Impact analysis of beam-hardening CT-artefacts in radiotherapy planning
The methods of X-ray computed tomography being improved every year and we have a sufficient amount of methods and tools for high-quality visualization at that very moment. However, as practice shows, there are still low quality CTs (with the presence of strongly pronounced artifacts) in the developi...
Gespeichert in:
| Datum: | 2017 |
|---|---|
| Hauptverfasser: | , , , |
| Format: | Artikel |
| Sprache: | English |
| Veröffentlicht: |
Національний науковий центр «Харківський фізико-технічний інститут» НАН України
2017
|
| Schriftenreihe: | Вопросы атомной науки и техники |
| Schlagworte: | |
| Online Zugang: | http://dspace.nbuv.gov.ua/handle/123456789/136079 |
| Tags: |
Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
|
| Назва журналу: | Digital Library of Periodicals of National Academy of Sciences of Ukraine |
| Zitieren: | Impact analysis of beam-hardening CT-artefacts in radiotherapy planning / V.P. Starenkiy, I.O. Samofalov, L.L. Vasyliev, A.V. Trofymov // Вопросы атомной науки и техники. — 2017. — № 3. — С. 50-54. — Бібліогр.: 10 назв. — англ. |
Institution
Digital Library of Periodicals of National Academy of Sciences of Ukraine| id |
irk-123456789-136079 |
|---|---|
| record_format |
dspace |
| spelling |
irk-123456789-1360792018-06-16T03:03:09Z Impact analysis of beam-hardening CT-artefacts in radiotherapy planning Starenkiy, V.P. Samofalov, I.O. Vasyliev, L.L. Trofymov, A.V. Ядерно-физические методы и обработка данных The methods of X-ray computed tomography being improved every year and we have a sufficient amount of methods and tools for high-quality visualization at that very moment. However, as practice shows, there are still low quality CTs (with the presence of strongly pronounced artifacts) in the developing countries. The quality of the scans is of great importance in radiotherapy planning. An incorrect mapping of the density distribution will distort isodoses calculation. In the present article, we will consider the impact magnitudes of the beam hardening artifacts on the radiotherapy planning. Незважаючи на те, що методи рентгенівської комп'ютерної томографії вдосконалюються із року в рік, а також з'являється багато методів і засобів для високоякісної візуалізації, продовжують зустрічатися комп'ютерні томограми низької якості (з наявністю яскраво виражених артефактів). Найбільш часто ця тенденція спостерігається в країнах з нестійким економічним розвитком, де поряд з сучасним обладнанням для комп'ютерної томографії є застаріле покоління томографів. Якість візуалізації має велике значення для планування променевої терапії - некоректне відображення розподілу густини спотворить розрахунок ізодоз. В даній статті ми розглянемо величину впливу артефактів збільшення жорсткості пучка на планування променевої терапії. Несмотря на то, что методы рентгеновской компьютерной томографии совершенствуются из года в год, а также появляется значительное количество методов и средств для высококачественной визуализации, продолжают встречаться компьютерные томограммы низкого качества (с наличием ярко выраженных артефактов). Наиболее часто эта тенденция наблюдается в странах с неустойчивым экономическим развитием, где наряду с современным оборудованием для компьютерной томографии соседствует устаревшее поколение томографов. Качество визуализации имеет большое значение для планирования лучевой терапии - некорректное отображение распределения плотности исказит расчет изодоз. В данной статье мы рассмотрим величину влияния артефактов ужесточения пучка на планирование лучевой терапии. 2017 Article Impact analysis of beam-hardening CT-artefacts in radiotherapy planning / V.P. Starenkiy, I.O. Samofalov, L.L. Vasyliev, A.V. Trofymov // Вопросы атомной науки и техники. — 2017. — № 3. — С. 50-54. — Бібліогр.: 10 назв. — англ. 1562-6016 PACS: 03.65.Pm, 03.65.Ge, 61.80.Mk http://dspace.nbuv.gov.ua/handle/123456789/136079 en Вопросы атомной науки и техники Національний науковий центр «Харківський фізико-технічний інститут» НАН України |
