Response Time of a Thin-Film Resistance Temperature Detector (RTD) for High-Intensity Focused Ultrasound (HIFU) Phantom Applications

Featured Application

The thin-film RTD array sensor quickly detects a sudden thermal change inside a tissue-mimicking material (TMM) to validate high-intensity focused ultrasound (HIFU) devices.

Abstract

The thermal response time of a thin-film resistance temperature detector (RTD) array sensor was measured for a high-intensity focused ultrasound (HIFU) phantom. As the temperature inside materials change rapidly within several seconds, it is important to have a temperature sensor with a fast response time to evaluate their performance. However, previous methods for measuring thermal response time were not suitable for thin-film sensors, and there were no quantitative data available. In this study, we used a liquid drop method to measure the thermal time constant of the thin-film RTD, which was found to be 1.0 ± 0.2 ms. This indicates that the thin-film RTD array sensor has a sufficiently fast response time to detect sudden temperature changes inside the tissue-mimicking material (TMM) for validating HIFU devices.

1. Introduction

High-intensity focused ultrasound (HIFU) is a technique that utilizes ultrasonic waves to focus and deliver mechanical energy into a small volume inside a material [1]. It is a non-invasive therapeutic technology used in clinics as an alternative or complement to conventional surgery. The mechanical energy delivered into the targeted tissues can stimulate or ablate the tissues. For example, HIFU had non-invasively permeabilized the blood–brain barrier to deliver drugs into brain tissue [2]. More importantly, HIFU can be used to ablate various benign and malignant tumors by applying temperatures above 60 °C [3,4,5,6].

In the case of using HIFU therapy, it is important to raise the temperature at the target as quickly as possible to reduce thermal damage to the surrounding tissues. If HIFU is exerted for a long time, nearby non-targeted normal tissue is damaged by the transferred heat energy. For the validation of these HIFU treatment devices, a tissue-mimicking phantom with an inserted thermal probe array has been developed [7,8,9]. The phantom should have acoustic properties that are similar to those of soft tissue, such as speed of sound and attenuation coefficient. To minimize ultrasound reflection inside the phantom and make it transparent to ultrasound propagation, thermal probes such as thermocouples [9] and thermochromic material [10] have been used instead of conventional PT-100 probes. The thickness of the thermal probes needs to be much thinner than the wavelength of ultrasound (1.57 mm for 1 MHz ultrasound frequency). Non-metallic thermal probes are also needed to minimize the acoustic impedance mismatch and increase the ultrasound transmission. Since the minimum beam diameter of HIFU deices is about 1 mm [3,11,12], a spatial resolution of 1 mm or less is also required for temperature monitoring systems. For this purpose, polyimide-based thin-film RTD arrays [7,13] have been developed as thermal probes inside the phantom. The thin-film includes a total of 100 RTDs that are arrayed in a 10 × 10 layout to measure the temperature distribution inside the phantom caused by HIFU [13].

The use of HIFU leads to rapid changes in temperature inside the phantom, which necessitates using thermal probes with fast response times. While the thin-film RTD array sensors are expected to have a fast response time due to their small size, there is a lack of quantitative results on their effectiveness. This study aims to characterize and discuss the thermal response time of the thin-film RTD sensor.

2. Materials and Methods

2.1. Water Drop Method

To measure the response time of thermal probes, an experimental setup capable of creating a sudden change in temperature is necessary. One simple method involves quickly submerging the probes into a heated environment and analyzing the resulting temperature readings. However, this approach is only suitable for probes with slow response times (>10 s) due to mechanical limitations. Another method is the Loop Current Step Response (LCSR), which uses electrical power switching to generate heat on the probe, allowing for remote and repeated response time measurements [14]. However, it requires additional electrical circuitry and is not suitable for voltage-sensing probes such as thermocouples. A pulsed laser can also be used to generate heat on probes [15]. It is suitable for fast response time measurement but requires an expensive optical setup. It also cannot be applied for high-reflective probes. Heated liquid drops can also deliver heat to the thermal probes for response time measurement [16]. Although the size of the probe is limited due to the drop volume, this method is as simple as the plunge method and does not need any additional experimental setup. Figure 1a shows the schematic view of the liquid drop test. Both micropipettes and precise fluid injections, such as syringe pumps, can be used to drop liquid on the surface.

