

A The rotating permanent magnet (RPM) actuator is teleoperated to exert in-plane torque required to steer the UMR toward the left renal artery. B The UMR is extracted from the left renal artery and no damage in the wall lining is observed. C A four-stage sequence is executed to transition the UMR from its location in the abdominal aorta to the renal artery. The small circles on the visual representation denote instances where the visual feedback of the UMR is obscured by the RPM actuator. The black-dashed arrows indicate the position of the UMR, and the pink arrows indicate the direction of blood flow. The white and red circular arrows indicate the direction of rotation of the RPM with respect to its rotation axis and the z axis, respectively.

Fig. 8. The 9-mm-long untethered magnetic robot (UMR) is controllably moved back and forth between the abdominal aorta and the proximal end of the left renal artery.

A A cone-beam computed tomography scan shows the xz plane of the renal bifurcation. B Motion control is achieved in a stationary blood flow ($\stackrel{°}{Q}~0$). C Motion control is achieved in blood flow of 35 mL/min.