Discussion
We faced some issues in the hardware design while building the circuit for TIS G2. Initially we had used a 3.6V battery to power the LT1932 chip and 74AC14 chip at the same time. However, when we constructed the circuit on the breadboard and tested it, we found out that the PWM dimming part was not working. Then we checked the entire circuit, and found that the voltage across the circuit is 1V instead of 3.6V. After checking the data sheet, we found the PIN VIN limits the input voltage to 1V. For chip 74AC14, the minimum operation voltage is 2V. Thus, we separate the powers for two chips and use one 3.6 V battery for each chip. After soldering the entire circuit, we connected the negative pins of two batteries to provide the same ground voltage. For the probe frame part, we increased the size of the detection area to 3.8cm × 3.8cm, and used the nylon tube as a lighting source. In this part, the position of the nylon tube did affect the image. If the tubes were close to the surface of the probe, more light will distribute on the surface. Also, when we do the indirect contact experiment, the whole screen will light up. In that condition, we cannot capture the image of the target. So, we put the light tube close to the glass, and make a glue frame on the top of the tubes. The distance between the PDMS and the tube also affects the quality of the image. When we put the tubes 1cm away from the PDMS, the image is little better than putting the tube right next to the PDMS. However, putting the tubes 1cm away from the PDMS will increase the size of probe frame, which will increase the required force to get the image. Since the images for the two conditions just have a slight difference, we decided to put the light tube close to the PDMS.
Conclusion
In this thesis, we described a new design of the tactile imaging system. The objective for this design was to improve the TIS performance in estimating mechanical properties of soft tissue with inclusion. Therefore, we proposed, designed and implemented three hardware upgrades. First, the light source configuration was redesigned to fit the larger probe. Second, we incorporated a wide angle lens, with a small focal distance to improve the image resolution. Third, a constant current driver circuit with PWM dimming control was designed and implemented to provide the uniform brightness. In order to compare TIS G1 and TIS G2, we performed a mechanical properties estimation experiment on phantoms. The results showed that TIS G2 improved the depth and size estimation over TIS G1 by 5.78% and 0.96% respectively. Furthermore, TIS G2 was more consistent in differentiating hardness based on elastic modulus than TIS G1. Finally, we performed t-test on system outputs (minimal normal force and pixel number). The results show that TIS G2 is more sensitive and more reliable than TIS G1. With these results, we concluded that the new TIS design improved the performance significantly. For future improvement, force cell can be considered to replace the heavier force gauge, which will make the system lightweight and comfortable for the patients.
