Metal reflector-enhanced thermoacoustic imaging as a guidance for puncture biopsy

1. Introduction

Puncture biopsy is a medical examination method that uses a metal puncture needle to obtain tissue or cell samples at the region of the lesion for pathological analysis¹ and can be used in a variety of organs, including joints,^2,3 liver,^4,5 thyroid,⁶ breast,⁷ etc. The pathological information provided by puncture biopsy is often the gold standard for cancer diagnosis,⁸ which helps doctors make an accurate assessment to formulate an appropriate treatment plan, and plays an important role in the evaluation of efficacy.⁹ Puncture biopsy is usually guided by imaging technology that can help doctors locate the lesion or target tissue and ensure the accuracy and safety of the puncture.¹⁰

Among the commonly used imaging techniques, computed tomography (CT) has a high image resolution and can provide accurate three-dimensional positioning to assist in the completion of puncture biopsy.¹¹ However, CT cannot provide real-time imaging of intraoperative puncture guidance and poses a risk of ionizing radiation to the people involved.¹² Magnetic resonance imaging (MRI) has a high resolution of soft tissue, multi-plane imaging capability, and no ionizing radiation, which can clearly display lesions and surrounding tissues¹³ and can locate small breast lesions more accurately than other imaging techniques.¹⁴ However, MRI has metal contraindications, and the operation cost of the equipment is high.¹⁵ Ultrasound imaging (US) is the most commonly used puncture guidance imaging technology.¹⁶ It has the advantages of real-time imaging, no radiation and low cost. However, the US has limitations in revealing deep tissue structure.^17,18 Traditional imaging techniques have played a great role in clinical puncture biopsy guidance, but there are still some usage restrictions.

Photoacoustic Imaging (PAI) is a new imaging method that combines high-contrast of optical imaging and high-resolution of ultrasonic imaging.^19,20 There have been studies using PAI to guide lymph node puncture and prostate cancer puncture.^21,22 But due to the limitation of light penetration depth, PAI is difficult to guide puncture biopsy of deep lesions. Microwave-induced thermoacoustic imaging (TAI) is a novel imaging method that combines microwave imaging and US imaging.²³ Microwaves have the ability to penetrate biological tissue much deeper than light, making it possible for the TAI technique to perform high-contrast and high-resolution imaging at depths that are difficult to achieve with conventional optical or PAI.²⁴ Up to now, TAI has been studied in breast cancer,^25,26 prostate,²⁷ joint,^28,29 thyroid³⁰ and other applications.^31,32 Huang et al. developed a real-time TAI system consisting of a 3 GHz microwave source and a 128-channel semi-ring ultrasound transducer to image the boundary between pork fat and background, breast tumor and puncture needle at a subcutaneous depth of 2 cm at the same time, achieving a preliminary assessment of the feasibility of TAI for breast cancer puncture guidance.³³

In this paper, we report a study on the application of a fast TAI system to puncture guidance imaging at different depths and propose a method to enhance the TAI signal intensity and imaging depth by using a metal reflector based on the reflection characteristics of electromagnetic wave of a metal plate, realizing the puncture needle imaging at a depth of more than 12 cm. The paper is arranged as follows. In Sec. 2, the basis of metal reflector-enhanced TAI is explained based on the imaging principle, and a fast TAI system for ex-vivo experiments is introduced. In Sec.3, a breast cancer model and a knee gout model are constructed based on CST simulation software, and puncture guidance simulation experiments were performed with or without reflectors at different depths. In addition, the ex-vivo experiments of tumor phantom puncture guidance with or without reflectors at different depths are presented. In Sec. 4, the principle of metal reflector enhancement and experimental results are discussed. In Sec. 5, the study is summarized.

2. Materials and Methods

2.1. Theoretical basis

The time domain features of thermoacoustic signals can be characterized by the acoustic pressure $p(r,t)$ at position r at time t. After the initial sound pressure, $p_0$ is generated at position r, as time t passes, the thermoacoustic wave $p(r,t)$ propagates in all directions in the form of a spherical wave. The thermoacoustic wave equation describes the relationship between thermoacoustic wave $p(r,t)$ and temperature $T(r,t)$, it can be expressed as the following equation³⁴ :