| institution |
Digital Library of Periodicals of National Academy of Sciences of Ukraine |
| collection |
DSpace DC |
| language |
English |
| topic |
Ядерно-физические методы и обработка данных Ядерно-физические методы и обработка данных |
| spellingShingle |
Ядерно-физические методы и обработка данных Ядерно-физические методы и обработка данных Starenkiy, V.P. Samofalov, I.O. Vasyliev, L.L. Trofymov, A.V. Impact analysis of beam-hardening CT-artefacts in radiotherapy planning Вопросы атомной науки и техники |
| description |
The methods of X-ray computed tomography being improved every year and we have a sufficient amount of methods and tools for high-quality visualization at that very moment. However, as practice shows, there are still low quality CTs (with the presence of strongly pronounced artifacts) in the developing countries. The quality of the scans is of great importance in radiotherapy planning. An incorrect mapping of the density distribution will distort isodoses calculation. In the present article, we will consider the impact magnitudes of the beam hardening artifacts on the radiotherapy planning. |
| format |
Article |
| author |
Starenkiy, V.P. Samofalov, I.O. Vasyliev, L.L. Trofymov, A.V. |
| author_facet |
Starenkiy, V.P. Samofalov, I.O. Vasyliev, L.L. Trofymov, A.V. |
| author_sort |
Starenkiy, V.P. |
| title |
Impact analysis of beam-hardening CT-artefacts in radiotherapy planning |
| title_short |
Impact analysis of beam-hardening CT-artefacts in radiotherapy planning |
| title_full |
Impact analysis of beam-hardening CT-artefacts in radiotherapy planning |
| title_fullStr |
Impact analysis of beam-hardening CT-artefacts in radiotherapy planning |
| title_full_unstemmed |
Impact analysis of beam-hardening CT-artefacts in radiotherapy planning |
| title_sort |
impact analysis of beam-hardening ct-artefacts in radiotherapy planning |
| publisher |
Національний науковий центр «Харківський фізико-технічний інститут» НАН України |
| publishDate |
2017 |
| topic_facet |
Ядерно-физические методы и обработка данных |
| url |
http://dspace.nbuv.gov.ua/handle/123456789/136079 |
| citation_txt |
Impact analysis of beam-hardening CT-artefacts in radiotherapy planning / V.P. Starenkiy, I.O. Samofalov, L.L. Vasyliev, A.V. Trofymov // Вопросы атомной науки и техники. — 2017. — № 3. — С. 50-54. — Бібліогр.: 10 назв. — англ. |
| series |
Вопросы атомной науки и техники |
| work_keys_str_mv |
AT starenkiyvp impactanalysisofbeamhardeningctartefactsinradiotherapyplanning AT samofalovio impactanalysisofbeamhardeningctartefactsinradiotherapyplanning AT vasylievll impactanalysisofbeamhardeningctartefactsinradiotherapyplanning AT trofymovav impactanalysisofbeamhardeningctartefactsinradiotherapyplanning |
| first_indexed |
2025-07-09T21:23:22Z |
| last_indexed |
2025-07-09T21:23:22Z |
| _version_ |
1837206029071286272 |
| fulltext |
IMPACT ANALYSIS OF BEAM-HARDENING
CT-ARTEFACTS IN RADIOTHERAPY PLANNING
V.P. Starenkiy,∗, I.O. Samofalov, L.L.Vasyliev, A.V. Trofymov
State Institution ”Grigoriev Institute for medical Radiology NAMS of Ukraine”, Kharkiv, Ukraine
(Received April 25, 2017)
The methods of X-ray computed tomography being improved every year and we have a sufficient amount of methods
and tools for high-quality visualization at that very moment. However, as practice shows, there are still low quality
CTs (with the presence of strongly pronounced artifacts) in the developing countries. The quality of the scans is of
great importance in radiotherapy planning. An incorrect mapping of the density distribution will distort isodoses
calculation. In the present article, we will consider the impact magnitudes of the beam hardening artifacts on the
radiotherapy planning.
