Comparison between A-mode and B-mode ultrasound in local hyperthermia monitoring

Abstract

Hyperthermia therapy is one of the therapy methods used for cancer treatment. It has shown to be an effective way of treating the cancerous tissue when compared to surgery, chemotherapy and radiation. However, real time monitoring method is capable in delivering a consistent heat and preventing any damages to the nearby tissue. Ultrasound is among the widely used technique in clinical setting. A-Mode ultrasound involves one-dimensional (1D) signal processing which enables a quantitative measurement on different types of breast tissues to be conducted faster as it has relatively simple signal processing requirement. On the other hand, B-Mode ultrasound offers good spatial resolution for thermal monitoring. Therefore, the aim of this study is to investigate and to compare the most optimum A-Mode and B-Mode ultrasound parameters to monitor hyperthermia in normal and pathological breast tissue. A series of experiment was conducted on 40 female Sprague Dawley rats. The pathological and normal rats were dissected and exposed to hyperthermia at variation temperature of 37oC (body temperature) and 40oC, 45oC, 50oC and 55oC for hyperthermia temperatures. A-Mode and B-Mode of 7.5 Mhz and 6Mhz was used simultaneously during the experiment for collecting acoustic information and scanning purposes before and after the hyperthermia exposure. Result obtained shows that, for normal tissue condition of both A-Mode and B-Mode, the attenuation calculation to mean of pixel intensity found to be (3.59±0.04)dB and 187.68 at temperature value of 50 oC. Meanwhile, in pathological tissue condition, the attenuation value with respect to pixel intensity was obtained by (3.36±0.26)dB at temperature value of 45oC and 199.26 was achieved at temperature value of 40oC. For backscatter coefficient to variance analysis, the result found that, in both A-Mode and B-Mode normal tissue condition, at temperature value of 40oC, (1.81±0.25) of backscatter coefficient was obtained while at 45oC, the variance value of 3298.94 was achieved. In pathological tissue, the temperature value of 40oC and 55oC was the most pronounce temperature dependent of (1.45±0.28) for backscatter coefficient with respect to 3275.35 of variance analysis. The result obtained from artificial neural network have shown that, 91.67% to 87.5% of testing to validation percentage accuracy of A-Mode was achieved, while in B-Mode, 88.89% and 81.25% of testing and validation data was obtained. Therefore, it is shown that, the use of A-Mode with comparison to B-Mode ultrasound can be used as another potential approach since its attenuation to pixel intensity and backscatter coefficient with respect to variance of A-Mode and B-Mode is very sensitive to the tissue structure in monitoring hyperthermia therapy with respect to the changes of temperature.