where $v_s$ is the speed of sound in tissue, ${\beta }_e$ is the volumetric expansion coefficient, k is isothermal compression coefficient and its equation as follows : where $C_p$ is the specific heat at constant pressure, $C_v$ is specific heat at constant volume. Since the microwave pulse width $\tau$ is very narrow, the temperature $T(r,t)$ changes linearly in time $\tau$, the relationship can be expressed as follows : The temperature rises $\mathrm{\Delta }T$ in Eq. (3) can be expressed as follows : The electromagnetic loss $P_{\mathrm{\text{loss}}}$ in Eq. (4) can be expressed as follows : By substituting Eqs. (2)–(5) into Eq. (1), we can get the following expression : where $I(t)$ is the time-domain waveform of the microwave pulse envelope, and $\sigma (r)$ is the tissue conductivity. In general, ${\beta }_e$ and $C_p$ can be treated as constant values.^35,36,37 Therefore, the thermoacoustic pressure difference is mainly determined by the electric field $\stackrel{⟶}{|E(r)|}$ and the tissue conductivity $\sigma (r)$. Due to the reflection feature of metal on microwave energy,³⁸ we propose to use a metal plate to reflect the microwave energy penetrating tissue back into the imaging region to improve the electric field $\stackrel{⟶}{|E(r)|}$ in the imaging region, thus enhancing the thermoacoustic signal intensity and improving the imaging depth of TAI.

2.2. Experimental system

A schematic diagram of our fast TAI system is shown in Fig. 1. The homemade microwave generator (frequency: 3GHz, repetition frequency: 10–200Hz, peak power: 65kW, pulse width: 70–500ns) was used as the excitation source. The pulse width and the repetition frequency selected in the ex-vivo experiment were 550ns and 50Hz, respectively. The pulsed microwave was radiated to the tissue via the coaxial wire (length: 1.5 m, insertion loss: 1.2dB) and the pyramidal horn antenna (aperture: $14\mathrm{\text{cm}}\times 11\mathrm{\text{cm}}$). The average power density at the antenna aperture was 8mW/cm². Considering the multiple reflections of the microwave in air, mineral oil and tissue, the actual average power density at the tissue is below the IEEE safety threshold (20mW/cm² at 3GHz).³⁹

With mineral oil as the coupling medium, the tissue absorbed microwave energy to produce thermoacoustic signals, which were detected by a customized 128-channel semi-ring ultrasonic transducer (center frequency: 3MHz, −6dB bandwidth ≥60%, curvature radius: 65mm). The detected signals were amplified by a home-made 128-channel array amplifier (voltage gain: 56dB, −3 dB bandwidth: 0.26–2.19MHz) and finally collected by two 64-channel data acquisition cards (sampling rate: 50MHz, resolution: 12 bits). The collected data were reconstructed by delay and sum algorithm to map the distribution of microwave energy absorption of tissue. It should be noted that in the ex-vivo experiment, due to one of two data acquisition cards burned out, only the signals detected by the 64 channels detection elements distributed at 180^∘ were actually collected for image reconstruction.

3. Results

3.1. Simulation of breast cancer model

Comprehensively referring to previous breast cancer simulation studies,^40,41,42 the setup of the breast cancer puncture model is shown in Figs. 2(a)–2(c). The shape of the breast model was semi-ellipsoid, with a height of 10cm and a diameter of 14cm at the widest point. The main component of the breast model was set as fat, a 1cm muscle layer was added at the bottom, and the internal tumor tissue was set as a uniform sphere with a radius of 5mm. The puncture needle was set with reference to the 14G needle (outer diameter: 2.1mm) commonly used in breast cancer puncture biopsy. Depending on the manufacturer and usage scenario, the inner diameter of the 14G puncture needle ranges from 1.5mm to 1.9mm, and the length ranges from 6cm to 20cm. A puncture needle with inner diameter of 1.7mm and length of 14cm was used in simulation of breast cancer model. The antenna was a pyramidal horn antenna with an aperture of $14\mathrm{\text{cm}}\times 11\mathrm{\text{cm}}$. The conductivity and relative permittivity of each tissue in the breast cancer model at 3GHz frequency are shown in Table 1.⁴²

According to the TAI process, the power loss density is directly related to the tissue microwave energy absorption.⁴³ The higher the dielectric properties and the stronger the electric field strength, the greater the corresponding power loss density and the stronger the thermoacoustic signal intensity.

In order to explore the effect of electric field polarization direction on the imaging of the puncture needle, simulation experiments were conducted at a depth of 2cm, where the angle between the puncture needle and the electric field polarization direction was 90^∘, 45^∘ and 0^∘, respectively. The simulation results are shown in Figs. 2(d1)–2(d3). As shown in Fig. 2(d3), when the angle between the puncture needle and the electric field polarization direction was 0^∘, the power loss density of the tissue around the puncture needle was the highest. As shown in Fig. 2(e), the power loss density of the tissue around the puncture needle and the tumor became stronger after using the reflector.

