Enter doses read from the target's dose-volume histogram. Each homogeneity index appears as soon as the doses it needs are filled in, so you can see how the definitions compare for the same plan.
Record which definition you used. The same plan gives different numbers under each, and they improve in different directions.
ICRU Report 83 divides the spread between the near-maximum and near-minimum doses by the median; a related version divides by the prescription dose instead (Kataria 2012). The D5% ÷ D95% ratio is the one the NRG brain SRS template calls the heterogeneity index. The RTOG index (Shaw 1993) divides the maximum isodose in the target by the reference (prescription) isodose.
ICRU 83 uses D2% and D98% rather than the true maximum and minimum because single-point doses depend on the calculation grid. See Homogeneity Index for when uniformity is wanted and when it is not.
Proposes a new homogeneity index. Its introduction lists the indices in use — Dmax/Dp, D5/D95, (D2 − D98)/Dp and the ICRU Report 83 version (D2 − D98)/D50 — notes that lower values of the last two mean a more homogeneous dose, and that each is based on only two or three points of the DVH. Full text read in Europe PMC (PMC6839365).
Retrospective study of 99 plans comparing five homogeneity index formulas. Its introduction summarizes the RTOG index (maximum isodose in the target ÷ reference isodose), D5%/D95%, and (D2% − D98%) ÷ prescribed dose, and explains that D2% and D98% are used because true point maximum and minimum doses are sensitive to calculation grid size and placement. Full text read in Europe PMC (PMC3532749).
The NRG Oncology template for the radiation section of a brain radiosurgery trial. Its example text asks for CT slices of 1.25 mm or less, strongly encourages MRI (contrast T1 and a 3D gradient-echo sequence) for target delineation, recommends fusing FDG-PET/CT when available, and asks that the fusion be done by a medical physicist and reviewed by the attending physician. Its plan-evaluation section defines the conformity index (CI, or R100%) as the prescription isodose volume ÷ PTV volume, R50% as the 50% prescription isodose volume ÷ target volume ("R50% defines dose falloff outside the target"), a heterogeneity index D5% ÷ D95%, and coverage as the PTV volume receiving the prescription dose ÷ PTV volume, and notes that criteria such as the CI should account for target size and treatment technology.
The template NRG Oncology gives investigators for writing the radiation section of a lung SBRT trial. Its imaging section allows PET/CT or MRI for structure definition, asks that the images be registered accurately to the planning scan (following AAPM TG-132 for deformable registration), notes that PET may be needed to separate tumor from atelectasis or fibrosis, and suggests an extra margin when a fused image is the primary image for the GTV. Its target and planning sections (5.2.4 and 5.2.6–5.2.8) define the plan-quality measures for lung SBRT trials: the prescription as the dose covering 95% of the PTV, a D99% requirement, R100% (prescription isodose volume ÷ PTV), R50% (50% prescription isodose volume ÷ PTV) and D2cm (maximum dose at 2 cm from the PTV in any direction, as a percentage of the prescription), with a table of allowable dose spillage by PTV volume. It describes a 20 mm PTV expansion structure used to control intermediate dose spillage, asks that dose above 105% of prescription be avoided outside the PTV, describes a deliberately heterogeneous dose with the hotspot in the central GTV, defines the maximum dose on a 0.03 cc volume, and ranks planning priorities when target coverage and organ limits conflict.
ICRU report on prescribing, recording and reporting intensity-modulated photon therapy. DoseLab has not read the report itself (it is sold by the ICRU). Its homogeneity index, (D2% − D98%) / D50%, is checked as quoted in the open-access Yan 2019 paper.
RTOG task-force guidelines for radiosurgery quality assurance, setting out a standard data set and minor and major protocol deviations. The paper itself is paywalled; DoseLab checked its conformity, coverage and homogeneity definitions as quoted in Kataria 2012 and Menon 2018. Citation checked against Crossref.