Optimizing
prostate needle biopsy
through 3-D
simulation[1]
Jianchao Zeng, Charles Kaplan[2], John Bauer5, Jianhua Xuan, Isabell A. Sesterhenn[3]
John H. Lynch2, Matthew T. Freedman, and Seong K. Mun
Imaging Science and Information Systems Center
Department of Radiology, Georgetown University Medical Center
2115 Wisconsin Avenue, NW, Suite 603, Washington, DC 20007[4]
ABSTRACT
Prostate needle biopsy is used for the detection of
prostate cancer. The protocol of needle biopsy that is currently routinely used
in the clinical environment is the systematic sextant technique, which defines
six symmetric locations on the prostate surface for needle insertion. However,
this protocol has been developed based on the long-term observation and
experience of urologists. Little quantitative or scientific evidence supports
the use of this biopsy technique. In this research, we aim at developing a
statistically optimized new prostate needle biopsy protocol to improve the
quality of diagnosis of prostate cancer. This new protocol will be developed by
using a three-dimensional (3-D) computer-based probability map of prostate
cancer. For this purpose, we have developed a computer-based 3-D visualization
and simulation system with prostate models constructed from the digitized prostate
specimens, in which the process of prostate needle biopsy can be simulated
automatically by the computer. In this
paper, we first outline our approach to developing a statistically optimized
prostate needle biopsy protocol using the visualization and simulation system that
is overviewed next. In the preliminary experiments, we develop an
interactive biopsy simulation mode in the system, and evaluate the performance
of the automatic biopsy simulation with the sextant biopsy protocol by
comparing the results by the urologist using the interactive simulation mode
with respect to 53 prostate models. This is required to confirm that the
automatic simulation is accurate and reliable enough for the simulation with
respect to a large number of prostate models. In addition, we compare the performance
of the existing protocols using the automatic biopsy simulation system with
respect to 107 prostate models, which will statistically identify if one
protocol is better than another. Since the estimation of tumor volume is
extremely important in determining the significance of a tumor and in deciding
appropriate treatment methods, we further investigate correlation between the
tumor volume and the positive core volume with 89 prostate models. This is done
in order to develop a method to estimate the tumor volume from the
corresponding positive core volumes. Preliminary experimental results are given for each
of these experiments.
Keywords:
3-D Probability map, cancer distribution, computer visualization and simulation,
automatic vs. interactive prostate needle biopsy, 3-D interface, tumor volume
vs. positive needle core volume, statistical optimization of biopsy protocol
1. INTRODUCTION
Prostate
cancer is the most prevalent male malignancy and the second leading cause of
death by cancer in American men. In 1997, more than 40,000 deaths are predicted
for prostate cancer, and 342,000 new cases will be diagnosed. Current screening
tests for prostate cancer include prostate specific antigen (PSA) and digital
rectal exam (DRE). The combination of these two tests has led to an increased number of patients undergoing prostate needle biopsy. However, as noted
above, the accuracy of current biopsy techniques needs to be improved. A recent
article in Urology Times reported that patients undergoing repeat prostate
biopsy (after negative prior biopsy) still had a negative biopsy rate of 18%
for tumors of 0.5 cc to 1 cc; in addition, 15% of the repeat biopsies were
negative when the tumor burden exceeded 1 cc [Bankhead 1997]. Daneshgari
[Daneshgari et al. 1995] developed a 2-D computer simulation of the prostate
based on 159 radical prostatectomy specimens. The computer then generated
random prostates and tumors. This computer model was used to simulate the
sextant biopsy protocol and verify its ability to detect low-volume tumors.
Various biases for the angle of biopsy and distribution of cancer foci were
incorporated in the model. The simulation showed that only 20.3% of the
simulated prostates had a tumor distribution in which sextant biopsy had a 95%
probability of tumor detection. In fact, 26.8% of prostates had a distribution
that was completely outside the range the sextant locations. These prior
findings show that a significant number of patients who have prostate cancer
are not diagnosed at their initial biopsy. Accordingly, improving the
predictive value of TRUS guided biopsy by optimizing biopsy protocols will
improve its value as a screening and diagnostic tool.
