Title: | Positron Emission Tomography Time-Activity Curve Analysis |
---|---|
Description: | To facilitate the analysis of positron emission tomography (PET) time activity curve (TAC) data, and to encourage open science and replicability, this package supports data loading and analysis of multiple TAC file formats. Functions are available to analyze loaded TAC data for individual participants or in batches. Major functionality includes weighted TAC merging by region of interest (ROI), calculating models including standardized uptake value ratio (SUVR) and distribution volume ratio (DVR, Logan et al. 1996 <doi:10.1097/00004647-199609000-00008>), basic plotting functions and calculation of cut-off values (Aizenstein et al. 2008 <doi:10.1001/archneur.65.11.1509>). Please see the walkthrough vignette for a detailed overview of 'tacmagic' functions. |
Authors: | Eric Brown [aut, cre] , Ariel Graff-Guerrero [dgs], Jon Clayden [rev] (<https://orcid.org/0000-0002-6608-0619>, Jon Clayden reviewed the package for ropensci, see <https://github.com/ropensci/software-review/issues/280>), Brandon Hurr [rev] (<https://orcid.org/0000-0003-2576-4544>, Brandon Hurr reviewed the package for ropensci, see <https://github.com/ropensci/software-review/issues/280>) |
Maintainer: | Eric Brown <[email protected]> |
License: | GPL-3 |
Version: | 0.3.0 |
Built: | 2024-11-27 03:54:35 UTC |
Source: | https://github.com/ropensci/tacmagic |
tac objects can be created from data.frame objects with 'as.tac()'. The time and activity units must be specified as arguments if not already set as attributes in the data.frame. The columns of the data frame are the regional time activity curves, with the column names the names of the ROIs.
as.tac(x, time_unit = NULL, activity_unit = NULL)
as.tac(x, time_unit = NULL, activity_unit = NULL)
x |
data.frame with start, end time and tac data |
time_unit |
NULL if in data.frame or set to "seconds" or "minutes" |
activity_unit |
NULL if in data.frame or set to "kBq/cc", "Bq/cc", "nCi/cc" |
If the time_unit and activity_unit attributes are already in the data.frame, they do not need to be set again, but otherwise they will need to be specified in the input parameters.
tac object
Other Loading functions: load_tac
,
load_voistat
, load_vol
manual <- data.frame(start=c(0:4), end=c(2:6), ROI1=c(10.1:14.2), ROI2=c(11:15)) manual_tac <- as.tac(manual, time_unit="minutes", activity_unit="kBq/cc")
manual <- data.frame(start=c(0:4), end=c(2:6), ROI1=c(10.1:14.2), ROI2=c(11:15)) manual_tac <- as.tac(manual, time_unit="minutes", activity_unit="kBq/cc")
For a vector of participant IDs and correspondingly named tac files, this loads the tac files. If roi_m = T, then can also merge ROIs into larger ROIs based on the optional parameters that follow.
batch_load(participants, dir = "", tac_file_suffix = ".tac", tac_format = "PMOD", roi_m = FALSE, PVC = NULL, vol_file_suffix = NULL, vol_format = NULL, merge = NULL, ROI_def = NULL, tracer_dose = NULL, dose_unit = NULL, weight_kg = NULL)
batch_load(participants, dir = "", tac_file_suffix = ".tac", tac_format = "PMOD", roi_m = FALSE, PVC = NULL, vol_file_suffix = NULL, vol_format = NULL, merge = NULL, ROI_def = NULL, tracer_dose = NULL, dose_unit = NULL, weight_kg = NULL)
participants |
A vector of participant IDs |
dir |
A directory and/or file name prefix for the tac/volume files |
tac_file_suffix |
How participant IDs corresponds to the TAC files |
tac_format |
Format of tac files provided: See load_tac() |
roi_m |
TRUE if you want to merge atomic ROIs into larger ROIs (and if not, the following parameters are not used) |
PVC |
For PVC, true where the data is stored as _C in same tac file |
vol_file_suffix |
How participant IDs correspond to volume files |
vol_format |
The file format that includes volumes: See load_vol() |
merge |
Passes value to tac_roi(); T to also incl. original atomic ROIs |
ROI_def |
Object that defines combined ROIs, see ROI_definitions.R |
tracer_dose |
optionally, a vector of tracer doses (in the same order as participants), for SUV |
dose_unit |
if tracer_dose is specified, note the unit (e.g "MBq") |
weight_kg |
optionally, a vector of participant weights in kg, for SUV |
See load_tac() for specifics.
