Correlations between B-mode ultrasonic image texture features and tissue temperature in microwave ablation

Study on microwave ablation temperature prediction model based on grayscale ultrasound texture and machine learning

BACKGROUND

Temperature prediction is crucial in the clinical ablation treatment of liver cancer, as it can be used to estimate the coagulation zone of microwave ablation.

METHODS

Experiments were conducted on 83 fresh ex vivo porcine liver tissues at two ablation powers of 15 W and 20 W. Ultrasound grayscale images and temperature data from multiple sampling points were collected. The machine learning method of random forests was used to train the selected texture features, obtaining temperature prediction models for sampling points and the entire ultrasound imaging area. The accuracy of the algorithm was assessed by measuring the area of the hyperechoic area in the porcine liver tissue cross-section and ultrasound grayscale images.

RESULTS

The model exhibited a high degree of accuracy in temperature prediction and the identification of coagulation zone. Within the test sets for the 15 W and 20 W power groups, the average absolute error for temperature prediction was 1.14°C and 4.73°C, respectively. Notably, the model's accuracy in measuring the area of coagulation was higher than that of traditional ultrasonic grey-scale imaging, with error ratios of 0.402 and 0.182 for the respective power groups. Additionally, the model can filter out texture features with a high correlation to temperature, providing a certain degree of interpretability.

CONCLUSION

The temperature prediction model proposed in this study can be applied to temperature monitoring and coagulation zone range assessment in microwave ablation.

A Review of Quantitative Ultrasound-Based Approaches to Thermometry and Ablation Zone Identification Over the Past Decade

Percutaneous thermal therapy is an important clinical treatment method for some solid tumors. It is critical to use effective image visualization techniques to monitor the therapy process in real time because precise control of the therapeutic zone directly affects the prognosis of tumor treatment. Ultrasound is used in thermal therapy monitoring because of its real-time, non-invasive, non-ionizing radiation, and low-cost characteristics. This paper presents a review of nine quantitative ultrasound-based methods for thermal therapy monitoring and their advances over the last decade since 2011. These methods were analyzed and compared with respect to two applications: ultrasonic thermometry and ablation zone identification. The advantages and limitations of these methods were compared and discussed, and future developments were suggested.

Correlations between B-mode ultrasound image texture features and tissue temperatures in hyperthermia

PURPOSE

The noninvasive monitoring of mild hyperthermia or thermal ablation is important to guarantee therapeutic safety and efficacy. The potential of ultrasound B-mode image texture features in monitoring temperature or coagulation zones studied in this article.

MATERIALS AND METHODS

The experiments carried out on eighteen in vitro porcine liver samples heated from 20°C to 60°C in the water bath. The ultrasound radiofrequency signal at different temperature collected to reconstruct B-mode ultrasound images. The texture features based on gray level histogram (GLH), gray level co-occurrence matrix (GLCM), and gray level-gradient co-occurrence matrix (GGCM) extracted, respectively. Accordingly, we analyze the correlations between these texture features and temperature based on the experiment results.

RESULTS

The results showed that five texture feature parameters closely related to temperature, including mean gray scale of GLH, homogeneity of GLCM, hybrid entropy, inverse difference moment, and correlation of GGCM. Some of these feature parameters have correlation coefficients larger than 0.9 within the temperature range of 20°C to 60°C.

CONCLUSIONS

The above-mentioned five feature parameters expected to apply for noninvasive monitoring of MH or TA.

Identification of Denatured Biological Tissues Based on Compressed Sensing and Improved Multiscale Dispersion Entropy during HIFU Treatment

Identification of denatured biological tissue is crucial to high-intensity focused ultrasound (HIFU) treatment, which can monitor HIFU treatment and improve treatment efficiency. In this paper, a novel method based on compressed sensing (CS) and improved multiscale dispersion entropy (IMDE) is proposed to evaluate the complexity of ultrasonic scattered echo signals during HIFU treatment. In the analysis of CS, the method of orthogonal matching pursuit (OMP) is employed to reconstruct the denoised signal. CS-OMP can denoise the ultrasonic scattered echo signal effectively. Comparing with traditional multiscale dispersion entropy (MDE), IMDE improves the coarse-grained process in the multiscale analysis, which improves the stability of MDE. In the analysis of simulated signals, the entropy value of the IMDE method has less fluctuation compared with MDE, indicating that the IMDE method has better stability. In addition, MDE and IMDE are applied to the 300 cases of ultrasonic scattered echo signals after denoising (including 150 cases of normal tissues and 150 cases of denatured tissues). The experimental results show that the MDE and IMDE values of denatured tissues are higher than normal tissues. Both the MDE and IMDE method can be used to identify whether biological tissue is denatured. However, the multiscale entropy curve of IMDE is smoother and more stable than MDE. The interclass distance of IMDE is greater than MDE, and the intraclass distance of IMDE is less than MDE at different scale factors. This indicates that IMDE can better distinguish normal tissues and denatured tissues to obtain more accurate clinical diagnosis during HIFU treatment.

