Nuclear and Radiological Engineering

Computational Phantoms and Skeletal Dose  Models for Adult and Paediatric Internal Dosimetry

Wesley Bolch, PhD, PE, CHPCommittee 2 of the ICRP and Chair, DOCAL Task GroupNuclear & Radiological Engineering, University of Florida

Michael Wayson and Deanna PafundiNuclear & Radiological Engineering, University of Florida

IAEA IDOS Symposium ‐

Session 3B

Internal Dosimetry for Diagnostic and Therapeutic Nuclear MedicineComputational Phantoms & Imaging Based Patient‐Specific Models

November 10, 2010

Nuclear and Radiological Engineering

The ICRP

C2 Task Groups – DOCAL and INDOSC3 Task Group – Radiopharmaceuticals

Nuclear and Radiological Engineering

NCRP Report 160 Trends in ionizing radiation exposure

Early 1980s 2006

~15% medical0.53 out of 3.6 mSv

~48% medical3.0 out of 6.2 mSv

Nuclear and Radiological Engineering

AAPM Press Release March 3, 2009

The NCRP report does not, however, "attempt to quantify the associated  health risks nor specify the actions that should be taken in light of these 

latest data," and AAPM experts are cautioning that these data do not  necessarily indicate that the U.S. population is at any higher risk due to 

this increased use of medical imaging. They caution that the new

report  should not deter patients from getting medically‐appropriate imaging 

exams. The NCRP findings on average population dose could be easily  misinterpreted if applied to an individual patient’s medical situation. 

Nuclear and Radiological Engineering

Impetus from NCRP Report 160 on Medical Dosimetry  Retrospective

Dosimetry Studies

•

Radiation epidemiological studies

•

Quantifying past exposures and construction of dose‐response  correlations

•

Emphasis on pediatric exposures

•

Examples ‐

NCI Radiation Epidemiological BranchStudy of pediatric CT imaging•

Retrieval of pediatric imaging records in the UK

•

Phase I –

Cohort study of 200,000 individuals (1985 to 2002)

•

Phase II – Nested case control study of 1000 individuals

•

Leukemia, brain, thyroid, breast cancers

Childhood Cancer Survivor Study (CCSS)

Nuclear and Radiological Engineering

Impetus from NCRP Report 160 on Medical Dosimetry  Prospective

Dosimetry Studies

•

Assignment of organ doses under specific imaging protocols

•

Recording of individual doses in electronic medical records

•

Optimization of patient dose versus image quality

•

Example – Pediatric Nuclear Medicine ImagingSurvey of 13 major pediatric hospitals (JNM 2008; 49:1024–1027)

•

16 radiopharmaceutical examinations were surveyed

•

Minimum / maximum activity

•

Activity per unit body mass or body surface area

Conclusions•

Maximum variations –

factor of 8.5 in amount administered

•

Average variation –

factor of 3

Nuclear and Radiological Engineering

Computational Anatomic Phantoms Essential tool for organ dose assessment

•

Definition

‐

Computerized representation of human anatomy for use in  radiation transport simulation of the medical imaging or radiation therapy 

procedure

•

Need for phantoms vary with the medical application–

Nuclear Medicine•

3D patient images sometimes not available, especially for children

–

Diagnostic radiology and interventional fluoroscopy –

no 3D image

–

Computed tomography•

3D patient images available, problem –

organ segmentation

•

No anatomic information at edges of scan coverage

–

Radiotherapy•

Needed for characterizing out‐of‐field organ doses

•

Examples – IMRT scatter, proton therapy neutron dose

Nuclear and Radiological Engineering

Computational Anatomic Phantoms Phantom Types and Categories

•

Phantom Format TypesStylized (or mathematical) phantoms

Voxel (or tomographic) phantoms

Hybrid (or NURBS/PM) phantoms

•

Phantom Morphometric CategoriesReference (50th percentile individual, patient matching by age only)

Patient‐dependent (patient matched by nearest height / weight)

Patient‐sculpted (patient matched to height, weight, and body contour)

Patient‐specific (phantom uniquely matching patient morphometry)