In order to explore the effect of depth on TAI-guided puncture needle imaging, the simulation experiments of breast cancer with or without reflectors at different depths were carried out. The power loss density distributions of the normal breast model at different depths are shown in Figs. 3(a1)–3(a4). The power loss density distributions of the breast cancer model at different depths are shown in Figs. 3(b1)–3(b4). The results showed that there was no abnormal TAI signal area in the normal breast model at different depths. At the depth of 4cm, the power loss density of the tumor in the breast cancer model was high, and the tumor outline was clearly visible. When the tumor depth reached 6cm, the power loss density of the tumor was weakened compared with that of the tumor at 4cm, but the outline of the tumor was still clear. When the depth reached 8cm, the power loss density of the tumor deteriorated, and the shape and outline became blurred, but the location of the tumor could still be recognized. However, when the tumor was at a depth of 10cm, its power loss density was insufficient and it was almost invisible in the image.

In Figs. 3(c1)–3(c4), the power loss density distribution was presented when a metal plate reflector with an angle of about 45^∘ from the horizontal plane was placed on the right side of the breast cancer model. The results showed that compared with the breast cancer model without reflective surface, the power loss density of the tumor with reflective surface at the corresponding depth was improved, the contrast between the tumor and the fat was higher, and the tumor outline was clearer. It was worth noting that when the reflector was not used, the tumor was almost invisible in the breast cancer model at a depth of 10 cm. When the reflector was used to reflect microwave energy to the tumor area, the tumor in the breast cancer model at the depth of 10 cm could be recognized. As shown in Figs. 3(d1)–3(d4). the power loss density distributions of breast cancer puncture models at different depths without reflectors were displayed. The results showed that at different depths, there were strong power loss densities in the local region where the puncture needle contacts the breast model and the tip of the needle. The distribution of power loss density around the puncture needle was roughly consistent with the shape of the puncture needle. With the increase of depth, despite the power loss density around the puncture needle in the breast cancer puncture model gradually decreased, the position and shape of the puncture needle could still be clearly identified. It should be noted that the tumor in the breast cancer model at the depth of 10cm was almost invisible. Therefore, in Fig. 3(d4), the breast cancer puncture model at the corresponding depth was not present.

As shown in Figs. 3(e1)–3(e4), power loss density distributions of breast cancer puncture models at different depths after using metal reflectors were presented. The results showed that compared with the breast cancer puncture model without a reflector, the power loss density of the tissue around the puncture needle at the corresponding depth was improved, and the contour of the puncture needle was more uniform. As shown in Fig. 3(e4), although the uniformity of the power loss density distribution around the puncture needle was reduced, the contour of the puncture needle could still be identified.

3.2. Simulations of knee gout model

In addition to the breast cancer model, a complex knee gout model was developed for puncture biopsy guidance studies. As shown in Figs. 4(a)–4(d), the knee joint model contained skin, fat, tendon, ligament, bone and cartilage. The dielectric properties of the above tissues at 3GHz frequency are shown in Table 2. In the knee gout model, the tophi was set as a sphere with a radius of 2.5mm, and its conductivity and relative permittivity were set to be the same as that of 3% saline.^44,45 The puncture needle was set with reference to the 18G needle (outer diameter: 1.1mm) commonly used in knee joint puncture biopsy. Depending on the manufacturer and usage scenario, the inner diameter of the 18G puncture needle ranges from 0.79mm to 1.04mm, and the length ranges from 6cm to 20cm. A puncture needle with inner diameter of 0.9mm and length of 7.5cm was used in simulation of the knee gout model. As shown in Figs. 4(c) and 4(d), the puncture angles of the shallow knee gout model and the deep knee gout model were shown, respectively. Among them, the tophi in the shallow knee gout model was 3cm below the skin, and the tophi in the deep knee gout model was 6cm below the skin.