A number of researchers have
investigated techniques for improving the accuracy of biopsy protocols;
however, several issues remain to be resolved. For example, Eskew et al.
introduced a new protocol called 5-region biopsy in which additional needles
are added systematically in addition to traditional sextant biopsy [Eskew et
al. 1997]. The 5-region biopsy and the traditional sextant biopsy were compared
with a total of 119 patients who underwent transrectal ultrasound guided needle
biopsy of the prostate. It was shown that of 48 cancer patients, 17 (35%) were
detected as having cancers only by the additional needles of the 5-region
biopsy method; within this group, 83% had Gleason scores of 6 or more. As a
result, the new 5-region biopsy method was claimed to improve biopsy results.
Eskew’s results are promising, but his study group of 48 patients is small.
Therefore, this protocol needs to be validated further, and the underlying
rationale for using 13 needles instead of some other number should be examined.
Our research group has developed a computer-based 3-D visualization of
digitized prostate specimens and has found that the 5-region protocol showed a
statistically significant advantage over the sextant method based on 107 3-D
prostate models [Kaplan, Zeng, Lynch et al. 1998a]. The computer simulates the
actual biopsy procedure for both the 5-region and sextant protocols by
calculating the location for each needle within the prostate and determining if
the needle has hit the cancer [Hayes et al. 1997]. Stamey [1995] conducted a
clinical study to evaluate existing biopsy protocols and to study the
correlation of the estimated findings with clinical significance. Goto [Goto et
al. 1996] has suggested that new biopsy strategies may be developed based on
probability maps of cancer distribution within the prostate. But issues such as
how these maps should be built and how new biopsy protocols could be derived
from the maps remain to be investigated.
Our research effort will
focus on the development of a statistically optimized biopsy protocol using a
3-D computer-based probability map of prostate cancer. A 3-D probability map
will be built by incorporating a large number of individual digitized prostate
specimens with localized cancers. This probability map gives the 3-D
distribution of prostate cancers and shows the probability of cancer detection
at each location in the prostate. Based on this probability map, a
statistically optimized needle biopsy protocol will be developed by
incorporating the needle locations with the highest probability of tumor
detection. This protocol will be compared and evaluated with existing biopsy
protocols, such as the sextant and 5-region, on our 3-D visualization and
simulation platform using 3-D prostate models reconstructed from the specimens
resected from the patients. Since the new biopsy protocol will be based on
statistical analysis of a quantitative database of digitized prostate
specimens, it may significantly improve the accuracy of prostate cancer
detection.
This paper is organized as
follows. Section 2 outlines the approach to optimizing a prostate needle biopsy
protocol. In section 3, the 3-D visualization and simulation system is
overviewed in some details. Preliminary experiments will be presented in
section 4 on performance evaluation of the automatic vs.
interactive biopsy simulation techniques,
on comparison of the biopsy protocols of sextant to 5-region using the
automatic biopsy simulation, and on the correlation of prostate tumor volume
and the positive needle core volume. Conclusions are made in section 5 with
future research directions.
2. OUTLINE OF
OPTIMIZING THE NEEDLE BIOPSY PROTOCOL
2.1 Construction of individual 3-D computerized
prostate models
The individual 3-D prostate models are constructed from radical prostatectomy specimens of prostate. The prostate specimens are step-sectioned in 4mm sections at 2.25mm intervals and then digitized with a scanning resolution of 1,500 dots per inch. Each digitized slice is then segmented by a pathologist to identify key pathological structures including surgical margin, capsule, urethra, seminal vesicles as well as tumor. The contours of each structure identified on each slice are then stacked and interpolation between the contours is carried out using a 3-D elastic model-based technique [Xuan et al. 1997]. The interpolation between adjacent contours C1 and C2 is completed by generating a force field that acts on C1 and forces it to gradually move and conform to C2. The 3-D model of each structure in the prostate is finally constructed by tiling triangular patches onto the interpolated contours using a deformable surface-spine model. This model uses a second order partial differential equation to control the deformation of the surface. After a 3-D model is constructed for each structure, an individual 3-D computerized prostate model can then be constructed by combining the structure models. Currently, more than 200 digitized prostate specimens have been acquired, and more than 100 3-D individual prostate models have been constructed using an SGI Onyx Infinite Reality 10000 Workstation. One example of a prostate model is shown in Figure 1.
