This part discusses the actual implementation of the data acquisition with the integrated software and its comparison with two original software. According to the requirements of
Figure 3.28. The Error Log File records the error information when it happens at run time. The information is very detailed which includes the time, error code, location and the reason.
this project, the integrated software was tested with phantom imaging with three different modes: EIT, OPT and OPT/EIT. In order to compare the measurement data fairly, for every single test, we apply the same settings and the same sample for the integrated and original software.
3.5.1 EIT Data Verification
To verify that EIT mode from the integrated software produces EIT data as the original EIT software does, a PCB phantom (Thomas Johann Seebeck Department of Electronics, TalTech, Estonia) is measured to test EIT mode. The PCB phantom is a spiderweb-like array of resistors with 16 equally distributed connections to the device and it was made by Marek Rist, TalTech, Estonia. This circuit serves as a stable EIT phantom, however, it is not meant for rotational EIT system. In this experiment, PCB phantom was plugged to the EIT device, which generates the same content data for each rotational position.
Therefore, by comparing the results of the integrated software and original EIT software verifies whether the EIT data of integrated software was correct or not. The data was acquired for different data-record tables in continuous modes from two different software, original EIT software and integrated software for further comparison and analysis.
The original EIT software and the EIT mode of the integrated software have the same functionalities. However, the biggest difference is that the original one does not support the rotational motor movement function. Therefore, to demonstrate the integrated soft-ware is working as expected, we tested the functionalities without rotating the sample, but only with single and repeated data-saving approaches in continuous mode. To ensure the accuracy of the results, all the configuration parameters for the EIT device such as ex-citation level, voltage channel gain, and current channel gain must be assigned exactly the same. We determined that the excitation level is set to be 0.71V, the voltage channel gain is 10 and the current channel gain is 1, in order to get less noise data. Besides, the step period is 5 ms for both the original and the integrated software. In this test, the EIT raw data contains 420 tetrapolar measurements with 15 different frequencies. The device configuration parameters are the same as the previous test. The excitation level is 0.71V, the voltage channel gain is 10, the current channel gain is 1, and the step period is 5 ms.
Due to the issue that the original EIT software does not include motor movement function-alities, it is hard to calculate the time costs for it. However, the time costs of the rotational EIT mode from the integrated software can be divided into two parts: EIT data acquisition time and motor rotation time. The EIT data is collected after every one cycle in the single saving mode, therefore, the EIT data acquisition time is the multiply of step period and table steps[3]. So the total time costs in seconds can be described in the formula 3.1:
T ime = 360
◦EIT Angle ∗ (1 + T able Steps ∗ Step P eriod) (3.1)
The raw data collected consists of complex impedance values. To visualize EIT image, data must be reconstructed with third- party software. With our own algorithm[33], MAT-LAB(MathWorks) R2019a is used for the EIT image reconstruction and visualization.
3.5.2 OPT Data Verification
To validate the OPT functionalities, the integrated software should be able to provide an OPT imaging with a flexible rotational angle, however, the original OPT software only can support 0.9 ◦ angular rotation, thus 0.9 degree rotations were used to verify OPT data to be of good quality. Same settings were configured for both integrated and original software. Therefore, in bright-field OPT test, the camera was configured with exposure time 0.025 seconds and binning value 1. However, in fluorescent OPT test, the exposure time was 0.5 seconds and the binning value was 1. Meanwhile, a 10X objective chosen for bright-field OPT and fluorescent OPT.
We conducted tests for both bright-field and fluorescent OPTs by using an exact same sample placed in fluorinated ethylene propylene tube and the sample included 15 µm fluorescence beads in 1% agarose hydrogel immersed in distilled water. The LED light used in the fluorescent OPT.
The time costs for OPT mode of the integrated software includes the motor rotation time and imaging time which counts only the exposure time. Because the integrated software handles the imaging saving asynchronously, the software does not need to wait until the image is saved. The motor also spends one second moving to the next data collection point. Therefore, the OPT mode time costs(s) is estimated in the formula 3.2:
T ime = 360
◦OP T Angle ∗ (1 + Exposure T ime) (3.2)
The OPT projection data were 3D reconstructed in MATLAB R2019a using a FBP algo-rithm and visualized in Fiji [34] software.
3.5.3 OPT/EIT Mode Test
The last but most important thing is to validate the OPT/EIT mode is functioning properly.
The OPT/EIT data acquisition is the key to prove the novelty and success of the integrated multimodal imaging software. For the OPT/EIT test, the bright-field OPT and rotational EIT were selected. The schematic layout for the OPT/EIT mode is displayed in figure 3.1 in chapter 3.1. The main components are a bright field LED (a), a sCOMS camera, a spectro-EIT device and sample manipulation platform. A rectangular piece of carrot sample (1.4x1.4 mm in the X-Y plane) is used as a phantom in the experiment. The sample is rotated in the center of the imaging chamber that is filled with saline solution and eight electrodes evenly distributed in both opposing sides. The XY plane size of the imaging chamber is 6.6x9 mm, and the height is 5 mm along Z-axis.
Then we configured the integrated software with same OPT parameters and planned EIT settings to enable acquire the data from two imaging methods simultaneously. For bright-filed OPT, the camera was configured with 0.013 seconds for exposure time, binning value 1, 2X objective and the rotational angle is 0.9◦. For EIT device settings, the excitation level is 0.59V, voltage channel gain is 10, current channel gain is 1, and the step period is 10ms. The data acquisition angle is 4.5◦and table 420.
For this OPT/EIT test, the total time costs would be the sum of OPT time and EIT time.
As we mention in the formula 3.1 and 3.2, in addition both data acquisition methods uses the same rotational motor, therefore the total time of OPT/EIT calculation is displayed in the formula 3.3:
T ime= 360◦
OP T Angle ∗ (1 +Exposure T ime) + 360◦
EIT Angle ∗ T able Steps ∗ Step P eriod (3.3) It is same as in solo OPT and EIT modes, the OPT data was reconstructed with FBP algorithm in MATLAB R2019a and visualized in Fiji. The EIT data reconstruction [33] and visualization were done in MATLAB R2019a.
4 RESULT AND DISCUSSION
This chapter assesses the user interface design, the results of the measurement data and the time costs performance of the integrated software. The results coming from EIT, OPT and OPT/EIT will be explained respectively. It also discusses the usability of the integrated software in detail, the limitation of this software, the challenges we faced during the software development and also the possible future work.