The power loss intensity distribution of the normal knee joint model and the shallow knee gout model were, respectively, as shown in Figs. 4(e1) and 4(e2). The results showed that the power loss density of the tophi in the shallow knee gout model was high, and the position and outline of the tophi could be clearly distinguished. Figure 4(e3) shows the power loss density distribution after the reflector reflected microwave energy to the tophi region. Compared with the shallow knee gout model without the reflector, the power loss density of the tophi increased after the reflector was used. In Figs. 4(e4) and 4(e5), the power loss density distributions of the shallow knee gout puncture model without a reflector and the shallow knee gout puncture model with a reflector were displayed, respectively. The results showed that the puncture needle could be clearly observed. In contrast, after the reflector was used, the power loss density around the puncture needle was higher, and the distribution of power loss density was more uniform.

As shown in Fig. 4(f2), the tophi in the deep knee gout model was 6cm below the skin, and the simulation results showed that the power loss density of the tophi was lower than that in the shallow knee gout model in Fig. 4(e2). As shown in Fig. 4(f3), after enhancing the deep tophi area with a reflector, the power loss density of the tophi was improved, and the outline of the tophi was more obvious. The simulation results of the deep knee gout puncture model are shown in Fig. 4(f4). The needle is inserted obliquely from the patella into the knee joint, and the power loss density distribution around the puncture needle could show the location and shape of the puncture needle. Figure 4(f5) shows the power loss density distribution of the deep knee gout puncture model after using the reflector. The results showed that the power loss density around the puncture needle increased, which was more beneficial to the puncture guidance process.

3.3. Ex-vivo experiments

To further validate the impact of the electric field polarization direction on the imaging of the puncture needle, ex-vivo experiments were conducted. Referring to existing research,^49,50 the experimental samples included pork fat with a height of 4cm and a diameter of 8cm to mimic the breast, and a muscle with a diameter of about 1cm embedded in the sample to mimic a tumor. A 6cm 18G puncture needle was used in the following ex-vivo experiments. As shown in Figs. 5(a1)–5(a3), these were the thermoacoustic reconstruction images of tumor phantom puncture without reflector enhancement when the angles between the puncture needle and the electric field polarization direction are different at a depth of 2cm. The results showed that when the puncture needle was at an angle of 0^∘ to the polarization direction of the electric field, the thermoacoustic signal intensity around the puncture needle was the highest, providing the best display effect. As the angle between the puncture needle and the electric field polarization direction increased, the thermoacoustic signal intensity around the puncture needle gradually weakened, until it was barely visible at an angle of 90^∘, resulting in the poorest visualization effect. Therefore, the subsequent penetration depth experiments were based on the puncture needle parallel to the direction of the electric field polarization. Figures 5(b1)–5(b3) showed the thermoacoustic reconstruction images of tumor phantom puncture with the use of a metal plate to reflect and recycle microwave energy at a depth of 2cm. The results indicate that compared to Figs. 5(a1)–5(a3), the signal intensity of the tumor phantom and puncture needle in the three different angle groups with reflectors was higher than that of the corresponding groups without reflectors. Moreover, the puncture needle shown in Fig. 5(a3) was barely visible, Fig. 5(b3) showed a clear outline after the enhancement of the reflector, proving the beneficial effect of the metal reflector in enhancing microwave-induced TAI. As shown in Fig. 5(c), the normalized TAI signal intensities of the puncture needle with or without reflector enhancement were quantified. The result showed that at the angles of 0^∘, 45^∘ and 90^∘ angles, the TAI signal intensity of the puncture needle with reflector enhancement was 33.4%, 20.7%, and 7.4% higher than that of the corresponding puncture needle without reflector enhancement, respectively.