Figure 1 An example of reconstructed prostate model.
2.2 Development of a 3-D integrated prostate model based on a large number of individual 3-D computerized models
To develop an integrated model, the average volume of
the sample computerized 3-D individual models are first calculated, which can
be automatically conducted by the current modeling system by accumulating
volumes confined between each pair of contours after contour interpolation
[Kaplan, Zeng, Lynch et al. 1998b]. Next, each individual model will be
normalized to the average volume. The normalization is conducted by uniform
scaling of the individual model in x, y and z directions in 3-D space [Udupa
and Herman 1989, Rogers and Adams 1990]. Each individual model will then be
registered in 3-D position to a common coordinate system (such as the screen
coordinate system) by translating its center of gravity to the coordinate
origin. Each model is also registered in 3-D orientation by first aligning its
central axis to the z-axis of the common coordinate system. The central axis is
the axis that is perpendicular to all the original contours of the prostate.
Then, the model is rotated about the z-axis to align the front of the prostate
(as defined in the contouring process) with the positive direction of the
y-axis.
2.3 Establishment of
a 3-D probability map of prostate cancer spatial distribution
Based on the 3-D integrated prostate model, the
spatial distribution of cancer will be obtained by calculating the occurrence
rate of the cancers at each 3-D position inside the prostate. Here the
occurrence rate Or(x,y,z) is defined as the number of tumors that contribute to
each location (x,y,z) inside the prostate divided by the total number of tumors
in the integrated prostate model. A greater occurrence rate represents a higher
probability of detection at the corresponding location.
2.4 Development of a statistically optimized biopsy protocol based on the 3-D probability map of tumor distribution.
A
biopsy protocol consists of several key parameters, including needle position,
needle angle, needle depth, and the number of needles. Since current clinical
instruments for needle biopsy fix the needle angle with respect to the
ultrasound probe, changing the needle angle will require new instrumentation
and therefore will not be considered in this study. In addition, the needle
depth using current triggering devices is fixed at 15-22 mm and therefore we
will not consider changing the needle depth in this study. We thus have two
parameters to investigate: needle position and the number of needles.
The 3-D probability map of
spatial distribution gives a quantitative measure of where the cancers are most
likely concentrated in the prostate. This map can be used to develop a
statistically optimized biopsy protocol with optimal needle positions. The
needle positions will be determined based on regions of high cancer
concentration. Here we propose the following algorithm to determine the
positions and the number of the needle in the new biopsy protocol.
Step 1: Calculate the occurrence rate Or(x,y,z) at
each location of the 3-D integrated model from the 3-D probability map of
prostate cancer spatial distribution.
Step 2: Select the location P(x1,y1,z1) that has the
largest the value of Or(x1,y1,z1) as the first needle location.
Step 3: Remove the individual models that have
cancers which have contributed to the location P(x1,y1,z1). Assume that the
number of such individual models is N1. It is anticipated that the number of
models that will be removed at this step will be relatively small compared to
the 200 sample cases. At this point, the sensitivity of cancer detection is
S1=N1 /200.
Step 4: Recalculate the Or(x,y,z) at each location of
the 3-D integrated model with N1 individual models removed as described in Step
3.
Step 5: Select the location P(x2,y2,z2) that has the
largest value of Or(x2,y2,z2) as the second needle location.
Step 6: Again, remove the individual models that have
cancers which have contributed to the location P(x2,y2,z2). Assume the number
of such individual models is N2. Therefore at this point, the sensitivity of
cancer detection is S2 = (N1 + N2 ) / 200. The sensitivity gain by the second
needle is SG2 = N2 / 200.