A list of data.frames, each is a participant's TACs
Other Batch functions: batch_tm
,
batch_voistat
# For the working example, the participants are full filenames. participants <- c(system.file("extdata", "AD06.tac", package="tacmagic"), system.file("extdata", "AD07.tac", package="tacmagic"), system.file("extdata", "AD08.tac", package="tacmagic")) tacs <- batch_load(participants, tac_file_suffix="")
# For the working example, the participants are full filenames. participants <- c(system.file("extdata", "AD06.tac", package="tacmagic"), system.file("extdata", "AD07.tac", package="tacmagic"), system.file("extdata", "AD08.tac", package="tacmagic")) tacs <- batch_load(participants, tac_file_suffix="")
For a list of tac data (from load_batch) this calculates specified models and saves in a tidy data.frame. Current model options are "SUVR", "Logan".
batch_tm(all_tacs, models, custom_model = NULL, ...)
batch_tm(all_tacs, models, custom_model = NULL, ...)
all_tacs |
A list by participant, of tac data (load_batch()) |
models |
A vector of names of the models to calculate |
custom_model |
A function that can be run like other models (advanced) |
... |
The arguments that get passed to the specified models/custom model, many are required; please check with model desired. |
For further details about how the models are calculated, see the individual functions that they rely on. "SUVR" uses suvr(), "Logan" uses DVR_all_ref_Logan().
A table of SUVR values for the specified ROIs for all participants
Other Batch functions: batch_load
,
batch_voistat
participants <- c(system.file("extdata", "AD06.tac", package="tacmagic"), system.file("extdata", "AD07.tac", package="tacmagic"), system.file("extdata", "AD08.tac", package="tacmagic")) tacs <- batch_load(participants, tac_file_suffix="") # Keeps only the ROIs without partial-volume correction (PMOD convention) tacs <- lapply(tacs, split_pvc, FALSE) batch <- batch_tm(tacs, models=c("SUVR", "Logan"), ref="Cerebellum_r", SUVR_def=c(3000,3300,3600), k2prime=0.2, t_star=23)
participants <- c(system.file("extdata", "AD06.tac", package="tacmagic"), system.file("extdata", "AD07.tac", package="tacmagic"), system.file("extdata", "AD08.tac", package="tacmagic")) tacs <- batch_load(participants, tac_file_suffix="") # Keeps only the ROIs without partial-volume correction (PMOD convention) tacs <- lapply(tacs, split_pvc, FALSE) batch <- batch_tm(tacs, models=c("SUVR", "Logan"), ref="Cerebellum_r", SUVR_def=c(3000,3300,3600), k2prime=0.2, t_star=23)
For a vector of participant IDs and correspondingly named .voistat files, this extracts the value from the files for the specified ROIs. participants can also be a vector of filenames, in which case set dir="" and filesuffix="", as in the example.
batch_voistat(participants, ROI_def, dir = "", filesuffix = ".voistat", varname = "VALUE")
batch_voistat(participants, ROI_def, dir = "", filesuffix = ".voistat", varname = "VALUE")
participants |
A vector of participant IDs |
ROI_def |
Object that defines combined ROIs, see ROI_definitions.R |
dir |
Directory and/or filename prefix of the files |
filesuffix |
Optional filename characters between ID and ".voistat" |
varname |
The name of the variable being extracted, e.g. "SRTM" |
See load_voistat() for specifics.
