RE : AUTOMATED PROSTATE VOLUME DETERMINATION WITH ULTRASONOGRAPHIC IMAGING

R.G.Aarnink, A.L.Huynen, R.J.B.Giesen, J.J.M.C.H.de la Rosette, F.M.J.Debruyne and H.Wijkstra

J.Urol.,153:1549-1554,1995

To the Editor. I was greatly impressed by the fact that the authors have been successful in outlining the prostate in the ultrasonic images automatically. Manual tracing of the prostatic contours has been inevitable not only in the planimetric volume calculation, but also in the ellipsoid volume calculation. In the prolate ellipse volume calculation, an operator has had to input six points to measure the three diameter. Among these volume calculation techniques, the planimetry is most accurate one but extremely time-consuming. Therefore, automated volume determination using edge detection technique and planimetry will become a routine. They stated that the automated method could not determine the prostate contours correctly only in one patient because of poor image quality, among 56 unselected patients. This rate of miscalculation is small enough, even though it may increase in a multicenter trial with urologists less experienced in ultrasound as they stated. Making use of this system, I wish they will calculate the parameters which I previously reported in order to discriminate prostatic cancer from hypertrophy, prostatitis, or normal prostates¹. About ten years ago, I calculated five parameters from prostatic sonograms by manually tracing the prostatic contours; prostatic volume (V), presumed circle area ratio (PCAR), anteroposterior diameter/transverse diameter (ATR), asymmetry index (ASI), and dissimilarity index (DSI). The weighted sum of V ,PCAR, ATR, and ASI was used to discriminate cancer and hypertrophy from prostatitis or normal prostates by the accuracy of 90%. The other value subtracting weighted sum of V and PCAR from that of ATR, ASI, and DSI discriminated cancer from hypertrophy by the accuracy of 70%. With an apparatus which outlines the prostate automatically, calculation of these values needs no additional procedure. These values are worth being calculated and printed on a report with prostatic volume, though the accuracy is yet insufficient for the automated diagnosis of prostatic diseases.

Respectfully,
Akira Kimura
Department of Urology
Branch Hospital
Faculty of Medicine
the University of Tokyo
3-28-6 Mejirodai, Bunkyo-ku
Tokyo
Japan


1. Kimura, A., Nakamura, S., Niizuma, M., Hoshino, T., Niijima, T., Ohashi, Y., and Higuchi, T.: Quantitative analysis of ultrasonogram of the prostate. J. Clin. Ultrasound, 14: 501, 1986.