An evidence-based model for determining treatment dosages in therapeutic ultrasound using thermometry: an in-vitro investigation using post-mortem pig tissues

Abstract

The aim of this study was to clarify the relationship between the dosage parameters and temperature increase at the target tissues (up to 5 cm below the skin surface), and to explore the possibility of proposing a preliminary model to guide clinicians and researchers in determining treatment dosages based on expected increase in temperatures at the target tissue. Prior to the conduct of the main study several protocol-related issues were investigated. These included the reliability of the measurement procedures, the optimum speed of movement of the transducer, the optimum size of the treatment area, and the maximum output intensity that could be considered safe for treatment applications and investigations. An in-vitro post-mortem pig model was chosen for the experimental design using only adult-sized pigs, weighing between 60 to 80 kilograms. A total of 76 specimens were obtained from the shoulder and thigh sections of 19 pigs. The therapeutic ultrasound machine used throughout the study was the Omnisound 3000TM (Physio Technology Inc., Topeka, Kansas, USA) Output from the Omnisound 3000TM was checked and calibrated as necessary prior to each experiment using a power meter (Model UPM-DT-10, Ohmnic Instruments Co., St. Michaels, Maryland 21663, USA). Calibration was only performed when the checks demonstrated an error in the output intensity of the machine exceeded ±10%. The Minolta spot thermometer (HT-11, Minolta Co. Ltd., Japan) and the Avio thermal video system (TVS) 2000TM (Nippon Avionics Co. Ltd., Japan) were used to measure the change in tissue temperatures (dependent variable) at the skin surface and subcutaneously (at l, 2, 3, 4 and 5 cm below skin surface) respectively. The prepared specimen was mounted on a fixed table, with the clean cross-section facing the infrared thermographic camera.The camera to specimen distance was standardised at 50 cm for all experiments. Markers corresponding to 1, 2, 3, 4, and 5 cm on the specimen were plotted on the display unit, and saved to a 3.5 inch floppy disk. Measurements were recorded at baseline (prior to commencement of the experiment) and subsequently at 1-minute intervals during 10 minutes of exposure to the ultrasound, and for a further 10 minutes post-exposure, until the end of the experiment at 20 minutes. In general, there were five main parameters for all the studies: the movement speed of the transducer, the size of the treatment, and the frequency, intensity and duration of exposure and post-exposure to ultrasound. These five parameters represented the independent variables for all the studies. The dependent variable throughout was change in tissue temperature (measured in °C) at the skin surface, and at 1, 2, 3, 4 and 5 cm below the skin surface. Data were analysed using the SPSS for Windows software, Version 10.0 (SPSS Inc., 444N Michigan Avenue, Chicago, Illinois 60611, USA). Analyses of the data, using a repeated measures analysis of variance procedure, were performed on change in temperature, rather than actual temperature measured at selected time points. Only data from the 5th, 10th, 15th, and 20th minutes were analysed. This corresponded to the middle and end of the ultrasound exposure phase (5th and 10th minute) and post-exposure phase (15th and 20th minute), as these were considered to be representative of both these phases of data collection.Data for all 20-minute sampling is provided in the table of means for each experiment. The level of statistical significance was set at 0.05. Results of the reliability study showed that both the infrared spot thermometer and the video thermography unit were reliable within acceptable limits (as defined in this study). The latter, however, was more reliable than the former. In addition, the reliability was better for the post-exposure phase compared with the exposure phase, and for deeper tissues compared with the superficial and surface tissues. An unplanned analysis of the twenty minutes of data (at one minute intervals) suggested the possibility of reducing the duration factor from 20 to 4 (5th, 10th, 15th and 20th minute). In this manner, the data analyses for subsequent studies could be simplified considerably without affecting the overall results. Results of the other protocol-related studies showed that: a. There was no difference in change in temperatures between the slow (60 beats/min or 7cm/s), moderate (120 beats/min or 14cm/s) and fast (180 beats/min or 21cm/s) movement speeds of the transducer. However, for practical reasons, the moderate speed was recommended for subsequent studies; b. There was a significant difference in change in temperatures between the small (2X ERA), medium (3X ERA) and large (4X ERA) treatment sizes. The small treatment size provided the most effective and deeper heating, and was the recommended treatment size for subsequent studies; c. For both 1 and 3 MHz, tissue damage did not occur for intensities up to 1.5 Watts/cm[superscript]2. However, irreversible thermal injury to the tissues occurred at 2.0 Watts/cm[superscript]2 (1 MHz).Therefore, the recommended maximum intensity at which investigations could be carried out without any risks of thermal injury to the tissues was 1.5 Watt s/cm[superscript]2 for both 1 and 3 MHz. The results from the main study demonstrated that the increase in temperature due to absorption of the ultrasonic energy at any of the investigated target sites (up to 5 cm below surface) was related to the ultrasound frequency, intensity and duration of exposure. For the frequency factor, the evidence seems to suggest that compared with the 3 MHz ultrasound, the 1 MHz frequency may be more appropriate for clinical applications as it does not overheat surface tissues, and at the same time, is able to increase the temperatures of target tissues up to a depth of 5 cm. For the intensity factor, the results suggest that the therapeutic range of intensities which can be considered neither too low (as to be ineffective) nor too high (as to be damaging) are 0.5 to 1.3 Watts/cm 2 and 0.3 to 0.5 Watts/cm[superscript]2 for 1 and 3 MHz respectively The narrow therapeutic range for 3 MHz could render it questionable for clinical applications. In contrast, the larger therapeutic range available for the 1 MHz frequency suggests that it is more suitable for clinical applications and research. For the duration factor, the results demonstrated that the temperatures at all tissue sites increased as the duration of exposure increased.However, for the post-exposure phase, while the superficial tissues decreased with time, the deeper tissues continued to increase in their temperatures, albeit gradually. In summary, the results demonstrated that a higher frequency, a higher intensity, a greater exposure time and a more superficial site all contribute to a greater change in mean temperature. From these results, a preliminary model to guide clinicians and researchers in determining treatment dosages, based on expected increase in tissue temperatures at the target site, was proposed. While the preliminary model provided is only a first step effort, it is hoped that it can be refined further through use by physical therapists and other users of therapeutic ultrasound.