Z1 - Characterizing Force for the µSiM Vasc-SkM Microphysiological System to Model Vascular-skeletal Muscle Interactions in Rheumatoid Arthritis

Saturday, October 11, 2025

10:00 AM - 11:00 AM PDT

Location: Exhibit Hall F, G & H

Presenting Author(s)

Ananya Sridhar

Undergraduate Student
Duke University, United States

Primary Investigator(s)

GT

George Truskey

Professor
Duke University
Durham, North Carolina, United States

Co-Author(s)

Mingzhi Xu

PhD Candidate
Duke University
Durham, North Carolina, United States

Introduction: : Rheumatoid arthritis (RA) is a chronic, inflammatory autoimmune disease in which the body’s immune system targets its own tissues. RA affects about 1% of the world’s population [1], and its symptoms include joint pain, muscle weakness, and muscle wasting [2]. In order to better understand and treat autoimmune disorders like RA, researchers can create in vitro models with cells donated by RA patients, such as myobundles made from primary myoblasts. These myobundles are placed in chips providing perfusion, allowing them to be relatively stable and better preserve their native phenotypes.

One important feature of engineered muscles is their contractility. Previously in our lab, we fixed one end of the myobundle in a tissue culture well and the other end on a force transducer, and then the myobundle was stimulated with carbon electrodes to induce contraction. Currently, we are developing a microphysiological system that integrates a blood vessel mimic and a myobundle, the μSiM Vasc-SkM system. The myobundle is attached to flexible PDMS beams, so that this force can be determined through beam deflection and the elastic modulus of PDMS.

Hence, a mechanism for determining the force exerted by the muscles on the beams is needed. The goal of this project was to develop an automated program that can analyze videos for beam deflection to determine the force exerted on the beam by the myobundles, and organize the data to produce force tracing plots and spreadsheets with key parameters such as max force and time to half relax.

Materials and

Methods: : Two different MATLAB programs were written for detecting the deflection of the beam and therefore the force exerted by the myobundle. Both programs divide a video of the muscle contraction into frames and then direct the user to drag a box around a rectangular region containing the line of interest. The first program then detects this line by looking at the darkest region of the image, assuming that the darkest region is where the beam is located (referred to as the RGB method). The second program uses the subPixel method [4] to detect the line in that region of the image.

Both programs will then output the coordinates, along with an image of the frame with the detected line drawn. These programs also save the coordinates, the frame’s time stamp, and calculate the force exerted so that these values can be analyzed at a later point in time. These programs will then plot the coordinates of the force generated against the time values and a curve is fit to the data using a two exponential model, modelling a sudden and sharp spike in the force followed by an exponential decay.

Results, Conclusions, and Discussions:: Figure 1 displays the myobundle stretched between the beams on the hMoC. The silver ribbons are the Pt electrodes on the glass under the hMoC for muscle stimulation. Figure 2 shows examples of the frames to be analyzed and the detected edge of the two different methods. In 2.A, the beam’s edge is very distinct and easy to identify, but 2.B has a blurrier edge. The program must be able to successfully detect both types of edges. 2.C shows the edge detected by the RGB method, which appears highly accurate. 2.D shows the edge detected by the subPixel method, which is much less accurate. Based on these results, the RGB method was selected for further analysis and improvement.

Figure 3 shows the various methods used to generate a line of best fit through the data. 3.A and 3.B show a line of best fit generated for the RGB and subPixel data respectively. This line was created using a cubic spline. However, this line misses the maximum force generated (seen in 3.A) and the data generated by the subPixel method resulted in an inaccurate line of best fit. 3.C uses a different spline tool built into MATLAB [5], which generates a spline through data collected with the RGB method. This spline is an improvement on the one in 3.A, as it includes the maximum force value, but the rest of the curve is not smooth. 3.D shows the current solution, which takes the force values calculated from the RGB displacement values and fits them to a double exponential model.

To create a more automated method of determining the force exerted on the beam by the lab-grown myobundles, a program was written using MATLABusing the RGB method, which looks at the darkest area of the selected region to identify the beam’s edge. After iterating through each frame of the video and determining the coordinates of the edge, the double exponential model is used to create a line of best fit that mimics the force exerted by the muscle.

Acknowledgements and/or References (Optional): :