Multiplet resonances in MRS data from normal and cancerous prostate - Signal Processing in MRS with Biomedical Applications - page 332

overdetection and overtreatment can entail morbidity and even mortality
[404]. Recently, the U.S. Preventive Services Task Force has recommended
against screening for prostate cancer among men over age 75 years, and con
siders the current evidence insu cient for making a decision regarding men
younger than age seventyfive [409]. The rationale for this recommendation
is that there is a lack of randomized, controlled studies, and that the cross
sectional and longitudinal evidence showed adverse psychological effects of a
falsepositive PSA screening result up to 1 year, while the benefits of screening
are considered to be uncertain [410].
Currently, once prostate cancer is suspected on the basis of, e.g., abnor
mal PSA plus digital rectal exam, prostate biopsy guided by transurethral
ultrasound (TRUS) is recommended. Since TRUS has limited sensitivity for
visualizing prostate cancers, at least eight, and according to some authori
ties twelve cores should be obtained [401, 402]. Even with such measures,
the sensitivity of TRUSguided biopsy is approximately 80%; in other words,
some 20% of prostate cancers are missed [402, 404]. On the other hand, given
the limited specificity of PSA, up to 75% of patients undergoing biopsy for
elevated PSA do not have cancer of the prostate [411].
Multiplanar T 2 weighted MRI has played an important role as a non
invasive tool for the diagnosis, staging as well as management of prostate
cancer, and is particularly valuable because of its high sensitivity [102]. How
ever, MRI poorly distinguishes benign from malignant lesions of the prostate,
such that its specificity is low [412]. This high negative predictive value of
MRI could help avoid unnecessary biopsies. However, because of its low pos
itive predictive value, many unnecessary biopsies would still be performed
based upon suspicious findings from MRI.
11.1.1 Initial detection of prostate cancer with in vivo MRS
and MRSI
Spectroscopic imaging through in vivo magnetic resonance has made a ma
jor impact in a wide range of areas related to prostate cancer diagnostics
and clinical decision making [102]-[104]. Compared to MRI alone, MRSI has
substantially improved the accuracy with which prostatic tumor and extra
capsular extension are detected, as well as helping to distinguish cancerous
prostate from benign prostatic hypertrophy [413]. As stated recently by Cas
ciani and Gualdi [414], “3D MRSI is emerging as the most specific tool for
noninvasive evaluation of prostate cancer”.
The major breakthroughs in prostate cancer diagnostics have relied heavily
upon assessment of the metabolic characteristics of larger areas of suspected
tumor, peritumoral regions and normal tissue. This is provided by MRSI.
Complete volumetric coverage of a suspected lesion is critical for accurate di
agnosis, and metabolically abnormal regions may not precisely correspond to
much larger T 2 abnormalities [415]. Vigneron et al. [216] point out the weak
ness of single voxel techniques that “provide one spectrum as in MRS which