Step 7: Repeat this process as long as the
sensitivity gain SGi is larger than 2%. Assume that the last location selected
this way is P(xk,yk,zk).
Following the above
procedure, the positions P(xi,yi,zi) (i = 1, 2, ..., k) and the number k
of needles required will be
statistically optimized based on the 200 digitized prostate specimens.
2.5 Evaluation of the new protocol
The
evaluation process compares the new protocol against existing biopsy protocols
(sextant and 5-region) by examining the detection rate (based on finding at
least one positive core) and the number of positive cores. An additional 100
3-D individual models are constructed from new digitized prostate specimens for
this purpose. With these new models, there are 80% power to detect a 12%
difference in sensitivity (i.e., 80% versus 92%) when comparing the proposed
optimized biopsy approach to an existing biopsy protocol. The power calculation
was based on a matched-pair test with a 5% significance level using the method
of [Breslow and Day 1987] (Section 7.6.b). After testing the new protocol on
the 3-D visualization and simulation platform, a clinical trial needs to be
conducted where the new biopsy protocol will be evaluated on a large number of
patient subjects (e.g., 200 patients) to verify the effectiveness of the new
protocol. The same parameters measured in the simulation study need to be
investigated here, namely, the detection rate and the number of positive cores.
2.6 Data analysis
Data analysis is designed to assess the improvement in sensitivity of the new protocol over the existing prostate biopsy techniques. Sensitivity is defined as the proportion of individuals with the disease who have a positive test. Specificity, defined as the proportion of individuals without the disease who have a negative test, is 1.0 in the context of needle biopsies and so needs not be considered. Statistical tests to assess the differences between the two protocols include McNemar's test [Breslow and Day 1980] for comparing detection rates and the paired t-test for comparing quantitative variables (number of positive cores, length of positive cores, etc.) and the Wilcoxon signed-rank test. Differences in rates or means are considered statistically significant if they attain the 0.05 level of significance.
3. 3-D
VISUALIZATION AND SIMULATION SYSTEM
The visualization and simulation system has two simulation modes: an automatic simulation and an interactive simulation. The whole process of a prostate needle biopsy with any specific protocol can be simulated based on the reconstructed 3-D prostate models. This simulation system can be used to evaluate the performance of difference biopsy protocols. It can also be used as a training or testing system for the residents to practice their skills of biopsy, or a planning system for the urologists before they undergo a complex real biopsy procedure.
In the automatic simulation mode, the locations for needle insertion on the surface of the prostate are calculated automatically by the computer based on the requirement of the specific protocol. Thirty degrees of angle with respect to the local normal vector of the prostate surface are also calculated automatically for each needle. Needles are then mounted to the positions in the calculated poses. After shooting the needles, the system then detects which needle or needles are hitting the tumors inside the prostate, and if any, calculates the positive core volumes and displays the results on the screen. Since this whole process is controlled by the system, it can be finished quickly, making it possible to apply this simulation to a large number of samples (3-D prostate models) for statistical analysis if its performance can be confirmed. Each step of the biopsy simulation process can be visualized from any perspective by manipulating the 3-D prostate model in real time with a two-dimensional mouse. Figure 2 shows the needle locations on the prostate for the sextant (pink) and 5-region (pink + blue) protocols. Figure 3 shows the needles mounted in their initial locations and poses. Figure 4 shows the side view of the needles after being fired in the prostate. An example of needle biopsy results for both the sextant and 5-region protocols is shown in Figure 5.
