Guidelines for the Safe Use of Diagnostic Ultrasound - Rationale

4. Rationale

4.1 Thermal Effects

One of the major mechanisms for adverse biological effects from ultrasound exposure of mammalian systems is the heating of tissue via absorption of the ultrasonic beam (NCRP 1992). Therefore, guidelines have been developed in terms of exposure parameters directly related to temperature rise and the biological effects of heating.

4.1.1 Exposure Parameters

Implementation of the recommendations in this document requires a basic knowledge of the meaning of the new primary exposure parameter, the Thermal Index (TI) (AIUM/NEMA 1998a, Lopez 1998). This index is an estimate of the maximum temperature rise which could occur in ultrasonically heated tissue during an ultrasound examination. To distinguish it from an actual temperature elevation, the TI is unitless, being normalized to a temperature elevation of 1 °C. However, in varying with changes in the user control settings, the TI is directly proportional to the potential for heating. The Thermal Index is computed from directly measurable properties of the ultrasonic field, as determined in water under standard conditions. The methods of measurement are described in the Acoustic Output Measurement Standard for Diagnostic Ultrasound Equipment (AIUM/NEMA 1998b). The methods of computation are described in the Standard for Real-Time Display of Thermal and Mechanical Acoustic Output Indices on Diagnostic Ultrasound Equipment (AIUM/NEMA 1998a). A computed index is used because it is not feasible to monitor the actual temperature elevation in a clinical examination. In addition, the complexity of the conditions also precludes calculation of the actual temperature elevation.

There are three user-selectable TI categories which can be displayed (AIUM/NEMA 1998a, Lopez 1998). The Soft Tissue Thermal Index (TIS) is meant to be displayed for examinations in which the ultrasound beam travels a path which is made up principally of homogeneous soft tissue or a soft tissue/fluid path, as in a first trimester fetal examination or an abdominal examination. The Bone Thermal Index (TIB) is applicable to examinations in which bone is exposed to ultrasound, as could occur during Doppler blood flow examinations of a second or third trimester fetus. The Cranial Bone Thermal Index (TIC) pertains to examinations in which bone is at or very near the surface of the transducer, such as during transcranial, Doppler blood flow examinations.

A number of experimental and theoretical studies provide support for the three types of Thermal Index. Earlier studies have been thoroughly documented in two major reports on the subject (NCRP 1992, AIUM 1993). Since the publication of these documents, more recent clinical (Ramnarine, et al., 1993, Siddiqi, et al., (1995)) and experimental (Bosward, et al., 1993, O'Neill, et al., (1994), Duggan, et al, . 1995, Doody, et al., 1999) research has provided further evidence to support the models used to determine the different types of Thermal Index. The Standard for Real-Time Display of Thermal and Mechanical Acoustic Output Indices on Diagnostic Ultrasound Equipment (AIUM/NEMA 1998a)also provides brief rationales for the different types of Thermal Index.

4.1.2 Biological Effects

The clinical effect of an exposure depends on the nature and degree of tissue injury. This can be assessed from biological effects studies. Several extensive reviews have been published regarding the adverse biological effects of ultrasonic heating based on animal studies, particularly in mammalian species (Lele 1985, NCRP 1992, WFUMB 1992, AIUM 1993, WFUMB 1998). With regard to adult tissues, the available literature suggests that tissue temperature elevations in the range of 8-10 °C, sustained for 1 to 2 minutes will cause tissue injury (Bly, et al., 1992, Lele 1985). The reviews have also considered studies of teratogenic effects, usually on the developing brain, due to whole body heating of the embryo or fetus. The recommendations resulting from these reviews can be succinctly expressed as follows (WFUMB 1998):

1. a diagnostic ultrasound exposure that produces a maximum in situ temperature rise of no more than 1.5 °C above normal physiological levels (37 °C) may be used clinically without reservation on thermal grounds,
2. a diagnostic ultrasound exposure that elevates embryonic and fetal in situ temperature above 41 °C (4 °C above normal temperature) for 5 minutes or more should be considered potentially hazardous,
3. the risk of adverse effects is increased with the duration of exposure.
   

In addition, it has been reported that water immersion body heating of rats yielded the development of encephalocoeles in the rat fetuses in as little as 1 minute at a temperature elevation of 5 °C above normal physiological temperature. (WFUMB 1998).

For temperature elevations greater than 1.5 °C above normal physiological levels (37 °C), this information can be approximately matched to a functional form recommended by the NCRP (NCRP 1992). This yields an equation for combinations of temperature elevation and time which should be considered potentially hazardous:

where t is the time in minutes at the specified temperature and is the temperature elevation above normal (37 °C).

Barnett, et al., (1997) have recently published an updated review of thermal effects, focussing on the potential for effects on the fetus. They note that there is little information on the teratogenic effects from localized heat damage by ultrasound.

4.1.3 Human Exposure and Clinical Significance

To determine whether an exposure is justified, the equipment operator must assess whether reliable diagnostic information can be achieved and the severity and likelihood of an adverse health effect. To address the potential effects due to heating, estimates of the maximum exposures from various devices have been made in terms of the AIUM/NEMA Thermal Index and a method recommended by the National Council for Radiation Protection and Measurement (NCRP 1992) for estimating maximum temperature elevations.

As indicated in Sections 3.2 and 4.1.2, the dwell time is also an important parameter when considering the potential for a biological effect due to heating. One study found that the dwell time in fetal Doppler carotid artery exams was in the range of 4 to 80 seconds with a mean of 31 seconds (Duggan and McCowan 1993).

