Thermocouple arrays for temperature measurements

							1984, The British Journal of Radiology, 57, 849-850                                                                 SEPTEMBER      1984

                                                         Correspondence
   We commend the author for his ingenuity and originality.           procedure other than direct injection of contrast medium into
However, we believe that in this patient if opacification of the      the relevant venous segments.
SVC and IVC was not optimal at contrast venography due to                                                          Yours, etc.,
the size of the collateral venous pathway then other imaging                                                          P. J. ROBINSON.
methods should have been utilised.                                    St. James's University Hospital
   Firstly, computed tomography with contrast injections given        Beckett Street,
via the venous catheters would have provided the necessary            Leeds LS9 7TF
information regarding the sites of obstruction, particularly          (Received April 1984)
with its inherent high contrast resolution (Engel, 1983).
   A second approach would have been to inject contrast                                        REFERENCE
medium via the venous catheters and use modern digital                COUPER, N. T. A., ROBINSON, P. J., HALL, T. J. & KESTER,
angiographic equipment to record and enhance the low iodine             R. C. 1982. Radionuclide and contrast venography following
signal.                                                                 caval interruption by the Mobin-Uddin umbrella filter.
   Thirdly, a radionuclide venogram would certainly have                Clinical Radiology, 33, 577-583.
confirmed the clinical suspicion of combined SVC and IVC
obstruction (Bentley, 1979).
                                              Yours, etc.,
                                               NOEL K. ALLAN,
                                                BHARAT RAVAL.
The University of Texas,
Medical School at Houston,
6431 Fannin,
Houston, Texas 77030
U.S.A.
{Received April 1984)
                                                                      Thermocouple arrays for temperature measurements
                        REFERENCES
                                                                      in vivo
BENTLEY, P. G., HILL, P. L., D E HAAS, H. A., MISTRY, F. &
  KAKKAR, V. V., 1979. Radionuclide venography in the                 THE EDITOR—SIR,
  management of proximal venous occlusion. A comparison                  The paper by Ter Haar and Dunn (1984) describes a method
  with X-ray contrast venography. British Journal of Radio-           of construction of a thermocouple array for use in temperature
  logy, 52, 289-301.                                                  measurements in vivo in mammalian tissue. It is suggested that
ENGEL, I. A., AUL, Y. O., RUBENSTEIN, W. A., SINDERMAN, K.,
                                                                      with such an array temperature measurement may be made
  WHALEN, J. P. & KAZAM, E., 1983. CT diagnosis of
                                                                      continuously during ultrasound application.
  mediastinal and thoracic inlet venous obstruction. American            However, during sonication local effects at the probe
  Journal of Roentgenology, 141, 521-526.                             junctions produce a rapid additional temperature rise. This is
ROBINSON, P. J., 1984. Access to the central circulation in the       due to the viscous forces acting between the probe and
  presence of combined superior and inferior vena caval               surrounding tissue, and results in measured probe tempera-
  obstruction. British Journal of Radiology, 57, 169-170.             tures being higher than surrounding tissue temperature while
                                                                      sonication continues. The magnitude of this rise will depend on
                                                                      the frequency and intensity of sonication, and on the
                                                                      orientation of the probe with respect to the field and the tissue
                                                                      (Fry & Fry, 1954a, b; Hynynen et al, 1982a, 1983a). This effect
                                                                      can give temperatures significantly higher than the surrounding
                                                                      tissue.
(Author's reply)                                                         Furthermore, the use of a coating on the probe junctions
                                                                      may add to this rapid rise. This is due to the larger ultrasound
THE EDITOR—SIR,                                                       absorption and shear viscosity coefficients of such materials
   Before embarking upon the catheter studies described in my         (Martin & Law, 1983). Particularly with focussed ultrasound
Technical Note, both radionuclide venography and contrast-            fields this effect becomes significant for clinical hyperthermia
enhanced CT were carried out. I agree that the radionuclide            where an accuracy of ±0.2°C or higher is usually required.
procedure is the preferred technique for confirming caval                Tissue temperature is better obtained by suspending
occlusion or patency (Couper et al, 1982); it certainly did so in     sonication shortly prior to, as well as during, measurement, in
this case. CT was helpful in excluding mediastinal and                order to allow for relaxation of the viscous forces and so that
retroperitoneal mass lesions and also showed the peripheral           the coating returns to thermal equilibrium with the surrounding
levels of occlusion nicely but did not demonstrate how much of        tissue. Normally, 200 milliseconds is adequate to achieve this,
the central caval segments were still patent.                         but a thick coating, due to its larger heat capacity, would
   The non-invasive tests established the diagnosis; the aim of       require a longer period of interruption to sonication before
the catheter studies was to pave the way for surgical treatment.      measurement.
Because some cases of mediastinal fibrosis have been reported            We have been using arrays of a similar construction to those
with pulmonary vascular involvement the surgeon needed to             of Ter Haar and Dunn with up to 16 junctions. Three such
know that right heart pressures were normal and that there            arrays have been monitored simultaneously with micro-
was no anatomical abnormality of the pulmonary vasculature.           processor control in phantom studies. Though there is no
It was also necessary to explore and define the limits of the         technical limit to the number of arrays monitored, in practice
central unobstructed segments of IVC and SVC with the                 patient tolerance to insertion of the probes may limit the
utmost clarity; I don't believe this could be achieved by any         number. The arrays have been used in phantoms (Hynynen

                                                                849