Although there is no limitation on the liquid in the droplet test, water has many advantages over other liquids. High specific heat capacity helps to maintain temperature as long as possible. A decrease in water temperature after dropping makes the response rate measurement inaccurate. In addition, since water has low vapor pressure and can stay on the sensor surface for a long time while minimizing evaporation, the response time of multiple sensors in an array can be measured at the same time.

2.2. Thin Fim RTD Array

The thin-film RTD array sensor on a thin polyimide (PI) substrate [13] was prepared for response time measurement. The sensor consists of 100 RTDs arranged in a 10 × 10 array with a 1 mm inter-RTD distance, with a serpentine design per unit of RTD. Platinum (Pt) was used as the material for both RTD electrodes and electrical lines. The thickness of the PI film was fabricated as thin as 2.5 μm.

The sensor was fabricated on a glass wafer using microelectromechanical system (MEMS) technology and then transferred onto a printed circuit board (PCB) using a previously reported process [13]. In brief, a polyimide liquid (VTEC PI-1388) diluted with N-Methyl-2-Pyrrolidone (NMP) was spin-coated and dried on a glass wafer. To create RTDs and electrode lines on the PI layer, a photoresist (PR) was spin-coated and patterned by a photolithography process using ultraviolet (UV) light and photomasks. Then, the Pt/Cr layer was deposited, and PR was removed to lift off Pt/Cr on PR. After attaching water-soluble tape to the wafer, the film was cut into individual devices and transferred from the wafer to the PCB. A square hole in the PCB corresponds to the sensing area (10 mm × 10 mm) of the thin-film RTD sensor. The electrode on the film was connected to the data acquisition (DAQ) system using anisotropic conductive film (ACF).

To perform the liquid drop test, an acrylic mold was assembled on the PCB, and polyurethane rubber (VytaFlex™40, Smooth-on, Inc., Macungie, PA, USA) was poured onto the top side of the sensor and cured. The speed of sound and attenuation coefficient of the rubber at the 1.1 MHz frequency were 1423 ± 4 m/s and 1.38 ± 0.07 dB/cm, respectively [13]. The PCB was then flipped so that the bottom PI substrate of the RTD array sensor was exposed to the air for the liquid drop method. The thin PI substrate acted as an insulating layer, preventing the RTD electrodes from contacting water. At the same time, the polyurethane rubber supported the thin-film sensors during the liquid drop test, preventing mechanical impact and deformation caused by the water drop.

2.3. Measurement Setup

For the resistance measurement of the RTD array, 50 µA of bias current was supplied to all RTDs using a source measurement unit (2450, Keithley manufacturer, Cleveland, OH, USA). Voltage drops at each RTD was measured using a 100-channel simultaneous data acquisition (DAQ) system (NI-cDAQ Chassis with 9202 Modules, National Instruments, Austin, TX, USA). The resistance at each channel was determined by dividing the voltage drop across the RTD by bias current. The sampling rate of the DAQ system was 10,000 Samples/s.

For response time measurement of the thin-film RTD array, 20 µL of deionized water was prepared, and colored ink was added to the water for visualization of the wet area. The water was preheated to several different temperatures ranging from 30 to 65 °C. The water volume was chosen not to cover the entire sensing area, allowing the wet and dry regions on the array to be distinguished using a micropipette. For response time analysis, the raw steamed data of a single RTD was used in the wetted area.