A table of values for the specified ROIs for all participants
Other Batch functions: batch_load
,
batch_tm
participants <- c(system.file("extdata", "AD06_BPnd_BPnd_Logan.voistat", package="tacmagic"), system.file("extdata", "AD07_BPnd_BPnd_Logan.voistat", package="tacmagic"), system.file("extdata", "AD08_BPnd_BPnd_Logan.voistat", package="tacmagic")) batchtest <- batch_voistat(participants=participants, ROI_def=roi_ham_pib(), dir="", filesuffix="", varname="Logan")
participants <- c(system.file("extdata", "AD06_BPnd_BPnd_Logan.voistat", package="tacmagic"), system.file("extdata", "AD07_BPnd_BPnd_Logan.voistat", package="tacmagic"), system.file("extdata", "AD08_BPnd_BPnd_Logan.voistat", package="tacmagic")) batchtest <- batch_voistat(participants=participants, ROI_def=roi_ham_pib(), dir="", filesuffix="", varname="Logan")
Change the radioactivity units of a tac or numeric object to the specified desired units (e.g. Bq, kBq, MBq, nCi, uCi, mCi, Ci). For convenience, if the unit is per volume ("x/cc" or "x/mL"), the "/cc" part is ignored for the conversion.
change_units(x, to_unit, from_unit)
change_units(x, to_unit, from_unit)
x |
time-activity curve or numeric object |
to_unit |
the desired unit (e.g. "kBq") |
from_unit |
not used for tac object (it is in the tac object), but for numeric objects, must be specified (e.g. "nCi") |
the converted object, same type as x
f <- system.file("extdata", "AD06.tac", package="tacmagic") AD06_tac <- load_tac(f, format="PMOD") AD06_tac_nCicc <- change_units(AD06_tac, to_unit = "nCi/cc") change_units(5, to_unit = "kBq", from_unit = "nCi") change_units(0.185, to_unit = "nCi", from_unit = "kBq")
f <- system.file("extdata", "AD06.tac", package="tacmagic") AD06_tac <- load_tac(f, format="PMOD") AD06_tac_nCicc <- change_units(AD06_tac, to_unit = "nCi/cc") change_units(5, to_unit = "kBq", from_unit = "nCi") change_units(0.185, to_unit = "nCi", from_unit = "kBq")
See the reference below and the tacmagic walkthrough vignette. Aizenstein et al. (2008) proposed a standardized method of calculating Pittsburgh Compound B (PIB) cutoff values to classify participants as PIB+ or PIB-. They used the distribution volume ratio (DVR) from several ROIs associated with amyloid deposition. The steps are summarized below. cutoff_aiz() implements 1-3, returning cutoff values for each ROI. It can be used to dichotomize participants, with pos_anyroi().
cutoff_aiz(modelstats, ROIs)
cutoff_aiz(modelstats, ROIs)
modelstats |
SUVR or DVR data for group of participants from batch_tm() |
ROIs |
list of variables (ROIs) to use for cutoff detection |
1. Remove outliers from a group of cognitively normal individuals. An outlier is defined as having any ROI with DVR > upper inner fence of that ROI (= 3rd quartile + (1.5 * IQR). 2. Iterate step 1 as needed until there are no more outlying participants. 3. From this subset of the group with outliers removed, the cutoff value for each ROI is set as the upper inner fence. 4. For all participants, if there is any ROI above the cutoff for that region, then the participant is deemed to be PIB+.
Cutoff values for each ROI based on the above method
Aizenstein HJ, Nebes RD, Saxton JA, et al. 2008. Frequent amyloid deposition without significant cognitive impairment among the elderly. Arch Neurol 65: 1509-1517.
