Ultrasound is sound waves with frequencies above the upper audible limit of human hearing, which is approximately 20 kHz. The ultrasound (US) wave interacts with tissue and reflects back depending on the properties of the tissues such as velocity of sound in the tissue and its density, which can be modeled using wave equation. Higher frequencies have better resolution (shorter wavelengths) but cannot penetrate as deep into the tissues. In diagnostic ultrasound frequencies above 1 MHz are used. For better tissue penetration frequencies of interest for our purposes are between 0.3 to 1 MHz yielding sufficient resolution and penetration at the same time Speed2001 [1]. Latest studies show that it is possible to use pulsed ultrasound for neurostimulation, ultrasound therapy such as treatment of pain and the repair of various tissues, however without clear knowledge which receptors or cell types we are targeting, and what is the mechanism of the ultrasound stimulation effects Vasquez2014 [2].

For the simple setup of the ultrasonic stimulation of cells we initially used ultrasonic baths (1) that are used to clean the laboratory equipment, small devices to clean jewelry or ultrasonic cell disruptors, which however offer little control over the intensity, frequency or pulse shapes and numbers of repetitions and are not appropriate to monitor activation of mechanosensors under the fluorescence microscope. However, one member of the team is a student of electrical engineering and this was the right challenge for him.

For the generation of specific shapes of ultrasound pulses researchers usually use a setup consisting of two signal generators (one for switching on and off the train of US pulses and the other one to produce the sinusoidal signal of suitable frequency). This signal is further fed to the amplifier and then to the ultrasonic transducer. Furthermore, an ultrasonic sensor is required to evaluate and control the magnitude of the ultrasound. Usually a hydrophone is used in combination with an amplifier and an oscilloscope. This setup is complex to establish and difficult to use. Therefore, the goal of a part of the iGEM 2016 group (student of electrical engineering) was to develop a device (named Moduson), which would be capable of providing appropriate signals required to perform specific ultrasonic experiments in a single apparatus and as such easy to use. The research and development of the device was performed in the Laboratory for Bioelectromagnetics/Faculty of Electrical Engineering/ University of Ljubljana under the advisor prof. Dejan Križaj and a company Noeto.

Adaptability

The device should be designed to be as flexible as possible in order to be capable of delivering a wide range and different shapes of stimulation signals to stimulate cells under the microscope, cells in a petri dish, cells in a microplate immersed into a bath and to stimulate animals. In order to fulfill this requirement we selected a dedicated embedded measurement card Red Pitaya acting as an embedded computer. This embedded device is based on Linux system and can be completely customized to the user’s needs. Furthermore, it can be Wi-Fi controlled so the final application is based on modern programming tools such as JavaScript, C++, Html, etc. The final application is basically a web page, accessible with any computer capable of the Wi-Fi connection. This application gives complete control of all stimulation parameters and operation of the device. Another advantage of the platform used its capability of simple integration with Matlab through so-called SCPI commands, usually used in the instrumentation for easy control and data acquisition making the device perfect research and development tool. Simple example of SCPI commands in Matlab for pulsed bursts is shown below:

On the Figure 2 a typical sequence of a stimulation signal constructed of a set number of sine waves of defined frequency that is repeated for required number of times with selected repetition period is shown.

Simple graphical user interface

Users of the device do not need to be computer experts and do not need to have knowledge on the signal generators, amplifiers and oscilloscopes. Therefore, the motivation was to design and develop a simple user-friendly graphical interface with all the relevant parameters that need to be set in order to run the experiments from the Web page. To accomplish this, dedicated software was written that enables users to design and run the experiment and evaluate the results. Figure 3 presents a developed user interface. On the right side (US burst settings -> STIMULATION PULSE) we can set parameters, such as the amplitude of the signal, frequency and number of sine waves. Below these settings, repetition parameters of the signal (PULSE REPETITION) such as the number of pulse repetitions and its frequency can be set. When all parameters are set, the signal can be previewed by clicking the SHOW BURST button. The pulse sequence is initiated with the START button. The acquired signal from hydrophone is seen on the main plot. At the top of the interface, there are options to export the image of a graph and csv data of the acquired signal. Chosen values of parameters can also be saved for the next use of the device.

Signal amplifier

Once a suitable signal is generated, it needs to be amplified from a signal level (amplitude of 2 V) to several hundred volts or even above 1000 V, which poses a serious engineering challenge. For this purpose, we first designed a circuitry on a conceptual level and simulated it with LT Spice (4). In the simulations an electric model of a transducer was considered as a combination of capacitors, inductors and resistors wired in parallel (5). This model enables simulation of ultrasonic transducers with several resonant frequencies. Simulations show that our device can reach the power at transducer around 92W, which is close to the measured 86W in the real system.

We built the circuitry with real elements on a prototype board where several additional alterations were required, as the real elements do not have ideal characteristics as the ones included in the simulation. Another challenge was the design of a suitable transformer to increase the voltage amplitude of the signal. The transformer is also required to operate as an impedance transformer. Lowering impedance of the ultrasonic transducer enables higher current levels and consequently higher power values.

Evaluation of the developed device

Functionality of the Moduson was first tested without any connected load. Figure 7 presents a typical output measured with an oscilloscope for a signal of frequency 310 kHz.

Developed device was tested with several transducers with resonance frequencies ranging from 300 kHz up to 1 MHz. Most experiments were performed with an unfocused transducer Olympus V318-SU (8) with a waterproof case, allowing it to be used in in vitro experiments.

Figure 9 presents measured electric power using the transducer V318-SU. After partial compensation of reactive part, the measured real power reached 86 W. This result exceeded the set requirements.

With compensation, voltage was considerably increased at contacts of transducer and reached around 900Vpp as shown in Figure 10.

We measured the emitted ultrasonic pressure indirectly with a hydrophone RP 31l shown in Figure 11. This hydrophone has a sensitivity of 50mV/100kPa at a frequency of 310kHz. It can be connected directly to the oscilloscope input to detect the pressure of the ultrasonic waves. Pressure measured 4 mm from the transducer reached 7 kPa which gives the intensity of 33 W/cm2.

Example of an output signal measured by a hydrophone is presented in the Figure 12.

Set-up for experiments with ultrasonic pulses

After we successfully tested the Moduson device, we aimed to design measurement set-up for stimulation of cells cultivated in plastic 6-well plates, which allows insertion of the ultrasound transducer. In order to ensure repeatable conditions in every experiment the position of a transducer was fixed with a suitable holder. Several models of holders were designed for different experimental configurations and a 3D-printer was used to fabricate a dedicated holder (13). This allowed positioning of the transducer at the fixed height above the bottom of a well, which was crucial due to the series of maximal and minimal intensity of generated pressure.