PACS: 03.65.Pm, 03.65.Ge, 61.80.Mk
1. INTRODUCTION
With the development of computerized means of in-
formation processing, more and more voluminous and
complex tasks are assigned to automated software
in many areas of human activity. A striking exam-
ple of such an area is radiation therapy, where au-
tomation of planning and technical implementation
of treatment has allowed achieving significant results:
three-dimensional planning replaced two-dimensional
one, the emergence and development of IGRT, IMRT,
RapidArc, VMAT and SRS techniques. Progressiv-
ity of automation is undeniable, but along with im-
proving the methods of processing information, we
have to rely increasingly on the accuracy and relia-
bility of the results of such methods and tools. Thus,
earlier (with manual RT planning) a medical physi-
cist (or radiologist) when receiving a low-quality im-
age or tomogram, could visually discard the obvious
image artifacts (any systematic discrepancy between
the CT numbers in the reconstructed image and the
true attenuation coefficients of the object [1]) dur-
ing planning and they did not significantly influence
dose calculation in the target volume; nowadays, vi-
sualization data of the internal body structures being
analyzed by software, usually not having the artifacts
auto-correction function. As a result the calculation
of isodoses can be made with gross dosimetric and
geometric errors. Such artifacts can be classified by
source of occurrence: on a hardware level, anatomical
(arising from the peculiarities of the internal struc-
tures of the studied object) and conditioned by the
human factor (incorrect research, whether it is a pa-
tient movement or a methodological error) [2]. Hard-
ware faults and the human factor are problems elim-
inated by sufficient qualification of maintenance per-
sonnel and by organization of a QA system. Thus,
features of the internal structures of the facility re-
quire the application of techniques for eliminating (or
weakening) such artifacts [3]. One of the most com-
mon artifacts is the consequence of the beam hard-
ening effect – an inhomogeneous weakening of the
beam over the photon energy spectrum. It means
that photons with lower energy in projections with a
significant density will be absorbed in a much larger
amount, which distorts the reconstructed image. A
particular (and most problematic) case of such an ef-
fect are metal artifacts (Fig.1). Over the past few
years, many methods of suppressing such artifacts
have been proposed [4-7], but they can be useful in
the primary tomogram processing only, i.e. at the
stage of designing and creating software for CT scan-
ners. A number of CT scanners (obsolete from this
point of view, most of which are in developing coun-
tries) still have weak opportunities for suppressing
such artifacts.
The authors of the article assume that correct
planning can be performed only with a satisfac-
tory quality tomogram (without contrast for diag-
nostic studies, without significant artifacts). How-
ever, in some cases the impossibility of performing re-
visualization during planning is still performed with
the available image.
The purpose of this article is to consider the
mechanism of influence on the process and the result
of RT planning of increasing / lowering the density in
the zone of interest on the reconstructed image due
to the beam-hardening effect and metallic artifacts
and to evaluate such influence.
∗Corresponding author E-mail address: imr@ukr.net, Tel.+38 0577255072, +38 0577255013
50 ISSN 1562-6016. PROBLEMS OF ATOMIC SCIENCE AND TECHNOLOGY, 2016, N3(109).
Series: Nuclear Physics Investigations (68), p.50-54.
a b c
Fig.1. Variations of the beam hardening effect in computed tomography:
a – in phantom study; b and c – in head CT with teeth implants
2. MATERIALS AND METHODS
As the purpose of this study is determination and
evaluation the artifact’s impact on planning, we need
to compare the estimate of the treatment plan per-
formed with a scans without artifacts, with an es-
timate of the plan with artifact ones. Phantoms
being often used for such purposes [8,9], but to bet-
ter approximate the real artefacts impact in this
study, we use tomographic images of real patients.
Even between two different visualizations of the same
internal structures of the same patient made in dif-
ferent period of time, differences appear not only
in the density distribution but also in the geometry
of these structures. In order to correctly estimate
the magnitude of the effect, the following research
method is applied: we duplicate the scans without
artifacts, add (with help of the planning system) the
changes in the density values of some regions typi-
cal for the manifestation of beam-hardening artifacts
in each localization considered. Then 3D conformal
radiotherapy (CRT) plans were made (photon radi-
ation 1.25MeV (Co-60)) and 6MeV (Varian Clinac
600C) using scans with added artifacts. After that
we imported scans without artifacts into the plan
and performed recalculation of isodoses with the
same fields parameters. Also, one should pay at-
tention to the localization choice for the studying of
the artifacts influence, as comparing the estimates of
treatment plans for different organs will not be cor-
rect. The question about the statistics of the areas
of such artifacts greatest propagation remains open,
but according to the preliminary collected data, the
most frequently pronounced artifacts of this type
occur during head and neck scanning (e.g., dental
implants) [10]; such artifacts, with the same localiza-
tion, presumably, will most heavily affect the result
of planning with irradiating volume located in the
oral cavity. To determine the value of the added den-
sity, statistical analysis of the average and maximum
density values (in HU) in visually conspicuous areas
of manifestation of artifacts are performed. It’s made
without taking into account the initial object of in-
creased density. To avoid the registration of random
calculated maxima, the maximum density values of
the artifacts were taken in order to be the maximum
of average density values of at least 9 nearby pixels
with a spread of values not more than 20%. Results
of analysis are shown in Table 1.