Nuclear and Radiological Engineering

Format Types ‐

Stylized Phantoms

1960sStylizedPhantom

Heart

LiverSpleen

Stomach

Small intestine

Ascending colonDescending colonUrinary bladder

Anatomy of ORNL stylized adult phantom

Flexible but anatomically unrealistic

Nuclear and Radiological Engineering

Format Types ‐

Voxel Phantoms

Anatomically Realistic but not very flexible

Lungs

Heart

LiverColon

Small intestine

Urinary bladder

Testes

Anatomy of Korean male voxel phantom

1980sVoxel

Phantom

Nuclear and Radiological Engineering

Format Types –

Hybrid Phantoms

2000sHybrid

Phantom

Realistic and flexible

LungsHeart

Liver

Stomach

Colon

Small intestine

Urinary bladder

Anatomy of UF hybrid adult male phantom

Nuclear and Radiological Engineering

Format Types –

Hybrid Phantoms

Hybrid phantom

Tomographic phantomStylized phantom

Mathematical Flexibility(NURBS / PM surface)

Anatomical Realism(CT images of patient)

Non‐Uniform Rational B‐Spline

Nuclear and Radiological Engineering

Format Types –

Hybrid Phantoms

Segmentation Polygonization

NURBS modeling Voxelization

Segment patient 

CT images using 

3D‐DOCTOR

Convert into 

polygon mesh 

using  3D‐DOCTOR

Make NURBS 

model from 

polygon mesh 

using

Rhinoceros

Convert NURBS 

model into voxel 

model using 

MATLAB code 

Voxelizer

Nuclear and Radiological Engineering

Format Types –

Hybrid PhantomsSome selected and recently published hybrid phantoms

• XCAT Series ‐

Segars and Tsui  (Proc IEEE, 2009)• RPI Series –

Zhang et al (PMB 2009) and Xu et al (PMB 2007)

• Virtual Family – Christ et al (PMB 2010)• FASH and MASH – Cassola et  (PMB 2010)• UF Series ‐

Lee et al (PMB 2010)

XCAT Series FASH Virtual Family

Nuclear and Radiological Engineering

Morphometric Categories –

Reference PhantomsReference Individual ‐

An idealised male or female with characteristics 

defined by the ICRP for the purpose of radiological protection, and with  the anatomical and physiological characteristics defined in ICRP

Publication 89 (ICRP 2002). 

Note –

While organ size / mass are specified in an ICRP reference phantom, organ shape, depth, position within the body are not defined by reference values

Nuclear and Radiological Engineering

Reference Phantoms Used by the ICRPEssentially all dose coefficients published to date by the ICRP are based on computational data generated using the ORNL stylized phantom series.

ORNL TM‐8381Cristy & Eckerman

One exception is ICRP Publication 74 on external dose coefficientsReference data taken from a variety of both stylized and voxel phantoms

Nuclear and Radiological Engineering

Reference Phantoms Adopted by the ICRPICRP Publication 110 – Adult Reference Computational Phantoms 

Upcoming Publications from ICRP using the Publication 110 Phantoms• Reference dose conversion coefficients (DCC) for external radiations (revision of ICRP 74)• Reference DDC for space radiation environments• Reference DCC for aircrew exposures• Reference absorbed fractions (AF) for internal dose coefficients / nuclear medicine

Nuclear and Radiological Engineering

Reference Phantoms Adopted by the ICRPIn September 2008, ICRP established that its future reference phantoms forpediatric individuals would be based upon the UF series of hybrid phantoms 

Nuclear and Radiological Engineering

Dosimetry Applications of Hybrid Phantoms Nuclear Medicine Imaging

Comparison of UFH15F to ORNL Stylized 15‐year

Nuclear and Radiological Engineering

Dosimetry Applications of Hybrid Phantoms Nuclear Medicine Imaging

Ratio of Tc‐99 S‐valueORNL  

/ S‐valueHybrid

Target Organs Heart Kidneys Liver Lungs Skeleton Spleen UB ContActive Marrow 2.4 0.8 1.4 0.9 1.4 1.4 0.8Brain 1.8 0.3 0.8 1.2 6.2 1.2 0.6Breasts 2.5 0.9 1.0 1.6 22.7 1.5 3.6Liver 2.3 0.9 1.3 1.3 10.0 1.0 2.7Lungs 1.1 0.4 0.6 0.8 1.3 0.7 1.1Ovaries 52.0 1.8 5.9 3.6 2.6 6.5 0.3Skin 3.8 0.8 1.6 1.7 11.7 1.0 2.0Stomach Wall 2.2 0.7 0.6 1.0 6.2 1.1 2.7Thyroid 1.0 0.3 0.6 0.6 4.4 0.8 1.2Urinary Bladder Wall 9.1 1.3 2.7 2.6 3.1 2.6 2.6

Source Organs

Differences are attributed to changes in 15‐year female  reference masses as well as inter‐organ spacing