Other Cutoff functions: pos_anyroi
cutoff_aiz(fake_DVR, c("ROI1_DVR", "ROI2_DVR", "ROI3_DVR", "ROI4_DVR"))
cutoff_aiz(fake_DVR, c("ROI1_DVR", "ROI2_DVR", "ROI3_DVR", "ROI4_DVR"))
This calculates the DVR using the non-invasive reference Logan method for all TACs in a supplied tac file. It uses DVR_ref_Logan if a target ROI is specified, otherwise will calculate DVR for all ROIs with DVR_ref_all_Logan()
dvr(tac, model = "logan", target = NULL, ref, k2prime, t_star, error = 0.1, method = "trapz")
dvr(tac, model = "logan", target = NULL, ref, k2prime, t_star, error = 0.1, method = "trapz")
tac |
The time-activity curve data from load_tac() or tac_roi() |
model |
Only model currently available is "logan" |
target |
Optional - otherwise will calculate DVR for all regions |
ref |
Required – The reference region, e.g. "cerebellum" |
k2prime |
Required – A fixed value for k2' must be specified (e.g. 0.2) |
t_star |
Required – If 0, t* will be calculated using find_t_star() |
error |
For find_t_star() |
method |
Method of integration, "trapz" or "integrate" |
For other model parameters, directly call DVR_ref_Logan().
Data frame with calculated DVRs
Logan, J., Fowler, J. S., Volkow, N. D., Wang, G.-J., Ding, Y.-S., & Alexoff, D. L. (1996). Distribution Volume Ratios without Blood Sampling from Graphical Analysis of PET Data. Journal of Cerebral Blood Flow & Metabolism, 16(5), 834-840. https://doi.org/10.1097/00004647-199609000-00008
Other Logan plot functions: DVR_all_ref_Logan
,
DVR_ref_Logan
, plot.ref_Logan
f <- system.file("extdata", "AD06.tac", package="tacmagic") fv <- system.file("extdata", "AD06_TAC.voistat", package="tacmagic") AD06_tac <- load_tac(f, format="PMOD") AD06_volume <- load_vol(fv, format="voistat") AD06 <- tac_roi(tac=AD06_tac, volumes=AD06_volume, ROI_def=roi_ham_pib(), merge=FALSE, PVC=FALSE) AD06_DVRs <- dvr(AD06, ref="cerebellum", k2prime=0.2, t_star=23) AD06_DVR <- dvr(AD06, target="frontal", ref="cerebellum", k2prime=0.2, t_star=23)
f <- system.file("extdata", "AD06.tac", package="tacmagic") fv <- system.file("extdata", "AD06_TAC.voistat", package="tacmagic") AD06_tac <- load_tac(f, format="PMOD") AD06_volume <- load_vol(fv, format="voistat") AD06 <- tac_roi(tac=AD06_tac, volumes=AD06_volume, ROI_def=roi_ham_pib(), merge=FALSE, PVC=FALSE) AD06_DVRs <- dvr(AD06, ref="cerebellum", k2prime=0.2, t_star=23) AD06_DVR <- dvr(AD06, target="frontal", ref="cerebellum", k2prime=0.2, t_star=23)
This calculates the DVR using the non-invasive reference Logan method for all TACs in a supplied tac file. It uses DVR_ref_Logan.
DVR_all_ref_Logan(tac_data, ref, k2prime, t_star, error = 0.1, method = "trapz", ...)
DVR_all_ref_Logan(tac_data, ref, k2prime, t_star, error = 0.1, method = "trapz", ...)