2.4. Response Time

The time constant, τ, is the time required for the sensor to undergo 63.2% of step change. The sensor will respond to the instantaneous temperature change from T₀ to T₁, and temperature as a function of time can be rewritten as follows:

$$T\left(t\right)=T_0+\left(T_1-T_0\right)[1-e^{-\frac{t}{\mathsf{\tau }}}].$$

(1)

3. Results

As shown in Figure 1b, colored water was dropped on the center of a thin-film RTD array. Figure 1c shows the temperature distribution right after the water drop. The RTDs inside the wet area responded instantly, while the other areas did not show any changes. No delays in signal change were observed depending on the RTD position. This indicated that the RTDs in the phantom layer were all independent, and only the areas under sudden temperature change due to water drop had a rise in resistance. After a few seconds, RTDs in the dry area also responded slowly due to the heat transfer through the film and the tissue-mimicking material.

Resistance before and after dropping is shown in Figure 2. For a water drop of 30 °C, the resistance changed by 0.7% from 1907.5 ± 1.0 Ω to 1920.0 ± 1.0 Ω. The maximum resistance increase was observed as 5.2% for 65 °C water. Since the temperature coefficient of resistance of thin-film platinum was 0.277%/°C [13], the resistance changes for a 65 °C water drop corresponded to about a 20 °C increase in temperature. Since the air continuously cools the water droplets pipetted from the hot water bath before making contact with the thin-film sensor, they were measured lower than the water bath temperature.

Regardless of water temperature, abrupt changes on the graph were observed upon the heated water droplet delivery. Once the heated water was dropped onto the thin-film surface, it took a certain amount of time for the resistance value to be stabilized and for the system to reach equilibrium. The temperature and humidity of the surrounding environment may also have an impact on the thermal stability of the RTD sensor. The time constant, τ, for the thermal response of the thin-film RTD array sensor was obtained as 1.0 ± 0.2 ms from five repeated measurements per each temperature condition. The measured response time was consistent regardless of drop conditions, such as the volume and height of the water drops.

4. Discussion

The advantage of the water drop method in this study is its simplicity. The response time could be characterized by just dropping liquid on the sensor, and no additional procedures or changes in the experimental setup are required. The temperature sensors within the wetted area could be measured simultaneously. However, the present methods had limitations in measuring the response time of conventional bulky thermal probes, such as a Pt-100 probe. A drop of water could not cover the entire probe surface simultaneously. Additionally, the high heat capacity of conventional probes also prevents water droplets from generating sufficient temperature changes. For these thermal probes, immersing them in a thermostat would be the best way to measure the response time.

The measured response would be affected by the thickness and thermal conductivity of the PI layer since the water was dropped to the PI layer for the electrical insulation of RTDs. A thick layer of low thermal conductivity material reduces the heat transfer rate, which results in a slow response of thermal probes. The difference in response time depending on the presence of the PI layer, however, would be negligible because the PI substrate was very thin (2.5 μm) compared to the thickness of the polyurethane rubber underneath it (10 mm). The response time of the thin-film RTD array sensor was sufficiently fast to evaluate HIFU devices that continuously apply thermal energy over a few seconds [6,17,18].

The most common method for measuring the thermal time constant is to apply adequate power to RTDs, keep it on until they reach thermal stability, and then turn off the power, called the “self-heating” method. In this case, the thermal time constant is the time it takes for the temperature of RTDs to decrease to the temperature representing a 63.2% difference. However, this method may not always be the best for all applications. For the applications we are targeting, such as using an RTD for the temperature measurement of a medium, it is best to measure the thermal time constant by using a step change of temperature, as we demonstrated in this work, rather than the self-heating method.

5. Conclusions

We measured the response time of thin-film RTD array sensors using a liquid drop method. A temperature sensor with a fast response is essential to evaluate the performance of HIFU devices that only heat up for a short period of time, i.e., within a few seconds. The response time measured in this study was confirmed to be sufficient for the application of HIFU validation.