Nuclear and Radiological Engineering

Definition ‐Expanded library of reference phantoms covering a range of height / weight percentiles

NHANES Database7320 individuals

AgeWeightStanding heightSitting heightBMIBiacromial breadthBiiliac breadthArm circumferenceWaist circumferenceButtocks circumferenceThigh circumference

ICRP - based UFHADM

US based phantom library10% 25% 50% 75% 90%

Reference weights @ 1 or more fixed anthropometric parameter(s)NHANES - based

UFHADM

Morphometric Categories –

Patient Dependent Phantoms

Nuclear and Radiological Engineering

Morphometric Categories –

Patient Dependent Phantoms

US Adult Male 

Standing Heights

US Adult Male 

Sitting Heights

Nuclear and Radiological Engineering

Method A, Scaling treePatient Dependent Phantoms – Adult Males

Nuclear and Radiological Engineering

Patient Dependent Phantoms – Adult Males

Same height / different weights Same weight / different heights

Nuclear and Radiological Engineering

Patient Dependent Phantoms – Pediatric Females

Nuclear and Radiological Engineering

Application of Patient Dependent Phantoms Organ Doses in Fluoroscopy

Patient‐dependent hybrid(a), (c), and (e)

Patient‐specific voxel(b), (d), and (f)

Males Females

Nuclear and Radiological Engineering

Application of Patient Dependent Phantoms Organ Doses in Fluoroscopy

Percent difference relative to patient‐specific voxel phantoms as an average for major organs within a RAO cardiac projection

Dose ImprovementThreshold~ 20‐30%

Nuclear and Radiological Engineering

Morphometric Categories –

Patient Sculpted Phantoms

Patient sculpted phantoms•Created by re‐shaping / re‐sizing the outer body contour (NURBS/PM) of a 

reference or patient‐dependent phantom to match the individual patient

•As a given body region is resized, 3D volumetric resizing of the

internal  organs is performed as well

Patient‐specific voxel Patient‐sculpted hybrid

ReferenceHybrid

Patient DependentHybrid

Nuclear and Radiological Engineering

Application of Patient Sculpted Phantoms Skin Doses in Fluoroscopy

Potential for real‐time dose 

monitoring with streaming 

RDSR DICOM files

Skin dose comparison between patient & anthropometrically matched hybrid patient dependent phantom (view is posterior).

Nuclear and Radiological Engineering

Continuum of Anatomic SpecificityPre‐computed dose library Patient‐specific dose calculation

Nuclear and Radiological Engineering

Current ICRP Model of Skeletal Dosimetry University of Leeds Studies 1960 ‐

1980

1907 - 1993

Nuclear and Radiological Engineering

Current ICRP Model of Skeletal Dosimetry University of Leeds Studies 1960 ‐

1980

Nuclear and Radiological Engineering

3D Image‐Based Skeletal Dosimetry

Nuclear and Radiological Engineering

PIRT Skeletal Model – 40 year male

Nuclear and Radiological Engineering

PIRT Skeletal Model – 40 year male

Homogeneous ICRP 70Skeletal Site Bone Volume (cm 3 ) Cortical Bone Spongiosa Med Cavity Trab Bone Marrow Shallow Marrow Cellularity

Craniofacial Bones 997.89 0.67 0.33 0.00Frontal / Facial 249.47 0.445 0.555 0.192 38%

Parietal 449.05 0.394 0.606 0.211 38%Occipital 299.37 0.912 0.088 0.061 38%

Mandible 48.49 0.40 0.60 0.00 0.089 0.911 0.078 38%Scapulae 488.91 0.35 0.65 0.00 0.153 0.847 0.100 38%Clavicles 68.60 0.40 0.60 0.00 0.116 0.884 0.080 33%Sternum 65.38 0.30 0.70 0.00 0.082 0.918 0.099 70%Ribs 310.80

Upper 85.66 0.40 0.60 0.00 0.103 0.897 0.072 70%Middle 135.25 0.40 0.60 0.00 0.142 0.858 0.109 70%Lower 89.89 0.40 0.60 0.00 0.101 0.899 0.109 70%

Cervical Vertebrae 96.67C3 (C1 ‐ C3) 42.43 0.35 0.65 0.00 0.159 0.841 0.155 70%C6 (C4 ‐ C7) 54.24 0.35 0.65 0.00 0.194 0.806 0.140 70%

Thoracic Vertebrae 309.42T3 (T1 ‐ T4) 81.18 0.25 0.75 0.00 0.132 0.868 0.111 70%T6 (T5 ‐ T8) 98.15 0.25 0.75 0.00 0.042 0.959 0.056 70%