tac_data |
The time-activity curve data from tac_roi() |
ref |
Required – The reference region, e.g. "cerebellum" |
k2prime |
Required – A fixed value for k2' must be specified (e.g. 0.2) |
t_star |
Required – If 0, t* will be calculated using find_t_star() |
error |
For find_t_star() |
method |
Method of integration, "trapz" or "integrate" |
... |
When called from tm_batch, unused parameters may be supplied |
Data frame with calculated DVRs for all ROIs
Logan, J., Fowler, J. S., Volkow, N. D., Wang, G.-J., Ding, Y.-S., & Alexoff, D. L. (1996). Distribution Volume Ratios without Blood Sampling from Graphical Analysis of PET Data. Journal of Cerebral Blood Flow & Metabolism, 16(5), 834-840. https://doi.org/10.1097/00004647-199609000-00008
Other Logan plot functions: DVR_ref_Logan
,
dvr
, plot.ref_Logan
f <- system.file("extdata", "AD06.tac", package="tacmagic") fv <- system.file("extdata", "AD06_TAC.voistat", package="tacmagic") AD06_tac <- load_tac(f, format="PMOD") AD06_volume <- load_vol(fv, format="voistat") AD06 <- tac_roi(tac=AD06_tac, volumes=AD06_volume, ROI_def=roi_ham_pib(), merge=FALSE, PVC=FALSE) AD06_DVR <- DVR_all_ref_Logan(AD06, ref="cerebellum", k2prime=0.2, t_star=23)
f <- system.file("extdata", "AD06.tac", package="tacmagic") fv <- system.file("extdata", "AD06_TAC.voistat", package="tacmagic") AD06_tac <- load_tac(f, format="PMOD") AD06_volume <- load_vol(fv, format="voistat") AD06 <- tac_roi(tac=AD06_tac, volumes=AD06_volume, ROI_def=roi_ham_pib(), merge=FALSE, PVC=FALSE) AD06_DVR <- DVR_all_ref_Logan(AD06, ref="cerebellum", k2prime=0.2, t_star=23)
This calculates the coefficient from the non-invasive Logan method, which is equal to DVR. Works for a single tac (target).
DVR_ref_Logan(tac_data, target, ref, k2prime, t_star, error = 0.1, method = "trapz")
DVR_ref_Logan(tac_data, target, ref, k2prime, t_star, error = 0.1, method = "trapz")
tac_data |
The time-activity curve data from tac_roi() |
target |
The name of the target ROI, e.g. "frontal" |
ref |
The reference region, e.g. "cerebellum" |
k2prime |
A fixed value for k2' must be specified (e.g. 0.2) |
t_star |
If 0, t* will be calculated using find_t_star() |
error |
For find_t_star() |
method |
Method of integration, "trapz" or "integrate" |
Data frame with calculate DVRs for all ROIs
Logan, J., Fowler, J. S., Volkow, N. D., Wang, G.-J., Ding, Y.-S., & Alexoff, D. L. (1996). Distribution Volume Ratios without Blood Sampling from Graphical Analysis of PET Data. Journal of Cerebral Blood Flow & Metabolism, 16(5), 834-840. https://doi.org/10.1097/00004647-199609000-00008
Other Logan plot functions: DVR_all_ref_Logan
,
dvr
, plot.ref_Logan
f <- system.file("extdata", "AD06.tac", package="tacmagic") fv <- system.file("extdata", "AD06_TAC.voistat", package="tacmagic") AD06_tac <- load_tac(f, format="PMOD") AD06_volume <- load_vol(fv, format="voistat") AD06 <- tac_roi(tac=AD06_tac, volumes=AD06_volume, ROI_def=roi_ham_pib(), merge=FALSE, PVC=FALSE) AD06_DVR_fr <- DVR_ref_Logan(AD06, target="frontal", ref="cerebellum", k2prime=0.2, t_star=0)
f <- system.file("extdata", "AD06.tac", package="tacmagic") fv <- system.file("extdata", "AD06_TAC.voistat", package="tacmagic") AD06_tac <- load_tac(f, format="PMOD") AD06_volume <- load_vol(fv, format="voistat") AD06 <- tac_roi(tac=AD06_tac, volumes=AD06_volume, ROI_def=roi_ham_pib(), merge=FALSE, PVC=FALSE) AD06_DVR_fr <- DVR_ref_Logan(AD06, target="frontal", ref="cerebellum", k2prime=0.2, t_star=0)
A fake dataset of 50 simulated participants in the format that the function
tm_batch() would be expected to produce with the "Logan" model specified.