Puncture guidance ex-vivo experiments under different depths with or without reflector enhancement were carried out on the basis of the angle between electric field polarization direction and puncture needle was 0^∘. A layer of pork fat with a thickness of 2cm was added below the tested fat block embedded muscle, and a 2cm thick acrylic plate was placed under the probe for the TAI-guided puncture experiment at a depth of 4cm. Deeper puncture guidance scenes could be simulated by layer-by-layer superposition. Figures 6(a1)–6(a3) showed the TAI reconstruction results for the tumor phantom group at depths of 4, 8 and 10cm without a reflector. The results showed that at a depth of 4cm, the tumor phantom was clearly visible with complete boundaries between fat and mineral oil. As the depth increases to 10cm, the signal intensity of the tumor phantom decreases but remains visible. The boundary of the fat became blurred with only partial boundaries visible, indicating the weakening effect of increasing depth on image quality. The corresponding image results with a reflector are shown in Figs. 6(b1)–6(b3). After adding a reflector, the tumor maintained a higher thermoacoustic signal intensity even at deeper depths of 8cm and 10cm. The boundaries between fat and mineral oil were more complete at depths of 4cm and 10cm. At a depth of 8cm, Fig. 6(b2) shows more boundary details than the corresponding results without a reflector shown in Fig. 6(a2). Figure 6(c) shows the statistical results of the normalized thermoacoustic signals of the simulated tumor phantom and fat. The results indicated a downward trend in the amplitude of the tumor’s thermoacoustic signal with increasing depth, while the changes in fat are not significant. At depths of 4, 8 and 10cm, due to the enhancement of the reflector, the contrast ratio between the tumor phantom and fat increased from 1.44, 1.35 and 1.31 to 1.74, 1.57 and 1.50, respectively. As shown in Figs. 6(d1)–6(d3), the normalized TAI reconstruction results of the tumor phantom puncture group at a depth of 4, 8 and 10cm without reflector were obtained. The results showed that the thermoacoustic signal intensity of the puncture needle was strong at both ends and weak in the middle, and gradually decreased with the increase of depth. The boundary between fat and mineral oil was only complete at a depth of 4cm and incomplete at 8cm and 10cm. Figures 6(e1)–6(e3) show the reconstruction result after using a reflector. Compared with the group without a reflector, the puncture needle signal of the group with a reflector was stronger and the boundary between fat and mineral oil was clearer at a similar depth. The statistical results in Fig. 6(f) showed that at the depth of 4, 8 and 10cm, the reflector improved the contrast between the puncture needle and fat, from 1.53, 1.42 and 1.33 in the group without a reflector to 1.90, 1.72 and 1.53 in the group with reflector, respectively, proving the ability of the reflector to improve the visualization of the puncture needle.

Figures 7(a1)–7(a3) show the normalized TAI reconstruction results for the tumor phantom puncture group at depths of 12, 14 and 16cm without a reflector. From the result, the tumor phantoms at 12cm and 14cm were visible, while the tumor phantom at the depth of 16cm was barely visible and appeared smaller. The boundary between mineral oil and fat was partially visible at the depth of 12cm, but not visible at 14cm and 16cm. The corresponding experimental results with the reflector are shown in Figs. 7(b1)–7(b3). The results showed that the visualization of the tumor phantom was improved. Especially at the depth of 16cm, the size of the tumor phantom shown in Fig. 7(b3) remained normal despite the weak imaging effect. Figure 7(c) shows the statistical results, indicating that the TAI signal of the tumor phantom decreased with increasing depth. The contrast ratios between the tumor phantom and fat at the depths of 12, 14 and 16cm were 1.25, 1.16 and 1.05 in the group without a reflector, and increased to 1.46, 1.37 and 1.11 after the enhancement of the reflector, respectively. Figures 7(d1)–7(d3) show the normalized TAI reconstruction results for the tumor phantom puncture group at the depths of 12, 14 and 16cm without a reflector. The results showed the good visualization effect of the puncture needle at the depth of 12cm. At 14cm, the visualization effect of the puncture needle decreased. When the depth reaches 16cm, the visualization effect of the puncture needle further decreases. The corresponding experimental results with the reflector are shown in Figs. 7(e1)–7(e3), where the signal of the puncture needle became stronger, but the boundaries between fat and mineral oil still could not be fully recognized. Figure 7(f) showed the statistical results, indicating a downward trend in the TAI signal of the tumor phantom with increasing depth. The contrast ratios between the puncture needle and fat at the depths of 12, 14 and 16cm without a reflector were 1.27, 1.23 and 1.10, respectively, and increased to 1.46, 1.35 and 1.20 with the enhancement of the reflector, respectively.

In summary, the reflector can enhance the TAI signal intensity of the tumor phantom and improve the clarity of the boundary between fat and mineral oil, which is crucial for precise localization. By using a reflector, the image of the puncture needle is clearer, helping to improve the accuracy and safety of the puncture.

To test the imaging effect of the puncture needle during deep puncture, an 11cm 18G model puncture needle was inserted into a rectangular pork fat sample (size: $10\mathrm{\text{cm}}\times 6\mathrm{\text{cm}}\times 4\mathrm{\text{cm}}$). The TAI reconstruction result of the puncture needle at 10cm deep is shown in Fig. 8(a). The results showed that the tissue around the puncture needle had a high thermoacoustic signal intensity with a uniform distribution. The TAI reconstruction result after the addition of a reflector is shown in Fig. 8(b). Compared to Fig. 8(a), the contrast between the puncture needle and pork fat was enhanced after using the reflector, and the outline of the puncture needle became clearer. In Fig. 8(e), the TAI signal intensity of the puncture needle with reflector enhancement was 43.3% higher than that of the puncture needle without reflector enhancement. The quantitative result further validated the findings obtained from the visual observation.

