An application of visible human database in radiotherapy: tutorial for
image guided external radiotherapy (TIGER)
Vincenzo Valentinia, Nicola Dinapolia,*, Stefania Norib, Gian Carlo Mattiuccia,
Giovanna Mantelloc, Laura Maruccid, Maria Elena Rosettoe, Numa Cellinia
aRadiotherapy Department, Cattedra Radioterapia, Istituto Radiologia, Universita ` Cattolica del Sacro Cuore, Policlinico Universitario ‘A. Gemelli’
l. go Gemelli, 8, 00168 Rome, Italy
bInstitute of Anatomy, Universita ` Cattolica del Sacro Cuore, Rome, Italy
cRadiotherapy, Ospedale Torrette, Ancona, Italy
dRadiotherapy, Ospedale San Camillo, Rome, Italy
eRadiotherapy, Ospedale Grande degli Infermi, Viterbo, Italy
Background and purpose: Three-dimensional conformal radiotherapy and intensity modulated radiotherapy allow accurate dose delivery
on target volumes. Due to the different background among specialists involved in target volume definition, the contouring emerges as one of
the most questionable steps in treatment planning procedures. A software tool devoted to contouring training, named tutorial for image
guided external radiotherapy (‘TIGER’), based on the Visible Human Project images data-set, is described.
Materials and methods: TIGER is addressed to facilitate the learning of axial anatomical images, to promote the training and
reproducibility in contouring process, to allow the availability of a tool to enhance the ‘drill and practice’ approach in training programs.
TIGER includes three different environments: Anatomic tutorial devoted to facilitate a self-learning approach to axial body sections;
Contouring tutorial addressed to practice contouring process of anatomical structures and to undergo a test program prepared by tutors;
Teacher’s tools to offer to tutors the opportunity to insert new outlines in TIGER-database, according to local needs or conventions, and to
use them in tutorial programs.
TIGER-database is grouped in six main anatomical sections: head and neck, male thorax, female thorax, abdomen, male pelvis, and female
pelvis. Overall 432 corresponding CT-VH images and 1189 contours of 134 different anatomical structures and lymphatic drainage areas are
available. The access to the TIGER software is allowed by ESTRO web site (http://www.estro.be).
Conclusions: TIGER provides an interactive human anatomy cross-sectional oriented source to facilitate the interpretation of CT scan
images usually contoured in daily practice. It offers a drill tool to facilitate the learning of a reproducible contouring procedure.
q 2004 Elsevier Ireland Ltd. All rights reserved.
Keywords: Target volume definition; Human anatomy; Visible Human Project; Didactical tool
The improvements of technology in radiotherapy allow
physicians to deliver the dose on target volume and to spare
normal tissues everydaymoreaccurately.Three-dimensional
target, enabling the sparing of normal tissues and dose-
escalation programs ; intensity modulated radiation
therapy (IMRT) adds the possibility to define a sharper dose
definition of target volume and critical structures is of a
great relevance both in 3D-CRT and IMRT [2,8].
Several authors focused that a valuable inter- or intra-
physician variability in contouring can often influence the
treatment planning and can represent one of the most critical
steps in 3D-CRT [2,7] and IMRT procedures. One of the
reasons of differences in contouring target volumes may be
due to educational training of physicians, which is not
always addressed to promote the learning of cross-sectional
human anatomy but it is more frequently based on textbook
and autoptical dissection studies.
We report the features of a software developed for
tutoring young radiotherapists or the technicians involved in
0167-8140/$ - see front matter q 2004 Elsevier Ireland Ltd. All rights reserved.
Radiotherapy and Oncology 70 (2004) 165–169
* Corresponding author.
the treatment planning process to contour structures on
cross-sectional human images. The aim of tutorial for image
guided external radiotherapy (TIGER) project is to provide
an interactive human anatomy cross-sectional oriented
source to facilitate the interpretation of CT scan images
usually contoured in daily practice, and to offer a drill tool
to facilitate the learning of a reproducible contouring
2. Material and methods
The Visible Human Projectw(VHP) images dataset was
the source of anatomical data selected for the project. The
Visible Human Projectwis an outgrowth of the National
Library of Medicine’s (NLM) 1986 Long-Range Plan. Its
aim was the creation of complete, anatomically detailed,
three-dimensional representations of the normal male and
female human bodies by the acquisition of transverse CT,
cryosection and MR images of representative cadavers.
A male and a female cadaver were frozen and underwent
imaging acquisition procedures by CT scan, MRI and
standard radiological imaging. Subsequently, the two
subjects were dissected by axial planes and high-resolution
anatomic images were acquired. The complete series of
anatomical and radiological images constituted the male and
female Visible Human datasets.