The data itself was generated as follows:
#higher <- matrix(rnorm(40, 1.9, 0.6), ncol=4, nrow=10)
#lower <- matrix(rnorm(160, 1.3, 0.3), ncol=4, nrow=40)
#fake_data <- as.data.frame(rbind(higher, lower))
#row.names(fake_data) <- paste0("p", 1:50)
#colnames(fake_data) <- c("ROI1_DVR", "ROI2_DVR", "ROI3_DVR", "ROI4_DVR")
#save(fake_data, "fake_DVR.Rda")
fake_DVR
fake_DVR
A data frame with 50 rows and 4 variables representing ROIs
This is the main function for loading an individual participant's TAC data. The minimal required information within the supplied files is the start and stop times and a time unit (either seconds or minutes), as well as the activity values for 1 or more ROIs, and units for activity. The currently supported formats (with the corresponding format argument), include:
"PMOD": PMOD .tac files
"voistat": PMOD TAC .voistat files used in combination with PMOD .acqtimes file for start/stop times.
"magia": magia pipeline .mat tac file
"DFT": Turku PET Centre's DFT format
load_tac(filename, format = "PMOD", acqtimes = NULL, time_unit = NULL, activity_unit = NULL)
load_tac(filename, format = "PMOD", acqtimes = NULL, time_unit = NULL, activity_unit = NULL)
filename |
(e.g. "participant01.tac") |
format |
A character string, with options listed above (e.g. "PMOD") |
acqtimes |
Filename for a .acqtimes file (as in PMOD), required for format="voistat" |
time_unit |
NULL if in file (e.g. PMOD .tac), or set to "seconds" or "minutes" if not in file or to override file |
activity_unit |
NULL if in file (e.g. PMOD .tac), or set to "kBq/cc", "Bq/cc", "nCi/cc" |
tac object
Other Loading functions: as.tac
,
load_voistat
, load_vol
f_raw_tac <- system.file("extdata", "AD06.tac", package="tacmagic") tac <- load_tac(f_raw_tac)
f_raw_tac <- system.file("extdata", "AD06.tac", package="tacmagic") tac <- load_tac(f_raw_tac)
PMOD can produce .voistat files with the average model values by ROI for its voxelwise binding potential (BPnd) models, such as Logan, SRTM, etc. This function reads the .voistat file and returns a data.frame with the ROI as rows and the model value as the column. Optionally, the ROIs can be combined into larger ROIs if ROI_def is specified, just as with TAC loading.
load_voistat(filename, ROI_def = NULL, model = "VALUE")
load_voistat(filename, ROI_def = NULL, model = "VALUE")
filename |
(e.g. participant_logan.voistat) |
ROI_def |
Optional ROI definitions to combine ROIs (e.g. roi_ham_pib()) |
model |
A string to name the variable being extracted, e.g. "Logan_DVR" |
data.frame with loaded model data in specified combined weighted ROIs
Other Loading functions: as.tac
,
load_tac
, load_vol
f <- system.file("extdata", "AD06_BPnd_BPnd_Logan.voistat", package="tacmagic") vs <- load_voistat(f, ROI_def=roi_ham_pib(), model="Logan")
f <- system.file("extdata", "AD06_BPnd_BPnd_Logan.voistat", package="tacmagic") vs <- load_voistat(f, ROI_def=roi_ham_pib(), model="Logan")
Loads ROI volumes from file for use by other functions
load_vol(filename, format = "voistat")
load_vol(filename, format = "voistat")
filename |
(e.g. participant.voistat) |
format |
(default is the TAC .voistat format from PMOD, also accepts "DFT and "BPndPaste") |
data.frame with loaded TAC data
Other Loading functions: as.tac
,
load_tac
, load_voistat
f_raw_vol <- system.file("extdata", "AD06_TAC.voistat", package="tacmagic") vol <- load_vol(f_raw_vol)
f_raw_vol <- system.file("extdata", "AD06_TAC.voistat", package="tacmagic") vol <- load_vol(f_raw_vol)
This plots the non-invasive Logan plot.