Figure 2 Needle locations calculated for Figure 3 Needles mounted in their
the sextant and 5-region protocols initial locations and poses
Figure 4 Side view of the needles after Figure 5 An example of needle biopsy
being fired in the prostate result for sextant and 5-region
For the interactive simulation, six-degree-of-freedom tracking device has been integrated to simulate the ultrasound probe used during actual prostate biopsy procedure. The tracking device consists of an ultrasound transmitter, a controller, and a freely movable receiver device that serves as a tracker. With this mode, the system can track both the position (x, y, z) and the orientation angles (Pitch, Yaw, Roll) of the receiver in real time (50Hz). The tracking information is simultaneously used in controlling movement of a virtual ultrasound probe in the visualization and simulation system. The synthesized ultrasound images are refreshed in real time to follow the movement of the probe, which display intersectional anatomical slices of the prostate as biopsy guidance for the user (a urologist). With this interactive simulation mode, the urologist can perform a virtual needle biopsy as though he/she is performing a real biopsy on a patient. In the interactive simulation mode, the urologist determines the location for each needle insertion based on a specific protocol under the guidance of the synthesized ultrasound image. The angle of the needle is fixed with the ultrasound probe, and the upcoming path of the needle is always overlaid on the ultrasound image so that the urologist knows where the needle will go through inside the prostate. The result of a biopsy is automatically calculated by the system after each biopsy and displayed to tell the urologist whether the biopsy is positive or negative and how much the positive needle core volume is. Figure 6 shows the virtual ultrasound probe and the needle in use, while Figure 7 shows the synthesized ultrasound image with needle path and the fired needle.
Figure 6 Virtual ultrasound probe Figure 7 Synthesized ultrasound image
and the needle in use with needle path and fired needle
4. PRELIMINARY
EXPERIMENTS
4.1 Automatic vs. interactive needle biopsy
In order to verify the performance of the automatic biopsy simulation, comparison of the automatic and interactive biopsy simulation modes has been conducted with 53 3-D prostate models using the following three variables: rates of positive biopsy, positive core volumes, and the number of positive needles for each sample. The sextant protocol is employed. McNemar’s test is used for analysis of the rates of positive biopsy, and the results are shown in Table 1. Given the level of significance a=.05, since the test statistic T = b = 4 which is larger than t(=2) and smaller than n-t(=b+c-t=7), the null hypothesis is accepted which indicates that the difference between the two simulation methods is not significant [Conover 1980]. We have used paired-t test to analyze the other two variables. For the positive core volume, the results are shown in Table 2. Since the p value is 0.12, which is larger than 0.05 with test statistic t=1.581, it also indicates that the difference is not significant between the 2 biopsy simulation methods. The 95% confidence interval is [-7.3E-04, 6.14E-03]. For the number of positive needles, the results are shown in Table 3. Since the p value is 0.107, which is also larger than 0.05 with test statistic t=-1.642, these results also suggest that the difference is not significant. The 95% confidence interval is [-0.67, 6.71E-02]. As a result of this analysis, we may conclude that the automatic biopsy simulation is performing as well as a human urologist.
Table
1 Number of positive biopsies for automatic
and interactive biopsy simulation
Interactive
|
|
Positive |
Negative |
|
Positive |
a = 31 |
b = 4 |
|
Negative |
c = 7 |
d = 11 |
Table 2 Mean values of the positive core volumes
for automatic and interactive
biopsy simulation
|
|
Mean values |
|
Automatic biopsy |
10.6E-03cc |
|
Interactive biopsy |
7.88E-03cc |
|
Difference |
2.71E-03cc |
Table 3 Number of positive needles for automatic
and interactive
biopsy simulation
|
|
Mean values |
|
Automatic biopsy |
1.42 |
|
Interactive biopsy |
1.72 |
|
Difference |
-0.30 |
4.2 Comparison of current biopsy protocols: sextant
vs. 5-region with automatic simulation
We have evaluated sextant protocol against 5-region protocol in order to identify if there is any difference that is statistically significant, using the prostate needle biopsy simulation system with 107 3-D prostate models [Kaplan, Zeng, Lynch et al. 1998a]. This evaluation can provide evidence to show that one protocol is advantageous over another before applying them to the patients. After the new protocol is developed in this research, it will be first evaluated using the simulation system before conducting a clinical trial. We have used McNemar’s test to evaluate the rates of positive biopsy and used paired t-test to evaluate the positive core volumes. The results of the rates of positive biopsy for the sextant and 5-region protocols are shown in Table 4. Given the level of significance a=.05, since the test statistic T = b = 0 which is smaller than t(=1), the null hypothesis is rejected which indicates that the difference between the two protocols is significant. Five-region protocol is advantageous over the sextant protocol. In addition, the results of the positive core volumes are shown in Table 5. Since the p value is 0.00, which is smaller than 0.05 with test statistic t=8.907, it also indicates that the difference is significant between the 2 biopsy protocols. The 95% confidence interval is [7.51E-03, 1.18E-02]. As a result of this analysis, we may conclude that the 5-region protocol is performing better than the sextant protocol.