Identification of Denatured Biological Tissues Based on Time-Frequency Entropy and Refined Composite Multi-Scale Weighted Permutation Entropy during HIFU Treatment

Identification of denatured biological tissue is crucial to high intensity focused ultrasound (HIFU) treatment. It is not easy for intercepting ultrasonic scattered echo signals from HIFU treatment region. Therefore, this paper employed time-frequency entropy based on generalized S-transform (GST) to intercept ultrasonic echo signals. First, the time-frequency spectra of ultrasonic echo signal is obtained by GST, which is concentrated around the real instantaneous frequency of the signal. Then the time-frequency entropy is calculated based on time-frequency spectra. The experimental results indicate that the time-frequency entropy of ultrasonic echo signal will be abnormally high when ultrasonic signal travels across the boundary between normal region and treatment region in tissues. Ultrasonic scattered echo signals from treatment region can be intercepted by time-frequency entropy. In addition, the refined composite multi-scale weighted permutation entropy (RCMWPE) is proposed to evaluate the complexity of nonlinear time series. Comparing with multi-scale permutation entropy (MPE) and multi-scale weighted permutation entropy (MWPE), RCMWPE not only measures complexity of signal including amplitude information, but also improves the stability and reliability of multi-scale entropy. The RCMWPE and MPE are applied to 300 cases of actual ultrasonic scattered echo signals (including 150 cases in normal status and 150 cases in denatured status). It is found that the RCMWPE and MPE values of denatured tissues are higher than those of the normal tissues. Both RCMWPE and MPE can be used to distinguish normal tissues and denatured tissues. However, there are fewer feature points in the overlap region between RCMWPE of denatured tissues and normal tissues compared with MPE. The intra-class distance and the inter-class distance of RCMWPE are less and greater respectively than MPE. The difference between denatured tissues and normal tissues is more obvious when RCMWPE is used as the characteristic parameter. The results of this study will be helpful to guide doctors to obtain more accurate assessment of treatment effect during HIFU treatment.

Wavelet-based Computationally-Efficient Computer-Aided Characterization of Liver Steatosis using Conventional B-mode Ultrasound Images

Hepatic steatosis occurs when lipids accumulate in the liver leading to steatohepatitis, which can evolve into cirrhosis and consequently may end with hepatocellular carcinoma. Several automatic classification algorithms have been proposed to detect liver diseases. However, some algorithms are manufacturer-dependent, while others require extensive calculations and consequently prolonged computational time. This may limit the development of real-time and manufacturer-independent computer-aided detection of liver steatosis. This work demonstrates the feasibility of a computationally-efficient and manufacturer-independent wavelet-based computer-aided liver steatosis detection system using conventional B-mode ultrasound (US) imaging. Seven features were extracted from the approximation part of the second-level wavelet packet transform (WPT) of US images. The proposed technique was tested on two datasets of ex-vivo mice livers with and without gelatin embedding, in addition to a third dataset of in-vivo human livers acquired using two different US machines. Using the gelatin-embedded mice liver dataset, the technique exhibited 98.8% accuracy, 97.8% sensitivity, and 100% specificity, and the frame classification time was reduced from 0.4814 s using original US images to 0.1444 s after WPT preprocessing. When the other mice liver dataset was used, the technique showed 85.74% accuracy, 84.4% sensitivity, and 88.5% specificity, and the frame classification time was reduced from 0.5612s to 0.2903 s. Using human liver image data, the best classifier exhibited 92.5% accuracy, 93.0% sensitivity, 91.0% specificity, and the classification time was reduced from 0.660 s to 0.146 s. This technique can be useful for developing computationally-efficient and manufacturer-independent noninvasive CAD systems for fatty liver detection.