## S3 method for class 'ref_Logan' plot(x, ...)
## S3 method for class 'ref_Logan' plot(x, ...)
x |
Reference Logan model data object from DVR_ref_Logan() |
... |
Additional parameters than can be passed to plotting function |
No return f <- system.file("extdata", "AD06.tac", package="tacmagic") fv <- system.file("extdata", "AD06_TAC.voistat", package="tacmagic") AD06_tac <- load_tac(f, format="PMOD") AD06_volume <- load_vol(fv, format="voistat") AD06 <- tac_roi(tac=AD06_tac, volumes=AD06_volume, ROI_def=roi_ham_pib(), merge=FALSE, PVC=FALSE) AD06_DVR_fr <- DVR_ref_Logan(AD06, target="frontal", ref="cerebellum", k2prime=0.2, t_star=0) plot(AD06_DVR_fr)
Other Logan plot functions: DVR_all_ref_Logan
,
DVR_ref_Logan
, dvr
Plots time activity curves from 1 or 2 participants or groups.
## S3 method for class 'tac' plot(x, tac2 = NULL, ROIs, ymax = 25, time = "minutes", title = "", colors = rainbow, ...)
## S3 method for class 'tac' plot(x, tac2 = NULL, ROIs, ymax = 25, time = "minutes", title = "", colors = rainbow, ...)
x |
A tac object containing time-activity curves to plot, e.g. from tac_roi() or load_tac() |
tac2 |
An optional, second TAC, to plot for comparison |
ROIs |
A vector of ROIs to plot, names matching the TAC headers |
ymax |
The maximum value on the y-axis |
time |
"seconds" or "minutes" depending on desired x-axis, converts tac |
title |
A title for the plot |
colors |
If null, rainbow palette is used, otherwise another palette can be specified (heat.colors, terrain.colors, topo.colors, cm.colors |
... |
Additional arguments |
Creates a plot
Other tac functions: save_tac
,
split_pvc
, tac_roi
# f_raw_tac and f_raw_vol are the filenames of PMOD-generated files f_raw_tac <- system.file("extdata", "AD06.tac", package="tacmagic") f_raw_vol <- system.file("extdata", "AD06_TAC.voistat", package="tacmagic") tac <- load_tac(f_raw_tac) vol <- load_vol(f_raw_vol) AD06_tac_nc <- tac_roi(tac, vol, roi_ham_full(), merge=FALSE, PVC=FALSE) plot(AD06_tac_nc, ROIs=c("frontal", "cerebellum"), title="Example Plot")
# f_raw_tac and f_raw_vol are the filenames of PMOD-generated files f_raw_tac <- system.file("extdata", "AD06.tac", package="tacmagic") f_raw_vol <- system.file("extdata", "AD06_TAC.voistat", package="tacmagic") tac <- load_tac(f_raw_tac) vol <- load_vol(f_raw_vol) AD06_tac_nc <- tac_roi(tac, vol, roi_ham_full(), merge=FALSE, PVC=FALSE) plot(AD06_tac_nc, ROIs=c("frontal", "cerebellum"), title="Example Plot")
Aizenstein et al. (2008) proposed a standardized method of calculating PIB+ cutoff values to classify participants as PIB+ or PIB-. They used the DVR from 7 ROIs associated with amyloid deposition. This function takes the ROI-based cutoff values, e.g. from cutoff_aiz(), and returns a table specifying which participants are positive, i.e. which have at least one ROI greater than the cutoff.
pos_anyroi(modelstats, cutoff)
pos_anyroi(modelstats, cutoff)
modelstats |
SUVR or DVR data for group of participants from batch_tm() |
cutoff |
cutoffs for ROIs as from cutoff_aiz() |
data.frame of participants and positive/negative status
Aizenstein HJ, Nebes RD, Saxton JA, et al. 2008. Frequent amyloid deposition without significant cognitive impairment among the elderly. Arch Neurol 65: 1509-1517.
Other Cutoff functions: cutoff_aiz
This includes the cortical regions of roi_ham_stand() but also other regions. It can be modified to suit the user's needs.
roi_ham_full()
roi_ham_full()
A list of lists, where each list is an ROI (e.g.) frontal lobe that specifies the atomic ROIs from the atlas that make it up.