Table 1. Analysis of average and maximum density values
Effective (average) initial Average artifact density Maximum artifact density Number of analyzed
object density, HU value, HU value, HU tomograms
1000...1500 297 677 19
1500...3000 504 1119 16
3000...6000 987 2168 15
6000...9000 1659 3874 13
Given the weakening of the artifact appearance at
a distance from the initial object, adding of the arti-
fact is performed in two regions, one of which has the
average density (according to the average density of
the initial object for each tomogram) and the second
with an increased density from average. Quality as-
sessment method of the treatment plan is to select the
dose-volume histogram (DVH). We compare DVH by
average (effective), maximum and minimum doses for
controlled volumes. Within the study, we will com-
51
pare DVHs for treatment plans for tomograms with
manifestations of the most common average values of
the initial objects density (1000...1500HU) for dif-
ferent energies, as well as DVH comparison for treat-
ment plans compiled from tomograms with close to
extreme initial density objects (6000...9000HU).
3. RESULTS
Treatment plans comparisons results for 16 patients
with oral cavity tumors and delineated CTV are an-
alyzed. 8 patients have got scans most commonly
with a displayed average density of the initial object
in the range 1000...1500HU , then increased density
areas of 500HU were added before the treatment
plan calculation and 300HU also (Fig.2,a). Calcu-
lated plans and performed isodoses recalculation with
fields parameters saving for 1.25MeV and 6MeV en-
ergies were transferred to the reference CT without
added artifacts. A DVH comparison was made for
plans with an artifact and without an artifact for of
6MeV and 1.25MeV energies (Figs.3,a and 3,b, re-
spectively), in Tables 2.1, 2.2 the mean values of such
comparisons are shown. In 4 cases, plans were made
using scans with a high effective density of the initial
object (6000...9000HU), but without significant man-
ifestations of the artifact. The study method was the
same – adding objects with increased density (1500
and 1800HU), making treatment plans for 1.25MeV
and 6MeV photons, importing field parameters to
the plan according to set of scans without an ar-
tifact, etc. DVHs comparison results are shown in
Tables 2.3, 2.4. In 4 more cases, plans were made us-
ing low-quality tomograms with a high effective den-
sity of the initial object (6000...9000HU) and with a
pronounced artefact (Fig.2,b). The plans calculated
using such tomograms for 1.25MeV and 6MeV were
transferred to tomograms with corrected densities to
0 in the soft tissue zone, with the following saving
field parameters. DVHs comparison results are shown
in Tables 2.5, 2.6.