T11 (T9 ‐ T12) 130.09 0.25 0.75 0.00 0.111 0.889 0.106 70%Lumbar Vertebrae 292.68

L2 (L1 ‐ L3) 169.47 0.20 0.80 0.00 0.114 0.887 0.112 70%L4 (L4 ‐ L5) 123.21 0.20 0.80 0.00 0.091 0.909 0.091 70%

Sacrum 197.17 0.25 0.75 0.00 0.118 0.882 0.111 70%Os coxae 923.96 0.25 0.75 0.00 0.100 0.901 0.112 48%

Homogeneous Bone Volume Fractions Spongiosa Volume Fractions

Nuclear and Radiological Engineering

PIRT Skeletal Model – 40 year male

Skeletal Site AM IM TBV CBV AM IM TM50 TBV CBV

Craniofacial Bones 54.84 85.13 341.48 1233.79 56.46 87.78 52.96 346.96 1253.59Mandible 10.08 15.65 4.76 35.79 10.38 16.13 2.27 4.83 36.36Scapulae 102.34 158.86 89.67 315.78 105.37 163.81 31.65 91.11 320.84Clavicles 12.02 23.21 8.78 50.64 12.37 23.94 3.28 8.92 51.45Sternum 29.42 12.00 6.92 36.20 30.29 12.37 4.61 7.03 36.78Ribs 115.00 46.89 41.06 229.41 118.41 48.35 18.70 41.72 233.09Cervical Vertebrae 36.14 14.74 20.72 62.44 37.21 15.20 9.34 21.05 63.44Thoracic Vertebrae 147.17 60.01 40.42 142.75 151.53 61.88 21.58 41.07 145.04Lumbar Vertebrae 146.92 59.91 44.91 108.02 209.81 61.77 24.56 45.63 109.75Sacrum 91.36 37.25 32.15 90.96 94.06 38.41 16.68 32.66 92.42Os coxae 299.65 308.86 127.24 426.26 308.52 318.47 77.74 129.28 433.10

Tissue Masses exclusive of MST (g) Tissue Masses inclusive of MST (g)

Similar data for extremities…

Totals 1170 2471 1159 4355 1263 2548 447 1177 44259414

ICRP 89 Values 1170 2480 1100 4400 ICRP 89 Value 9350Ratio 1.00 1.00 1.05 0.99 Ratio 1.01

Total Skeletal Mass

Nuclear and Radiological Engineering

PIRT Skeletal Model – 40 year male

Nuclear and Radiological Engineering

PIRT Skeletal Model – 40 year male

Nuclear and Radiological Engineering

Active Marrow Self‐Dose – PIRT vs CBIST

UF Adult

Hybrid Reference Phantom UF Newborn

Hybrid Reference Phantom

PIRT – Paired Image Radiation TransportCBIST –

Chord‐Based Infinite Spongiosa Transport

Nuclear and Radiological Engineering

Photon Skeletal DosimetryTwo methods of photon skeletal dose in phantoms:

• Dose response functionsscore energy‐dependent photon fluence in spongiosa regionsconvolve the photon fluence with the fluence‐to‐dose DRF

• Three‐factor methodscore energy deposition in homogenous spongiosa regionsscale that energy deposition by three terms as belowDRBM ~ Dspongiosa when latter  two factors approach unity

Nuclear and Radiological Engineering

Dose Enhancement Factors S(E)  to AM and TM50

Target ‐

Active Marrow (AM) Target – Total Shallow Marrow (TM50

)

Dose enhancement due to photoelectrons created in the bone trabeculae thatthen exit and irradiate the adjacent marrow tissues

Nuclear and Radiological Engineering

Concluding Remarks• With the develop of hybrid phantom technology and the construction of 

patient‐dependent phantom libraries, existing dosimetry software can be  extended away from its historical reliance on reference phantoms.

• Phantom assignment can thus be made based upon patient height / weight and  not only patient age.

• As image‐processing techniques become increasing automated, patient‐ sculpted phantoms and even patient‐specific phantoms with real‐time MC 

assessment of organ dose can move from the research realm into daily clinical  practice.

• Assessment of skeletal tissue dose is increasingly being refined

through micro‐ imaging and cadaver‐based reference models.

• Challenges for the future include adjustments of these models to include  patient‐specific changes in skeletal size, marrow cellularity, and bone 

microstructure.

Nuclear and Radiological Engineering

QuestionsQuestions……