Robotic platform and path planning algorithm for in situ bioprinting

Current in vitro bioprinting methods are mostly based on scaffolds, which have limitations such as fragile structure, high risk of contamination, and mismatch of shape and defect sites. Carmelo De Maria’s team from the University of Pisa published an article titled “Robotic platform and path planning algorithm for in situ bioprinting” in the journal Bioprinting. The team designed a robotic in-situ bioprinting platform called IMAGObot that injects biomaterials directly into damaged sites, capable of fabricating 3D structures on irregular surfaces.



Background introduction


Currently, scaffold-based tissue engineering is
clinical
There are some limitations in the application. Handling and implanting 3D tissue in vivo may result in: (1) disruption of micro and macro structures, (2) risk of contamination due to transport and artificial implantation, and (3) the requirement for a highly sterile environment. In addition, inaccurate design inputs due to the resolution limitations of computed tomography or magnetic resonance imaging scans can cause the shape of the fabricated structures to differ from actual defects.


In situ bioprinting is a solution to the above problems. It delivers biological material directly to the damaged site following a predefined path. Currently, in situ bioprinting methods are mainly divided into handheld and robotic. Handheld in situ bioprinting is flexible and simple to operate, and can easily fabricate simple structures. Robotic in situ bioprinting can print a variety of biological materials and has the ability to reconstruct complex tissue levels. On the other hand, robot-based in-situ printing methods have more than 3 degrees of freedom and involve less human intervention compared to hand-held ones.


This study aimed to investigate the potential and limitations of a 5DOF robotic arm as an in situ bioprinting platform. The use of 5DOF ensures a larger workspace; at the same time, compared to the traditional 3DOF3D bioprintingmachine that allows material to be deposited on curved and non-smooth surfaces, enabling complex geometries at defect sites to be repaired by precise and continuous bioink deposition


Materials and Methods


The robot platform is developed based on BCN3D’s 5DOF open source robot MOVEO (Fig. 1A), and its mechanical structure adopts3D printing(
FDM
), the electronic hardware is based on an Arduino board.


1. Hardware


The main modifications to the original MOVEO hardware are as follows:


(1) The original end effector (gripper) was replaced by a syringe pump extrusion module (Fig. 1B)


(2) Some connection parts were redesigned to accommodate the optical incremental encoder (Fig. 1C)


(3) The encoder is mounted on each axis, and a plug-and-play connector is added for easy maintenance (Fig. 1D)


(4) Open source for use with LinuxCNC
software
The electronics were upgraded (Figure 1E)

2. Software


In LinuxCNC, the ini file contains the basic configuration of the robot, such as name, firmware version, number and properties of axes, etc. In order to adapt to the IMAGOBot robot, 5 axes are defined in the ini file, that is, one for each joint. The kinematics of the robot are set to trivial, which means that each axis in the software corresponds directly to a physical joint. This makes it possible to set the angle of each axis in g-code to control the robot, effectively using the kinematics module external to LinuxCNC. All axes are defined as rotational motion in degrees. In addition, the following main parameters were set for each axis: maximum/minimum speed and acceleration, travel range, homing position and behavior (i.e., homing speed and homing sequence), and motor drive settings, as shown in Table 1.

The ini file also contains the PID controller settings, the specific parameters are as follows:


・ P, I, D values


・ Feedforward parameters FF0, FF1 and FF2


・ Output offset amount BIAS


・ Dead zone DEADBAND


・ Maximum output MAX_OUTPUT


Additionally, an Arduino UNO board for controlling the pressure regulator is integrated with a LinuxCNC controlled robotic arm using HAL logic.






3. Path planning


To manage all stages of the bioprinting process, a path planning algorithm was developed in Matlab, as shown in Figure 2. The algorithm can be formulated as follows: Project the generic print pattern onto the surface representing the print area, and extract the local area-related coordinates and corresponding normal vector for each intersection point. For each point, inverse kinematics is used to evaluate the joint angles of the manipulator and constrain the end effector to the normal direction.

For a faster and more intuitive use of the algorithms, a graphical user interface (GUI) was developed using Matlab App Designer, where all the algorithms described earlier can be managed and the simulation printing process can be visualized, as shown in Figure 3.