The male dataset consists of axial CT scans of the entire
body taken at 1 mm intervals at a resolution of 512 pixels by
512 pixels. Each pixel is made up of 12 bits of grey tone.
The axial anatomical images are 2048 pixels by 1216 pixels
where each pixel is defined by 24 bits of color, each image
consisting of about 7.5 MB of data. The anatomical cross-
sections are also at 1 mm intervals and coincide with the CT
axial images. There are 1871 cross-sections for each mode,
CT and anatomy, obtained from the male cadaver. MR
images of the head and neck taken at 4 mm intervals and
longitudinal sections of the rest of the body at 4 mm
intervals are also available. The MR images have 256 pixel
by 256 pixel resolution; each pixel has 12 bits of grey tone
The dataset from the female cadaver has the same
characteristics as the male cadaver with one exception. The
axial anatomical images were obtained at 0.33 mm intervals
instead of 1.0 mm intervals. This results in over 5000
anatomical images. The female dataset is about 40 GB in
size. Spacing in the Z direction was reduced to 0.33 mm in
order to match the pixel spacing in the XY plane which is
0.33 mm [4,9].
NLM agreed us the access to the whole VHP data set
from their Web site for using the images for TIGER project.
Some technical problems had to be solved to utilize the
VHP database for the TIGER project: the difference in the
format and size of anatomical and radiological images and
the presence in the images of some part of the body, namely
arms or hands, useless for the project. Due to the different
sizes of anatomical and CT scan images a semi-automatic
procedure in order to increase the size of the CT scans and to
obtain the best overlapping result as possible was realized.
Nevertheless, the fine adjustment between anatomic and
radiological images was performed manually, by consider-
ing the reference of bone structures. Actually, there are little
differences in the position among inner anatomical struc-
tures in the two different sets of images, due to the different
chronological moments of the acquisition for anatomical
and radiological sections and to the little movements
occurred to the subjects between these two moments. To
optimize the graphical outcome the less useful portions of
the images, as arms or hands located over the abdomen,
were manually deleted by using an imaging elaboration
After the images’ preparation a software platform was
defined as a stand-alone Win32 compatible application. The
application was optimized to run under a Win32 operating
system (from Windowsw95 up to Windows XPw). The
optimal hardware features of the system require an MMXe
instructions set capable CPU, in order to take the advantages
of the enhanced graphic routines of the software in
managing wide images, with at least 64 MB (under
Windows 95–98) or 128 MB (under Windows NT, 2000,
XP) of RAM memory available.
TIGER database is grouped in six main anatomical
sections: head and neck, male thorax, female thorax,
abdomen, male pelvis, and female pelvis. Overall 432
corresponding CT-VH images and 1189 contours of 134
different anatomical structures and lymphatic drainage areas
The entire application, with the selected images, the
contours database and the application files and resources
fills a normal cd-rom space (less than 650 MB).
A multidisciplinary team with radiotherapists, anat-
omists, diagnosticians and surgeons, had in charge to define
the aim, the methodology and the validation program of
TIGER project, since 1999 . The team identified as the
main target of the project the training radiotherapists or the
technicians involved in the treatment planning process.
The TIGER team defined the features of the project: to
facilitate the learning of anatomy of axial images of the
body, to promote the training in the contouring process
and its reproducibility, to allow the availability of a
learning/teaching tool to enhance the ‘drill and practice’
approach in the training program.
The TIGER team had also in charge to outline a contour
database. It was addressed namely to support the identifi-
cation of the lymphatic drainage areas in the different
sections of the male and female body, because they coincide
very often with the CTV contours. The TIGER software
allows us to outline for each slice many different contours:
the team choose for head and neck to follow as guidelines
the lymphatic drainages outlines published by Gregoire ,
Nowak  and Martinez-Monge  (Fig. 1), for the
V. Valentini et al. / Radiotherapy and Oncology 70 (2004) 165–169166
remaining body only the lymphatic outlines defined
according to Martinez-Monge.
The structure of TIGER software includes three different
environments, everyone reachable by the main menu of the
1. Anatomic tutorial. The aim of this section is to
facilitate a self-learning approach to the axial body sections.
