DR.fit.DoseVolume: Function for fitting Vdose and Dvolume to a given 'dvhmatrix'...

Description Usage Arguments Details References Examples

View source: R/dvh_mod.R

Description

This function can be used to find the best Vdose or Dvolume values that describe a given clinically observed outcome

Usage

1
2
DR.fit.DoseVolume(dvh, outcome, model, type = c("Vdose", "Dvolume"),
  CI = FALSE, CI.width = 0.95, epsilon = 1e-06)

Arguments

dvh

A dvhmatrix object

outcome

A vector of binary values (if not it will coerced to binary values) showing the observed outcome

model

The model that will be used to fit the dose-volume-response data (see examples). Models with available the logLik method can be used belonging to classes lm, glm and coxph

type

A character value representing the two type of dose-volume fitting as Vdose or Dvolume

CI

Logcal Value for calculating the confidence interval of the dose-volume fitting parameter and for the αβ if modLQ is optioned TRUE

CI.width

Width of confidence interval

epsilon

Error limit for calculating the Log Likelihood function in determining the confidence interval of dose-volume parameter

Details

The Vdose is the value of the volume of a structure in a dose volume histogram that receives a given level of dose. It is a widely used indicator and predictor of possible outcome when DVHs have to be clinically evaluated. There are many papers that exploit the correct use of Vdose for clinical evaluation of treatment plans. The DR.fit.DoseVolume function used inside the moddicom package allows to fit clinical data with dvhmatrix objects in order to detect the dose-volume relationship achievable by dosimetric data. The dose-volume-outcome fitting is performed by iterative search, so, as far as the author know, this is the first example of dose-volume-outcome fitting function that allows to find out the best fitting value of Vdose or Dvolume, rather than using an empyrical search from pre-fetermined values as usually reported in literature. When setting the model parameter the users have to chose between a glm model and a coxph model by writing the formula of the model without the dvhmatrix object

References

Special Considerations Regarding Absorbed-Dose and Dose-Volume Prescribing and Reporting in IMRT. J ICRU. 2010 Apr;10(1):27-40. doi: 10.1093/jicru/ndq008. PubMed PMID: 24173325.

Graham MV, Purdy JA, Emami B, Harms W, Bosch W, Lockett MA, Perez CA. Clinical dose-volume histogram analysis for pneumonitis after 3D treatment for non-small cell lung cancer (NSCLC). Int J Radiat Oncol Biol Phys. 1999 Sep 1;45(2):323-9. PubMed PMID: 10487552.

Examples

1
## NOT RUN DR.fit.DoseVolume(dvh = D, outcome = Dout, model = glm(formula = Dout ~ 1 , family = binomial(link = 'logit')), type = 'Vdose')

kbolab/moddicom documentation built on Nov. 29, 2020, 9:11 p.m.