An assessment of the clinical significance of ultrasonic heating from diagnostic ultrasound devices depends upon estimates of the potential acoustic exposure. For equipment available prior to the implementation of the Output Display Standard in 1993, Patton, et al, . (1994) found no real-time B-mode devices with Thermal Indices exceeding 1, consistent with the conclusions of the World Federation for Ultrasound in Medicine and Biology (WFUMB 1998). The vast majority of M-mode devices yielded Thermal Indices less than 1, with the largest TIB being 1.4. The maximum Thermal Indices for general pulsed Doppler devices were 1.3 for soft tissue exposure (TIS) and 2.8 for exposure of bone (TIB). The majority of the Doppler console/transducer/mode/intended use combinations (samples) yielded Thermal Indices less than 1. For peripheral vascular devices, most of the samples yielded TIB values greater than 1. The maxima were 2.2 and 2.8, for soft tissue and bone respectively. For colour flow Doppler, six of the samples (19%) yielded TIS values greater than 1, with the maximum Thermal index being 2.3.

The results of one major survey of the directly measured properties of the ultrasound field, acoustic pressure and power, suggest that acoustic output may have risen since the implementation of the Output Display Standard. However, this is not clear from the information made available in the published study (Henderson, et al., 1995). It is plausible that changes to the U.S. FDA 510(k) guidance document in 1993 have led to an increase in the number of devices with acoustic output approaching the limits recommended in Section 2.4. Although two surveys have been made of Thermal Index values since that time (Shaw, et al., 1997, 1998), a comparison between devices sold before and after 1993 cannot be made from the published report.

If the path through the bladder for a 1^st trimester fetal examination is more than 5 cm, computed estimates of the maximum temperature elevation according to the NCRP method (NCRP 1992) can exceed those given by the AIUM/NEMA Thermal Indices by as much as a factor of 2-3.. The evidence on propagation paths during ultrasound examinations in the study by Ramnarine, et al., (1993), suggests that the TIS might be exceeded by a factor of 2-3 for as many as 40% of 1^st trimester transabdominal examinations where the path through the bladder is more than 5 cm.

However, other evidence supports the expectation that, only in more unusual circumstances, would the NCRP estimate indicate a significantly different need for exposure reduction than would be provided by the TI. For each diagnostic ultrasound device in the sample studied by Patton, et al., (1994), the TI was compared to the NCRP estimates. It was found that about 95% of the TIS values were within a factor of 2 of the corresponding NCRP estimates. In addition, Bly, et al., (1992) found that for 236 samples, the NCRP estimate of maximum temperature elevation did not exceed 1.6 °C for trans-abdominal, pulsed Doppler examination of a first trimester fetus. Other calculations made for non-autoscanning transducers with well defined focusing geometries and a derated spatial peak time average intensity, ², (Bly, et al., 1992, AIUM 1993) also ^Ispta.3 , of 720 mW/cm yielded computed estimates of maximum temperature elevation which were approximately 2 °C.

There is also evidence to support the expectation that for 2^nd and 3^rd trimester exposures, only in relatively unusual circumstances would the NCRP estimate indicate a significantly different need for exposure reduction than would be provided by the TI. First, in the study by Ramnarine, et al., (1993), all of the propagation paths yielded ultrasound attenuation greater than or equal to 0.3 dB/cm-MHz, the value used as the basis for the tissue models in the Output Display Standard. Furthermore, the vast majority of samples in the study by Bly, et al, . (1992) yielded maximum temperature elevations of less than 4 °C for heating of second trimester fetal bone. In the study by Patton, et al, . (1994), the largest temperature elevation calculated according to the NCRP method was 5.9 °C. This occurred for a transducer with a 10.5 cm focal depth. This evidence suggests that only in highly unusual circumstances would there be a significantly different need for exposure reduction than would be provided by the TI.

Carstensen, et al., (1992) addressed the potential for ultrasonic heating during an echocardiography examination. They considered the primary heating concern to be the patient's rib(s), particularly if the patient was not able to provide an indication of discomfort or pain. It was noted that this would be the case in some pediatric examinations. Carstensen and co-workers calculated that most diagnostic ultrasound devices would not be able to ultrasonically heat the ribs by more than 1.5 °C. However, they did indicate that 1 pulsed Doppler model appeared to have the capability of heating to about 3-6 °C.

The information provided above suggests that, in most clinical examinations, exposures are not sufficient to cause adverse health effects due to ultrasonic heating. However, the maximum temperature elevations resulting from ultrasound exposure during Doppler blood flow examinations can be well above the normal diurnal variation of 1 °C. Also, in rare circumstances, the maximum potential temperature elevations appear to be near to thresholds for tissue injury for plausible clinical dwell times.

Transducer self-heating has been known to occur with diagnostic ultrasound devices and it has the potential to be a substantial source of heating (Duck, et al., 1989). Although the tissue heating should be localized to the region near the contact surface, the additional heating could be a cause for concern during transcranial, transvaginal, transrectal or transesophogeal examinations (NCRP 1992, WFUMB 1998).

Surveys of diagnostic ultrasound devices sold in Canada were last made in 1990. They indicated that, normally, devices were sold with acoustic intensities less than the limits recommended in Section 2.4 of these guidelines. These limits effectively constrain the potential for ultrasonic heating by diagnostic ultrasound devices. Therefore, having these limits on the device and using the real-time display of Thermal Indices, it should be possible to ensure that there is very little risk to the patient from ultrasonic heating.