Having discussed the contextual influences in Section 2.4, this section will discuss the combined implications on MRg thermal therapy performance.
MRg thermal therapy systems have an image update time which is presently in the order of several seconds. During this period of time, a thermal dose can accumulate without evidence of cell-death immediately showing up in the images, which can be unsafe. In order to address this safety issue, current systems monitor the accumulation of thermal dose in order to predict the probability of cell-death, rather than wait for the indications of cell-death having occurred.
Monteris Medical has demonstrated that surgeons, using the NeuroBlateR sys-
tem, can effectively deal with today’s quality of information being updated at eight second intervals[11]. This section describes the ways in which that performance space may be expanded when, for example, MR is able to deliver faster image updates.
Temperature at the periphery of the target treatment volume is a key factor in the performance space within which thermal therapy systems operate. The periphery represents an expanding region of cell-death. Regardless of instantaneous tempera- ture, tissue death becomes more likely with a greater accumulated thermal dose. A safety margin of two CEM43 has previously been identified[11]. Van Rhoon et al. [63] suggests that actual safety margins might be as high as 2 - 9 CEM43.
Figure 2.6 illustrates the performance space of this problem. The x-axis shows the temperature at the periphery of a tumour and increases, from left to right, from 37◦C (normal body temperature) to 53◦C (higher than typically desired at the periphery). The y-axis shows the time, in minutes, passing between updates of thermometry data. Five curves are plotted in the graph. Each curve represents thermal dose error tolerance. For any temperature, a time can be determined for a point on the lines. Any point on or below such a line will satisfy that desired thermal error tolerance. The change in angle at 43◦C is due to the formula for calculating CEM43. Equivalent thermal dose accumulates more slowly above 43◦C.
§2.5 System Performance 25
MR Guided Hyperthermia Performance Space
Instantaneous Peripheral Temperature ( ° C)
Time (min utes) 2 CEM43 1 CEM43 (8/60) CEM43 (1/60) CEM43 (0.2/60) CEM43 Faster MR Thermometry Finer Thermal Error Tolerance
Higher Temperatures 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 0 0.01 0.1 1 10 100 1000 10000 0.2 sec 1 sec 8 sec 1 min 2 min
Figure 2.6: An Evaluation Framework - The x-axis shows the temperature. The y-axis shows time. Five curves, representing different amounts of thermal dose, are plot- ted. 2 CEM43 (the top curve) is a commonly-used thermal error tolerance. The green shaded region represents the parameters within which current hyperthermia systems operate (i.e. with desired thermal error tolerance ≤ 2CEM43 and thermometry ca- pable of being delivered not faster than every 8 seconds.) The three arrows indicate the effects that temperature, time, and thermal error tolerance have on the operating range of device systems. Increased operating temperature, faster MR thermometry, or finer thermal error tolerance, all increase the range of conditions in which the
system may need to operate.
The area above the eight second line represents the performance space within which current systems such as NeuroBlateR operate. The area between the eight
second line and the one second line represents the performance space that systems will likely be required to operate in within the next five years. The area between the one second line and the 0.2 second line represents a performance space that systems may need to operate within, but for which a timeframe has not been predicted. The area below the 0.2 second line represents a performance space that would likely require significant automatic assistance to operate within, because a human surgeon
would not be able to react quickly enough to such information. Such a performance space would require a change in responsibilities, with the focus of the “human in the loop” shifting further towards supervision and exception handling.
Essentially, a typical hyperthermia system performs within the green region of the graph. As the capabilities of the system progressively increase, however, con- temporary visualisations become increasingly insufficient to the task (below and to the right of the green area). An increasing understanding of thermal damage mecha- nisms might affect the desired thermal error tolerance (2 CEM43 in this case). It might also limit the maximum peripheral temperature as a result of limiting the maximum tolerable amount of thermal accumulation between thermometry refreshes.
An example of how this figure may be used: If the system delivers thermal dose refreshes at eight-second intervals and the surgeon desires to apply heat with an accuracy of plus or minus 2 CEM43, then the maximum peripheral temperature fitting within these parameters would be approximately 47◦C. With time, the bottom of the green region can be expected to move downwards. The question, in this performance space illustration, is: How closely can the current refresh rate approach one second, before the surgeon starts to have difficulty dealing with that rate of information flow using contemporary interfaces? It is the objective of this research to be prepared for that time, by providing the “new contemporary” interface.