Feasibility of Using Ultrasonic Nakagami Imaging for Monitoring Microwave-Induced Thermal Lesion in Ex Vivo Porcine Liver

The feasibility of using ultrasonic Nakagami imaging to evaluate thermal lesions induced by microwave ablation (MWA) in ex vivo porcine liver was explored. Dynamic changes in echo amplitudes and Nakagami parameters in the region of the MWA-induced thermal lesion, as well as the contrast-to-noise ratio (CNR) between the MWA-induced thermal lesion and the surrounding normal tissue, were calculated simultaneously during the MWA procedure. After MWA exposure, a bright hyper-echoic region appeared in ultrasonic B-mode and Nakagami parameter images as an indicator of the thermal lesion. Mean values of the Nakagami parameter in the thermal lesion region increased to 0.58, 0.71 and 0.91 after 1, 3 and 5 min of MVA. There were no significant differences in envelope amplitudes in the thermal lesion region among ultrasonic B-mode images obtained after different durations of MWA. Unlike ultrasonic B-mode images, Nakagami images were less affected by the shadow effect in monitoring of MWA exposure, and a fairly complete hyper-echoic region was observed in the Nakagami image. The mean value of the Nakagami parameter increased from approximately 0.47 to 0.82 during MWA exposure. At the end of the postablation stage, the mean value of the Nakagami parameter decreased to 0.55 and was higher than that before MWA exposure. CNR values calculated for Nakagami parameter images increased from 0.13 to approximately 0.61 during MWA and then decreased to 0.26 at the end of the post-ablation stage. The corresponding CNR values calculated for ultrasonic B-mode images were 0.24, 0.42 and 0.17. This preliminary study on ex vivo porcine liver suggested that Nakagami imaging have potential use in evaluating the formation of MWA-induced thermal lesions. Further in vivo studies are needed to evaluate the potential application.

Texture analysis in gel electrophoresis images using an integrative kernel-based approach

Texture information could be used in proteomics to improve the quality of the image analysis of proteins separated on a gel. In order to evaluate the best technique to identify relevant textures, we use several different kernel-based machine learning techniques to classify proteins in 2-DE images into spot and noise. We evaluate the classification accuracy of each of these techniques with proteins extracted from ten 2-DE images of different types of tissues and different experimental conditions. We found that the best classification model was FSMKL, a data integration method using multiple kernel learning, which achieved AUROC values above 95% while using a reduced number of features. This technique allows us to increment the interpretability of the complex combinations of textures and to weight the importance of each particular feature in the final model. In particular the Inverse Difference Moment exhibited the highest discriminating power. A higher value can be associated with an homogeneous structure as this feature describes the homogeneity; the larger the value, the more symmetric. The final model is performed by the combination of different groups of textural features. Here we demonstrated the feasibility of combining different groups of textures in 2-DE image analysis for spot detection.

A survey of ultrasound elastography approaches to percutaneous ablation monitoring

Percutaneous thermal ablation has been widely used as a minimally invasive treatment for tumors. Treatment monitoring is essential for preventing complications while ensuring treatment efficacy. Mechanical testing measurements on tissue reveal that tissue stiffness increases with temperature and ablation duration. Different types of imaging methods can be used to monitor ablation procedures, including temperature or thermal strain imaging, strain imaging, modulus imaging, and shear modulus imaging. Ultrasound elastography demonstrates the potential to become the primary imaging modality for monitoring percutaneous ablation. This review briefly presented the state-of-the-art ultrasound elastography approaches for monitoring radiofrequency ablation and microwave ablation. These techniques were divided into four groups: quasi-static elastography, acoustic radiation force elastography, sonoelastography, and applicator motion elastography. Their advantages and limitations were compared and discussed. Future developments were proposed with respect to heat-induced bubbles, tissue inhomogeneities, respiratory motion, three-dimensional monitoring, multi-parametric monitoring, real-time monitoring, experimental data center for percutaneous ablation, and microwave ablation monitoring.

Feasibility of non-invasive temperature estimation by the assessment of the average gray-level content of B-mode images

