Redesign of the control system of a cyclic stretch bioreactor for cardiac tissue engineering and development of a computer vision software for characterizing beating of engineered cardiac tissues

Cardiovascular diseases (CDs) are the major causes of death in first world countries. Specifically, myocardial infarction (MI), commonly known as heart attack, causes irreparable loss of cardiomyocytes and impairment of the heart muscle. Over time, this leads to over stressed non-impaired regions, weakening the myocardium through fatigue. In this context, cardiac tissue engineering (CTE) aims to develop functional engineered cardiac tissues (ECTs) to be used as functional in vitro models for cardiac research or drug testing, with the ultimate but still challenging goal of repairing the damaged heart tissue. Multiple studies have shown that physical stimuli are crucial for the development of functional engineered tissues in vitro. In particular, tensile stimulation increases receptivity of cultured ECTs. ECTs can be stimulated in vitro within dynamic culture devices, known as bioreactors, that provide biomimetic culture conditions. In this work, the control system of an existing bioreactor for cyclic stretch culture of ECTs was optimized. In detail, the bioreactor power supply circuit was updated from a multiple to a single power source. Then, the control algorithm, on which the simulation accuracy relies, was modified from a previous open-loop configuration into a closed-loop quasi real-time configuration by integrating an optical encoder. Specifically, the developed control algorithm allowed an easy integration of different stretching waveforms, such as triangular form, which was added to the preset sinusoidal one. Also, the bioreactor user interface was improved by adding new features (e.g., the elapsed stimulation time) and making the respective hardware modular. In parallel, a computer vision software was developed in MATLAB to analyze acquired videos of beating cardiomyocytes utilizing Finite element method (FEM). Such software, including a variety of segmentation techniques (e.g., thresholding, adaptive thresholding, gradient methods), allows the video processing through a user-friendly interface which enables to modify the parameters adapting to multiple segmentation scenarios depending on investigated tissue type and conditions of capture. In addition, the software was interfaced with a predeveloped digital image correlation software in order to exploit additional functionalities such as FEM analysis and data plotting. Lastly, validation tests were performed on the stimulation control algorithm by using a linear position sensor to visualize the generated stimulation motion, whereas the developed segmentation software was used to analyze videos derived from two separate biological experiments on the stimulation of cardiac tissue. Results from validation tests demonstrated the absence of motion drift for long-term functioning (up to 3 days) and the correctness of the stimulation waveforms. Furthermore, the computer vision software allowed to appreciate and quantify the effect of ECT stimulation showing a larger strain than the control tissue sample (not stimulated).