The user can see single or multiple contours rendered on
anatomic or CT scan images of the VHP dataset, selecting
the outlines through a list of anatomical site. Automatically
the system can select the sections with the chosen outline
and allows the user to move up and down in the head–feet
direction. The selection of the images is also facilitated by
choosing the level on a silhouette of the visible human
bodies, which can appear as skeleton model or as whole
body figure (Fig. 2). It is also possible to realize a blending
effect between the anatomical and CAT scan image (Fig. 3)
so that the user can focus on the differences among the
features of the anatomical structures shown in the two
corresponding images. The blending algorithm is really fast
despite the wide size of images managed, since it is
optimized also for recent computer hardware technologies
as Intele MMXe processors instructions set.
2. Contouring tutorial. The goal of this section is to allow
the user to practice contouring process of anatomical
structure and to undergo the test program prepared by the
tutor. In this section, the user selects the anatomical
structure, which he wants to test, and using the same
software tools available in the treatment plan console he
draws his outline. At the completion of the contouring
procedure the program compares the user outlines with the
‘validated’ outlines by the teacher stored in the TIGER
database, and evaluates the size of areas that are: (a) covered
by both the contours—it represents the successfully traced
contour; (b) covered only by the user’s contour—it indicates
the exceeding portion of the traced contour compared with
the teacher’s one; (c) covered only by the teacher’s
contour—it indicates the portion of the contour not covered
by the user. The user can also print a report (Fig. 4) or export
its data in a MicrosoftwExcel compatible format spread-
sheet. In the TIGER contour database besides the lymphatic
compartment according to Gregoire , Novak  and
Martinez-Monge  any further contours defined by the
teacher could be available.
3. Teacher’s tools. The aim of this section is to offer to
the tutors the opportunity to add new outlines to the TIGER
database, according to local needs or conventions, and to
use them in the tutorial program. Furthermore, in this
section the tutor can define a group of outlines that have to
be contoured by each student attending a course to
accomplish his participation at the practice session. It is
also allowed to monitor the track of each student in the
contouring training to evaluate and to certificate the
individual participation in the tutorial. This environment
can be safely protected to deny the access to common users
in order to protect the information stored in the database by
using a password.
TIGER is now currently used in the anatomy course of
the Postgraduated School in Radiotherapy of our University
and validated as a tool of the drill and practice program for
virtual simulation training of our residents.
An evaluation free version of TIGER is available from
the ESTRO web site (http://www.estro.be) for download;
the package contains a smaller series of images for each
anatomical sections collected in order to facilitate the
download operation. The radiotherapists can also request to
ESTRO secretary the full version of the software, which will
be sent by mail, after the acceptance of the user License
Fig. 1. Contours’ rendering in a TIGER CT image: on the left side there are
the lymphatic drainage contoursfor that slice according to Martinez-Monge
, on the right side according to Gregoire .
Fig. 2. Selection of the outlines facilitated by the indication of the level on a
silhouette of the visible human bodies.
V. Valentini et al. / Radiotherapy and Oncology 70 (2004) 165–169167
Several authors focused that a valuable inter- or intra-
physician variability in contouring can often influence the
treatment planning and can represent the most variable step
in 3D-CRT and IMRT procedures: Tai  in an
investigation about the variability of target volume
delineation in cervical esophageal cancer observed that a
substantial inconsistency in defining the planning target
volume among different radiation oncologists could offset
the potential benefits of 3D treatment planning with high
precision dose delivery. Fiorino  evidenced a substantial
inter-observer variability outlines definition also for
well-trained radiotherapists involved in CRT contouring
procedures for prostatic cancer. Cazzaniga  analyzed the
interphysician variability in the irradiation of prostate and
seminal vesicles focusing on the need of a specific training
for the physician and of detailed protocols for reducing the
variability in interpreting diagnostic images. Senan 
studied the impact of the implementation of a protocol for
the contouring of target volumes in lung cancer and they
found significant inter-radiation oncologists variations in
target volumes despite the use of this institutional contour-
ing protocol. Finally, Seddon  described an investigation
of inter-observer variability in GTV and normal structure
outlining for clinical cases of prostatic cancer, they
indicated particular regions of clinical uncertainty (as
prostatic apex, superior aspect of the prostate projecting
Fig. 3. An example of interactive ‘blending’ tool, to display the contours’ location on the anatomical images and on CAT scan slices in real time.
Fig. 4. Print out of the report after a contouring test.
V. Valentini et al. / Radiotherapy and Oncology 70 (2004) 165–169168
into the bladder, seminal vesicles and superior rectum) and
indicated the need for requiring a uniform approach to the
outlining of any individual structure that could also be
achieved by education of clinicians.
This aim of TIGER project is to provide a radiotherapist
oriented solution to the problem of teaching and learning the
axial human anatomy: it favors comparison between highly
detailed anatomical section and CAT images, training in
contouring outlines of anatomic structures and of lymph-
node drainage compartments, evaluation of individualized
tutorial programs, development of inter-comparison con-
Tutorial of different level of users can be managed easily.
