Digital Signal Processing Reference
In-Depth Information
After therapy with transrectal highintensity focused ultrasound in ten pa
tients, MRI but not MRS, was found to provide some help in followup of
patients in conjunction with elevated PSA levels [430]. Imaging via MRI
used for surveillance after radical prostatectomy has also helped guide patient
management strategies [431].
11.1.4
Treatment planning and other aspects of clinical man-
agement
Other areas of clinical decisionmaking with respect to prostate cancer have
been impacted as well by MRSI. These include selection of therapeutic modal
ity, timing, and other aspects of treatment planning.
Comparisons have been made using MRSI in order to assess the effective
ness of threedimensional conformal external beam radiation therapy versus
permanent prostate implantation among fifty patients with low risk prostate
cancer. These comparisons showed the permanent implants to be more effec
tive in generating metabolic atrophy [432].
Detection of dominant intraprostatic lesions that should receive an esca
lated dose of radiation therapy with better sparing of surrounding normal
tissue has also been aided by MRSI [433]. More recently, it was found among
67 patients with biopsyproven prostate cancer who were followed up after
external beam radiation, that MRI and MRSI findings prior to therapy were
stronger independent predictors of outcome compared to clinical variables
[434]. The criteria for considering a volume of tissue “having unequivocal ma
lignant metabolism” on MRSI was that the choline peak was “clearly greater
than the citrate peak” (p. 666) [434].
11.1.5
Limitations of current applications of in vivo MRSI
directly relevant to prostate cancer
Notwithstanding the results heretofore achieved with MRSI in various aspects
of prostate cancer diagnostics and clinicaldecision making, shortcomings of
current applications hamper wider implementation of the potential for molec
ular imaging through magnetic resonance in this area of oncology.
We will now briefly summarize some of the major limitations of MRSI
based upon conventional, i.e., Fourierprocessing that are directly relevant to
prostate cancer. First of all, it should be noted that low citrate levels are not
pathognomonic of prostate cancer. Low citrate is also characteristic of normal
tissue in the stromal region of the prostate. In the face of metabolic atrophy,
citrate levels are also low without cancer being present. On the other hand,
hypertrophic prostate typically has high citrate levels, and this can still be
the case despite coexistent malignancy [7]. Thus, it is clear that reliance upon
ratios can be tenuous, since these can be affected by a number of processes
unrelated to the presence or absence of malignant prostate tissue.
Search WWH ::
Custom Search