a b
Fig.2. Internal structures in planning:
a – added density distribution example; b – CT with high-density artefact example
a b
Fig.3. DVH comparisons: a – for plans 6MeV photons; b – for plans 1.25MeV photons
52
Table 2.1. DVH compares 6 MeV
Tomogram CTV mean CTV max CTV min brainstern brainstern brainstern
dose, % dose, % dose, % mean dose max dose min dose
with artefact 100 105.3 93.7 9.6 74.5 2.3
without artefact 99.6 105.6 93.6 9.6 74.6 2.3
with artefact 100 105.3 93.7 11.0 40.9 3.9
without artefact 100.2 105.6 93.6 10.9 40.4 3.9
with artefact 100 107.2 92.7 3.5 12.6 0.9
without artefact 102.8 106.6 93.3 3.9 14.5 0.9
with artefact 100 109.2 90.8 4.1 15.5 1.1
without artefact 105.8 111.7 93.5 6.2 21.3 1.1
with artefact 100 107.1 93.3 5.2 11.2 1.2
without artefact 102.7 106.8 93.9 5.8 12.1 1.2
with artefact 100 109.4 91.2 6.2 16.2 1.8
without artefact 105.1 112.1 92.9 6.7 18.8 1.8
Tab. 2.2 DVH compares 1.25 MeV
Tab. 2.3 DVH compares 6 MeV extreme density
Tab. 2.4 DVH compares 1.25 MeV extreme den-
sity
Tab. 2.5 DVH compares 6 MeV extreme density
Tab. 2.6 DVH compares 1.25 MeV extreme den-
sity
4. CONCLUSIONS
According to DVH comparison results, for tomo-
grams with the most common manifestations of
metallic artifacts in aforementioned location, they in-
troduce insignificant error (up to 0.4% for CTV mean
dose) even with 1.25MeV photons. However, ex-
treme values of artifact density can lead to an error
in the CTV mean dose of up to 3% for 6MeV and up
to 5...6% for 1.25MeV , which exceeds the accuracy
of the dose release set by the IAEA. In the subse-
quent article authors are up to studying the statistics
of the manifestations of metallic artifacts frequency
in detail, and also to consider and compare possible
methods for their suppression.
References
1. F. Julia. Barrett and Nicholas Keat: Artifacts in
CT: Recognition and Avoidance. RadioGraphics.
2004, v.24, p.1679-1691.
2. F. Edward Boas, Dominik Fleischmann. CT arti-
facts: causes and reduction techniques // Imag-
ing in Medicine. 2012, v.4, Issue 2, p.229-240.
3. Vincentas Veikutis. Artifacts in computer tomog-
raphy imaging: how it can really affect diagnostic
image quality and confuse clinical diagnosis? //
Journal of Vibroengineering. 2015, v.17 Issue 2,
p.995-1003.
4. R. E.Alvarez, A.Macovski. Energy-selective re-
constructions in X-ray computerised tomography
// Phys. Med. Biol.. 1976, v.21, p.733.
5. M.Bal, L. Spies. Metal artifact reduction in CT
using tissue-class modeling and adaptive pre-
filtering // Med. Phys. 2006, v.33, p.2852-2859.
6. M.Abdoli, M.R.Ay, A.Ahmadian,
R.A.Dierckx, H. Zaidi. Reduction of dental
filling metallic artifacts in CT-based attenuation
53
correction of PET data using weighted virtual
sinograms optimized by a genetic algorithm //
Med. Phys.. 2010, v.37, p.6166-6177.
7. E.Van de Casteele, D.Van Dyck, J. Sijbers, and
E.Raman. A model-based correction method for
beam hardening artefacts in X-ray microtomog-
raphy // Journal of X-ray Science and Technol-
ogy. May 2004, N12(1), p.43-57.
8. Kakuya Kitagawa. Characterization and Correc-
tion of Beam-hardening Artifacts during Dy-
namic Volume CT Assessment of Myocardial Per-
fusion // Radiology. July 2010, v.256, N1.
9. Junjun Deng. Beam hardening correction using a
conical water-equivalent phantom for preclinical
micro-CT. Nuclear Science Symposium and Med-
ical Imaging Conference (NSS/MIC), 2011 IEEE.
10. C. P. Law: Imaging the oral cavity: key concepts
for the radiologist // Br. J. Radiol. 2011, Oct,
N84(1006), p.944-957.