The teacher can define individualized sets of outlines, which
the users have to contour, and the minimum percentage of
correct overlapping area accepted for each outline. The
system updates the record of each student and allows the
tutor to monitor the test program.
TIGER project has some limitations: it allows only the
analysis of VHP images, the contouring of pathological
slices is till now not available. In the following release a
DICOM importing module will provide the possibility to
browse, to display, to analyze and to contour the images
acquired by a DICOM compliant CT scan equipment. It
would allow the extension of the human anatomy learning to
actual clinical cases. Another limitation is the possibility to
obtain only a two-dimensional analysis of data of structures
contoured. At the moment no three-dimensional visualiza-
tion or volume analysis tool is available, but also this feature
could be introduced with the future developments of this
The contouring of the tumor and of the anatomical
structures can influence deeply the treatment planning in
3D-CRT and IMRT procedures. Several authors focused
that a valuable inter- or intra-physician variability in
contouring can represent the most variable step in
approaching new technologies in radiotherapy. One of the
reasons of differences in contouring target volumes may be
due to different educational training of physicians and
technologists involved in the process and to the missing of
tools to stimulate inter-comparision in the staff.
TIGER project allows us to promote the learning of
cross-sectional human anatomy, starting from the same
axial images utilized for treatment planning of patients who
have to receive a radiation therapy treatment. The source of
anatomical and corresponding CT images is The Visible
Human Project (VHP). The software allows us to identify
anatomical structure outlines, facilitating a real time
visualization of the correspondence between anatomical
and CT images.
TIGER is drawn to facilitate the training of people
committed to perform contouring of anatomical structures
by CT images. Tutors can define individualized sets of
outlines, which the user has to contour to test and to evaluate
his skill. This tool can also be used to promote inter-
comparison between radiotherapy staff. TIGER is an open
tool and it will be updated according to the suggestions of
the users and to the aims of the project.
 Cazzaniga LF, Marinoni MA, Bossi A, et al. Interphysician variability
in defining the planning target volume in the irradiation of prostate
and seminale vescicles. Radiother Oncol 1998;47(3):293–6.
 Fiorino C, Reni M, Bolognesi A, Cattaneo GM, Calandrino R. Intra-
and inter-observer variability in contouring prostate and seminal
vesicles: implications for conformal treatment planning. Radiother
 Gregoire V, Coche E, Hamoir M, Reychler H. Selection and
delineation of lymph node target volumes in head and neck conformal
radiotherapy. Proposal for standardizing terminology and procedure
based on the surgical experience. Radiother Oncol 2000;56:135–50.
 Martinez-Monge R, Fernandes PS, Gupta N, Gahbauer R. Cross-
sectional nodalatlas:a tool forthe definitionof clinical target volumes
in three-dimensional radiation-therapy planning. Radiology 1999;
 Nowak PJ, Wijers OB, Lagerwaard FJ, Levendag PC. A three-
dimensional CT-based target definition for elective irradiation of the
neck. Int J Radiat Oncol Biol Phys 1999;45(1):33–9.
 Seddon B, Bidmead M, Wilson J, Khoo V, Dearnaley D. Target
volume definition in conformal radiotherapy for prostate cancer:
quality assurance in the MRC RT-01 trial. Radiother Oncol 2000;56:
 Senan S, van Sornsen de Koste J, Samson M, et al. Evaluation of a
target contouring protocol for 3D conformal radiotherapy in non-
small cell lung cancer. Radiother Oncol 1999;53:247–55.
 Spitzer VM, Whitlock DG. National Library of Medicine atlas of the
visible human male: reverse engineering of the human body. Boston:
Jones & Bartlett; 1997.
 Tai P, Van Dyk J, Yu E, Battista J, Stitt L, Coad T. Variability of
target volume delineation in cervical oesophageal cancer. Int J Radiat
Oncol Biol Phys 1998;42:277–88.
 Valentini V, Dinapoli N, Nori SL, Maviglia R, Mattiucci GC,
Morganti AG, Cellini N. TIGER project: tutorial for image guided
external radiotherapy. Radiother Oncol 2000;56(Suppl. 1):S208.
 Vijayakumar S, Chen GT. Implementation of three dimensional
conformal radiation therapy: prospects, opportunities, and challenges.
Int J Radiat Oncol Biol Phys 1995;33:979–83.
V. Valentini et al. / Radiotherapy and Oncology 70 (2004) 165–169 169