4. Printing performance evaluation


Replace the extrusion pump at the end of the robotic arm with a marker and place a piece of graph paper on the print platform to evaluate the repeatability and highest resolution of the system.




Repeatability test: Print a series of dots at known locations three times and measure how much each test deviates from the previous one. The print path consists of 9 dots arranged in a square pattern, 20mm apart from each other. The robot ascends 10mm in its path from one point to another, touching the platform at Z=0 at each point. Another test was then carried out, printing a series of dots (11 dots 5 mm apart) along the two main directions (x and y), measuring how far off the ideal straight line was and how perpendicular the two lines were. This test is repeated twice.




Resolution Test: Print parallel lines at decreasing distances (5 mm, 2 mm, 1 mm, 500 μm, 200 μm, 100 μm) to determine the minimum distance at which the lines can be resolved. Each round of testing was repeated at 10, 20 and 30 mm/s, and images were acquired and analyzed using Matlab.




5. Preliminary testing of in situ bioprinting


30% w/v Pluronic Acid (
Sigma
-Aldrich, Italy) conducted preliminary tests of in situ bioprinting in deionized water, a hydrogel for extrusion bioprinting, extruded onto different irregular surfaces. The test was performed at a line speed of 10 mm/sec, which is typical for bioprinting applications (range 2-20 mm/sec). Three substrates were designed with different slopes and curvatures to simulate a defective humeral head, and in situ bioprinting trials were performed on them.




result


1. Repeatability test


Print tests for analysis of repeatability were performed in triplicate and images were acquired after each test (Figure 4A is the result of the third trial). The origin of the reference frame was fixed at the center point of the pattern, and Matlab was used to calculate the coordinates of 9 points (the results at 10 mm/s printing speed are shown in Table 2). Tests at all print speeds yielded similar results.

Table 2 Repeatability test (printing speed 10 mm/sec):


Coordinates of 9 points on the printed pattern (mm)




2. Collinearity and vertical testing


The red legend in Figure 4B shows the origin and XY axes as reference axes for the Matlab analysis. Fitting the data points using a linear regression method yields two straight lines. For each test, the correlation coefficient R2 was also calculated to evaluate the quality of the model, as shown in Table 3. The minimum R2 is 0.71, so in the worst case the linear model is sufficient to represent the data.

Table 3 Collinearity test (printing speed 10 mm/sec):


R2 value for each trial




From these data, the angle between the two printed lines and the angle between the printed line and the system reference line were calculated, as shown in Table 4. Similar results were obtained for all tested print speeds.

Table 4 Squareness test (printing speed 10 mm/sec):


Evaluate the angle between the print line and the main direction




3. Resolution test


As shown in Figure 4C, the minimum spacing at which lines can be resolved is 200um. Therefore, the robot is able to print at a resolution of at least 200um. Similar results were obtained for all tested print speeds.




4. Preliminary testing of in situ bioprinting


The first printing trials were carried out on a scaffold with three different ramp zones. As shown in Figure 4D, the extrusion axis was always kept perpendicular to the scaffold surface, demonstrating the reliability of the algorithm.




Two other tests were performed on two different surfaces: the first had areas with dramatic changes in slope (Fig. 4E), and the second had a gradual and continuous curvature over the entire surface (Fig. 4F). Finally, in situ bioprinting experiments were performed on the bone model, as shown in Figure 4G. In all cases, the algorithm proved to be robust, ensuring continuous extrusion of the material.

Summarize


This paper presents the design and development of a robotic in situ bioprinter and path planning algorithms to control all stages of the bioprinting process. Higher degrees of freedom open up the possibility of printing materials on irregular surfaces. In this context, in situ bioprinting may become a reality in the near future, especially for the most accessible organs, such as skin. At the same time, we must also face new challenges of safe human-computer interaction:
surgical
Doctors and robots will be
Operation
Collaboration in Room 4.0. Having a “collaborative” bioprinter can assist surgeons during the surgical phase, allowing for more precise interventions and minimizing human error.




references


Gmfa B , Gr A , Afba B , et al. Robotic platform and path planning algorithm for in situ bioprinting[J]. Bioprinting, 2021.




https://doi.org/10.1016/j.bprint.2021.e00139

(responsible editor: admin)