Table 4 Number of
positive biopsies for the sextant and 5-region protocols
Five-region
protocol
|
|
Positive |
Negative |
|
Positive |
a = 76 |
b = 0 |
|
Negative |
c = 8 |
d = 23 |
Table 5 Mean
values of the positive core volumes for the sextant
and 5-region protocols
|
|
Mean values |
|
Five-region protocol |
2.38E-02cc |
|
Sextant protocol |
1.42E-02cc |
|
Difference |
9.60E-03cc |
4.3 Correlation of tumor volume vs. positive core
volume
As a
step toward estimation of tumor volume from the corresponding positive core
volumes, we have investigated the correlation between these two kinds of
volumes. We have used 89 3-D prostate models on the needle biopsy simulation
system with both sextant and the 5-region protocols [Kaplan, Zeng, Lynch et al.
1998b]. Figure 8 shows a correlation result of tumor volume vs. positive core
volume with the sextant biopsy protocol. Table 6 gives the corresponding
correlation coefficient and the level of significance. Since the level of
significance is 0.002 which is smaller than 0.01, the correlation is
significant between the two volume variables with the sextant protocol. The
results with the 5-region protocol are shown in Figure 9 and Table 7, which
also indicates that the correlation is significant
Figure 8 Plot of tumor volume vs. core
Figure 9 Plot of tumor
volume vs. core volume
volume with sextant protocol with 5-region
protocol
Table 6 Correlation coefficient and the Table 7 Correlation coefficient and the level
of significance with sextant
protocol of significance
with 5-region protocol
5. CONCLUSIONS
In order to optimize the prostate needle biopsy protocols, we have developed a 3-D computer visualization and simulation system, and have conducted various experiments with a large number of 3-D prostate models based on the simulation system. To verify the performance of the developed visualization and simulation, an interactive biopsy mode is developed and integrated in the simulation system. A urologist can hold a six-degree-of-freedom tracking device to control a virtual ultrasound probe on the screen to guide his/her biopsy procedure. Since the tracking device can provide a real time tracking ability, the urologist can perform the virtual prostate needle biopsy as though he/she is using a real ultrasound probe on a patient. A synthesized ultrasound image is refreshed in real time following the movement of the tracking device in the user’s hand. With 53 sample prostate models on which the needle biopsies have been conducted by both the computer and the urologist using the sextant protocol, it is shown statistically that the biopsy simulation by the computer system is performing as well as the urologist. On the other hand, different biopsy protocols can be automatically simulated and evaluated for their performance analysis on our developed simulation system, which provides a quantitative measure to statistically compare various biopsy protocols before trying them on the patients. With 107 sample models, we have shown that the systematic 5-region protocol gives better results than the systematic sextant protocol. Further, to explore the possibility of estimating tumor volumes from the positive needle core volumes, we have investigated the possible correlation between the tumor volume and the positive needle core volumes with 89 3-D sample prostate models. A significant correlation is found between these two kinds of volumes, which provides an evidence that the tumor volume may be reasonably estimated from the positive needle core volumes. Finally, we have proposed a detailed algorithm to find a statistically optimized prostate needle biopsy protocol, which is currently being implemented in this research. Evaluation and data analysis methods are also provided and are being implemented.
With the advancement of new
imaging technologies, 3-D prostate models may be obtained in vivo, which will have a great impact on the way the prostate
needle biopsy is performed. More 3-D prostate models can be conveniently
acquired directly from the patients at screening tests, making a statistical
analysis more useful and powerful and reducing the time cycle of research.