This paper assesses the potential of the average gray-level (AVGL) from ultrasonographic (B-mode) images to estimate temperature changes in time and space in a non-invasive way. Experiments were conducted involving a homogeneous bovine muscle sample, and temperature variations were induced by an automatic temperature regulated water bath, and by therapeutic ultrasound. B-mode images and temperatures were recorded simultaneously. After data collection, regions of interest (ROIs) were defined, and the average gray-level variation computed. For the selected ROIs, the AVGL-Temperature relation were determined and studied. Based on uniformly distributed image partitions, two-dimensional temperature maps were developed for homogeneous regions. The color-coded temperature estimates were first obtained from an AVGL-Temperature relation extracted from a specific partition (where temperature was independently measured by a thermocouple), and then extended to the other partitions. This procedure aimed to analyze the AVGL sensitivity to changes not only in time but also in space. Linear and quadratic relations were obtained depending on the heating modality. We found that the AVGL-Temperature relation is reproducible over successive heating and cooling cycles. One important result was that the AVGL-Temperature relations extracted from one region might be used to estimate temperature in other regions (errors inferior to 0.5 °C) when therapeutic ultrasound was applied as a heating source. Based on this result, two-dimensional temperature maps were developed when the samples were heated in the water bath and also by therapeutic ultrasound. The maps were obtained based on a linear relation for the water bath heating, and based on a quadratic model for the therapeutic ultrasound heating. The maps for the water bath experiment reproduce an acceptable heating/cooling pattern, and for the therapeutic ultrasound heating experiment, the maps seem to reproduce temperature profiles consistent with the pressure field of the transducer, and in agreement with temperature maps developed by COMSOL®MultiPhysics simulations.

Ultrasound thermal mapping based on a hybrid method combining physical and statistical models

Non-invasive temperature measurement of tissues deep inside the body has great potential for clinical applications, such as temperature monitoring during thermal therapy and early diagnosis of diseases. We developed a novel method for both temperature estimation and thermal mapping that uses ultrasound B-mode radiofrequency data. The proposed method is a hybrid that combines elements of physical and statistical models to achieve higher precision and resolution of temperature variations and distribution. We propose a dimensionless combined index (CI) that combines the echo shift differential and signal intensity difference with a weighting factor relative to the distance from the heat source. In vitro experiments verified that the combined index has a strong linear relationship with temperature variation and can be used to effectively estimate temperature with an average relative error <5%. This algorithm provides an alternative for imaging guidance-based techniques during thermal therapy and could easily be integrated into existing ultrasound systems.

Ultrasonic evaluation of microwave-induced thermal lesions based on wavelet analysis of mean scatterer spacing

The microwave ablation has become an important manner for tumor treatment. In this paper, we proposed a new method for evaluation of microwave-induced thermal lesions using the wavelet analysis of the mean scatterer spacing (MSS). First, the ultrasonic radiofrequency (RF) data of normal and coagulated porcine liver tissues was collected through the temperature-controlled water bath heating experiments. The convex array ultrasound probe with a center frequency of 3.5 MHz was used. Second, the wavelet analysis was used to compute the MSS of normal and coagulated porcine liver tissues, respectively. Finally, the microwave-induced thermal lesions were detected based on the differences in the MSS between normal tissues and thermal lesions. Eighteen cases of microwave ablation experiments and 20 cases of water bath heating experiments were conducted on fresh porcine liver samples. The MSS of normal porcine liver tissues was 1.15±0.12 mm, and the MSS of coagulated porcine liver tissues was 0.93±0.07 mm. Six cases of thermal lesions were compared between the MSS-detected area and the caliper-measured area, and the MSS-detected area had an error of (13.55±5.29) %. The experimental results indicated that the proposed method could be used in preliminary detection and evaluation of microwave-induced thermal lesions.

Ultrasound thermal mapping based on a hybrid method combining cross-correlation and zero-crossing tracking

A hybrid method for estimating temperature with spatial mapping using diagnostic ultrasound, based on detection of echo shifts from tissue undergoing thermal treatment, is proposed. Cross-correlation and zero-crossing tracking are two conventional algorithms used for detecting echo shifts, but their practical applications are limited. The proposed hybrid method combines the advantages of both algorithms with improved accuracy in temperature estimation. In vitro experiments were performed on porcine muscle for preliminary validation and temperature calibration. In addition, thermal mapping of rabbit thigh muscle in vivo during high-intensity focused ultrasound heating was conducted. Results from the in vitro experiments indicated that the difference between the estimated temperature change by the proposed hybrid method and the actual temperature change measured by the thermocouple was generally less than 1 °C when the increase in temperature due to heating was less than 10 °C. For the in vivo study, the area predicted to experience the highest temperature coincided well with the focal point of the high-intensity focused ultrasound transducer. The computational efficiency of the hybrid algorithm was similar to that of the fast cross-correlation algorithm, but with an improved accuracy. The proposed hybrid method could provide an alternative means for non-invasive monitoring of limited temperature changes during hyperthermia therapy.