4. Discussion

The metal reflector enhances the electric field intensity in the tested sample by superimposing the incident microwave energy and the microwave energy passing through or diffracted by the sample, thereby enhancing the TAI signal intensity. Therefore, the metal reflector should be placed in the microwave propagation path and should face the antenna and sample. And the reflector needs to avoid being too far away from the sample to avoid the loss of microwave energy.

In theory, when the size of the metal reflector is much larger than the wavelength, it can be considered as an ideal emitting mirror.⁵¹ When it is smaller than the wavelength, the scattering phenomenon increases and reflection efficiency decreases.⁵² The large metal reflector can achieve electric field enhancement in a large area, which is conducive to the large-scale screening required for puncture biopsy guidance.

The smoother the metal reflector and the higher the conductivity, the better the reflection effect. As a common light metal, aluminum (At 3GHz, conductivity is $3.77\times 10^7$ S/m, wave impedance is $0.017+0.017$i$\mathrm{\Omega }$, and reflection coefficient is 0.9999$\angle 179.9^{\circ }$) has low cost and good corrosion resistance, making it the preferred material for a reflector.

Due to the propagation of TAI signals in the form of spherical waves,⁵³ semi-ring ultrasonic transducers are easier to achieve all-round signal detection than linear ultrasonic transducers and have a larger sensitivity range.^54,55 Compared with semi-ring ultrasound transducers, linear ultrasound transducers are flexible in use and more suitable for clinical applications due to their small size and lightweight. The method of enhancing the electric field strength and thus increasing the signal intensity by the metal reflector is not limited to the type of transducer. In the future, the linear ultrasonic transducers combined with the reflector enhancement method can be used for clinical pilot studies.

Although one of our two data acquisition cards burned out during the ex-vivo experiments, the reconstructed images could show the puncture needle and the tested sample completely. If more channels are used to acquire data, the reconstructed images will be clearer and more conducive to the puncture guidance.

The statistical charts in all ex-vivo experiments displayed the quantitative analysis results of three repeated experiments, demonstrating the stability of the metal reflector method in improving the image quality of TAI-guided puncture. However, it should be noted that due to the long interval between the first experiment and the repeated experiments, the energy difference between the original aging microwave source and the current calibrated microwave source led to the differences in the TAI signal intensity of fat, as shown by the large error bars for fat in Figs. 6(c) and 6(c). However, as a homogeneous background, the fat signal remained minimal and did not affect the identification of tumors.

5. Conclusions

In summary, we have presented a reflector-enhanced TAI approach and validated it as a puncture guidance tool using simulated and ex-vivo experiments. We found that the imaging effect of the puncture needle is best when the angle between the puncture needle and the electric field polarization direction is 0^∘. The TAI signal intensities of both the tumor phantom and puncture needle at different depths are improved after the metal reflector is used. With the reflector, the breast tumor and fat and puncture needle at depths of more than 12cm are still clearly visible, which proves the ability of the reflector to enhance imaging depth. This suggests the potential of TAI enhanced by reflectors for deep tissue puncture guidance.

While the results overall are positive, there are still some shortcomings. First, the arrangement of the metal reflector may affect its effect on the enhancement of the TAI signal intensity. Thus, the angles and positions of reflectors can be adjusted in a small range to achieve the best imaging quality. Second, using shorter microwave pulses or ultrasonic transducers with higher center frequency can improve image resolution and the accuracy of puncture guidance. Finally, only simulated and ex-vivo experiments were performed at present. In the future, in-vivo experiments can be carried out to further confirm the application of TAI as puncture guidance.

Acknowledgments

This work was supported in part by the Chinese Postdoctoral Science Foundation (2022MD723722), in part by the National Natural Science Foundation of China (62001075), and in part by the Chongqing postdoctoral research project (special funding project 2021XM2026). Shuang Du and Tao Qiang contributed equally to this work.

Conflicts of Interest

The authors have no conflicts of interest to disclose.

ORCID

Shuang Du  https://orcid.org/0000-0001-8778-735X

Tao Qiang  https://orcid.org/0009-0001-5858-7823

Zihui Chi  https://orcid.org/0000-0003-1278-4706

Huabei Jiang  https://orcid.org/0000-0002-9665-4919