Meanwhile, it becomes possible to develop an on-line prostate needle biopsy
system which provides real time augmented 3-D images to help a urologist to
quickly and precisely identify the abnormality in the prostate inside the
patient’s body. This idea is not limited to the prostate needle biopsy, it may
be applied to any type of biopsy, such as kidney biopsy. It may also be applied
to other surgeries beyond biopsies, making it possible to realize a real
on-site image-guided minimally-invasive surgery system.
6. ACKNOWLEDGEMENTS
We
would like to thank Dr. John Hanfelt of the Lombardi Cancer Canter of the
Georgetown University Medical Center for his helpful comments and discussion.
7. REFERENCES
Bankgead, C.: Sextant biopsy helps in prognosis of Pca, but it’s not foolproof. Urology Times, Vol. 25, No. 8, August 1997.
Breslow, N. E. and Day, N. E.: Statistical methods in cancer research: Volume I--The analysis of case-control studies. Lyon: International Agency for Research on Cancer, 1980.
Breslow, N. E. and Day, N.
E.: Statistical methods in cancer research: Volume II--The design and analysis
of cohort studies. Lyon: International Agency for Research on Cancer, 1987.
Conover,
W.J.: Practical nonparametric statistics (2nd edition). John Wiley
& Sons, Inc., 1980.
Daneshgari, F., Taylor, G.
D., Miller, G. J. and Crawford, E. D.: Computer simulation of the probability
of detecting low volume carcinoma of the prostate with six random systematic
core biopsies. Urology, 45: 604,
1995
Eskew, A. L., Bare, R. L. and
McCullough, D. L.: Systematic 5 region prostate biopsy is superior to sextant
method for diagnosing carcinoma of the prostate. J. Urol., 157: 199, 1997.
Flanigan, R. C., Catalona, W.
J., Richie, J. P., Ahmann, F. R., Hudson, M. A., Scadino, R. C., DeKernion, J.
B., Ratliff, T. L., Kavoussi, L. R., Dalkin, B. L., Waters, W. B., MacFarlane,
M. T. and Southwick, P. C.: Accuracy of digital rectal examination and
transrectal ultrasonography in localizing prostate cancer. J. Urol., 152: 1506, 1994.
Goto, Y., Ohori, M., Arakawa,
A., Kattan, M. W., Wheeler, T. M. and Scadino, P. T.: Distinguishing clinically
important from unimportant prostate cancers before treatment: value of
systematic biopsies. J. Urol., 156:
1059, 1996.
Hayes, W., Sesterhenn, I.,
Xuan, J., Wang, J., Lynch, J. and Mun, S. K.: Interactive 3-D modeling of
localized prostate cancer and computer simulation of needle biopsy techniques.
Annual Meeting of American Urology Society, 1997.
Kaplan, C. R., Zeng, J.
Lynch, J. H., et al.: Comparison of sextant to 5-region biopsy technique using
3-D computer simulation of actual prostate specimens. Accepted for presentation
at the 1998 Annual Meeting of American Urology Society, 1998a.
Kaplan,
C. R., Zeng, J. Lynch, J. H., et al.: Using three dimensional modeling of
localized prostate cancer to compare needle core volume to tumor volume.
Accepted for presentation at the 1998 Annual Meeting of American Urology
Society, 1998b.
Ohori, M., Wheeler, T. M.
Dunn, J. K., Stamey, T. A. and Scadino, P. T.: The pathological features and
prognosis of prostate cancer detectable with current diagnostic test. J. Urol.,
152: 1714, 1994.
Rogers, D. F. and Adams, J.
A.: Mathematical elements for computer graphics (2nd Edition). McGraw-Hill
Publishing Company, 1990.
Stamey, T. A.: Making the
most out of six systematic sextant biopsies. Urology, 45: 2, 1995.
Udupa, J. K. and Herman, G.
T.: 3D imaging in medicine. CRC Press, 1989.
Xuan, J., Hayes, et al.:
Surface reconstruction and visualization of the surgical prostate model. SPIE
Medical Imaging, 1997.