ÀÍÀËÈÇ ÂËÈßÍÈß ÊÒ-ÀÐÒÅÔÀÊÒΠÓÆÅÑÒÎ×ÅÍÈß ÏÓ×ÊÀ ÍÀ
ÏËÀÍÈÐÎÂÀÍÈÅ ËÓ×ÅÂÎÉ ÒÅÐÀÏÈÈ
Â.Ï.Ñòàðåíüêèé, È.À.Ñàìîôàëîâ, Ë.Ë.Âàñèëüåâ, À.Â.Òðîôèìîâ
Íåñìîòðÿ íà òî, ÷òî ìåòîäû ðåíòãåíîâñêîé êîìïüþòåðíîé òîìîãðàôèè ñîâåðøåíñòâóþòñÿ èç ãîäà â ãîä,
à òàêæå ïîÿâëÿåòñÿ çíà÷èòåëüíîå êîëè÷åñòâî ìåòîäîâ è ñðåäñòâ äëÿ âûñîêîêà÷åñòâåííîé âèçóàëèçàöèè,
ïðîäîëæàþò âñòðå÷àòüñÿ êîìïüþòåðíûå òîìîãðàììû íèçêîãî êà÷åñòâà (ñ íàëè÷èåì ÿðêî âûðàæåííûõ
àðòåôàêòîâ). Íàèáîëåå ÷àñòî ýòà òåíäåíöèÿ íàáëþäàåòñÿ â ñòðàíàõ ñ íåóñòîé÷èâûì ýêîíîìè÷åñêèì
ðàçâèòèåì, ãäå íàðÿäó ñ ñîâðåìåííûì îáîðóäîâàíèåì äëÿ êîìïüþòåðíîé òîìîãðàôèè ñîñåäñòâóåò óñòà-
ðåâøåå ïîêîëåíèå òîìîãðàôîâ. Êà÷åñòâî âèçóàëèçàöèè èìååò áîëüøîå çíà÷åíèå äëÿ ïëàíèðîâàíèÿ ëó-
÷åâîé òåðàïèè � íåêîððåêòíîå îòîáðàæåíèå ðàñïðåäåëåíèÿ ïëîòíîñòè èñêàçèò ðàñ÷åò èçîäîç.  äàííîé
ñòàòüå ìû ðàññìîòðèì âåëè÷èíó âëèÿíèÿ àðòåôàêòîâ óæåñòî÷åíèÿ ïó÷êà íà ïëàíèðîâàíèå ëó÷åâîé
òåðàïèè.
ÀÍÀËIÇ ÂÏËÈÂÓ ÊÒ-ÀÐÒÅÔÀÊÒI ÇÁIËÜØÅÍÍß ÆÎÐÑÒÊÎÑÒI ÏÓ×ÊÀ ÍÀ
ÏËÀÍÓÂÀÍÍß ÏÐÎÌÅÍÅÂÎ� ÒÅÐÀÏI�
Â.Ï.Ñòàðåíüêèé, I.Î.Ñàìîôàëîâ, Ë.Ë.Âàñèëü¹â, À.Â.Òðîôiìîâ
Íåçâàæàþ÷è íà òå, ùî ìåòîäè ðåíòãåíiâñüêî¨ êîìï'þòåðíî¨ òîìîãðàôi¨ âäîñêîíàëþþòüñÿ iç ðîêó â ðiê,
à òàêîæ ç'ÿâëÿ¹òüñÿ áàãàòî ìåòîäiâ i çàñîáiâ äëÿ âèñîêîÿêiñíî¨ âiçóàëiçàöi¨, ïðîäîâæóþòü çóñòði÷àòèñÿ
êîìï'þòåðíi òîìîãðàìè íèçüêî¨ ÿêîñòi (ç íàÿâíiñòþ ÿñêðàâî âèðàæåíèõ àðòåôàêòiâ). Íàéáiëüø ÷àñòî
öÿ òåíäåíöiÿ ñïîñòåðiãà¹òüñÿ â êðà¨íàõ ç íåñòiéêèì åêîíîìi÷íèì ðîçâèòêîì, äå ïîðÿä ç ñó÷àñíèì îáëàä-
íàííÿì äëÿ êîìï'þòåðíî¨ òîìîãðàôi¨ ¹ çàñòàðiëå ïîêîëiííÿ òîìîãðàôiâ. ßêiñòü âiçóàëiçàöi¨ ì๠âåëèêå
çíà÷åííÿ äëÿ ïëàíóâàííÿ ïðîìåíåâî¨ òåðàïi¨ � íåêîðåêòíå âiäîáðàæåííÿ ðîçïîäiëó ãóñòèíè ñïîòâîðèòü
ðîçðàõóíîê içîäîç.  äàíié ñòàòòi ìè ðîçãëÿíåìî âåëè÷èíó âïëèâó àðòåôàêòiâ çáiëüøåííÿ æîðñòêîñòi
ïó÷êà íà ïëàíóâàííÿ ïðîìåíåâî¨ òåðàïi¨.
54
|