Hammers, Alexander, Richard Allom, Matthias J. Koepp, Samantha L. Free, Ralph Myers, Louis Lemieux, Tejal N. Mitchell, David J. Brooks, and John S. Duncan. 2003. Three-dimensional Maximum Probability Atlas of the Human Brain, with Particular Reference to the Temporal Lobe. Human Brain Mapping 19 (4): 224-247. doi:10.1002/hbm.10123
Other ROI definitions: roi_ham_pib
,
roi_ham_stand
roi_ham_full()
roi_ham_full()
This includes the ROIs from roi_ham_full and also the PIB cortical composite ROI as defined in the PMOD documentation and as widely used in PIB studies. See PMOD Neuro Tool (PNEURO) (Version 4.0) documentation.
roi_ham_pib()
roi_ham_pib()
A list of lists, where each list is an ROI (e.g.) frontal lobe that specifies the atomic ROIs from the atlas that make it up.
Hammers, Alexander, Richard Allom, Matthias J. Koepp, Samantha L. Free, Ralph Myers, Louis Lemieux, Tejal N. Mitchell, David J. Brooks, and John S. Duncan. 2003. Three-dimensional Maximum Probability Atlas of the Human Brain, with Particular Reference to the Temporal Lobe. Human Brain Mapping 19 (4): 224-247. doi:10.1002/hbm.10123
Other ROI definitions: roi_ham_full
,
roi_ham_stand
roi_ham_pib()
roi_ham_pib()
Return a list of merged ROIs made up of the atomic ROIs in the Hammer's atlas.
roi_ham_stand()
roi_ham_stand()
A list of lists, where each list is an ROI (e.g.) frontal lobe that specifies the atomic ROIs from the atlas that make it up.
Hammers, Alexander, Richard Allom, Matthias J. Koepp, Samantha L. Free, Ralph Myers, Louis Lemieux, Tejal N. Mitchell, David J. Brooks, and John S. Duncan. 2003. Three-dimensional Maximum Probability Atlas of the Human Brain, with Particular Reference to the Temporal Lobe. Human Brain Mapping 19 (4): 224-247. doi:10.1002/hbm.10123
Other ROI definitions: roi_ham_full
,
roi_ham_pib
roi_ham_stand()
roi_ham_stand()
Saves a tac object, created by load_tac(), tac_roi() or manually, and saves it as a PMOD-formatted tac file. Using the .tac extension in the file name is recommended.
save_tac(tac, outfile)
save_tac(tac, outfile)
tac |
The time-activity curve data, e.g. from load_tac() or tac_roi() |
outfile |
The output filename |
Does not return an object, only saves a file
Other tac functions: plot.tac
,
split_pvc
, tac_roi
When partial volume correction (PVC) is used in PMOD, the saved tac files have ROIs with and without PVC. When loaded with load_tac()) it may be desirable to keep only either the PVC or non-PVC tacs. This returns a tac object that is a subset of the input tac object with only the PVC or non-PVC tacs. This relies on PMOD's convention of labelling tac columns with "_C".
split_pvc(tac, PVC = TRUE)
split_pvc(tac, PVC = TRUE)
tac |
The time-activity curve data from loading function (PMOD) |
PVC |
If TRUE, includes columns with "_C", if FALSE, ones without "_C" |
Time-activity curve object
Other tac functions: plot.tac
,
save_tac
, tac_roi
# f_raw_tac and f_raw_vol are the filenames of PMOD-generated files f_raw_tac <- system.file("extdata", "AD06.tac", package="tacmagic") tac <- load_tac(f_raw_tac) tac_pvc <- split_pvc(tac, TRUE) tac_nc <- split_pvc(tac, FALSE)
# f_raw_tac and f_raw_vol are the filenames of PMOD-generated files f_raw_tac <- system.file("extdata", "AD06.tac", package="tacmagic") tac <- load_tac(f_raw_tac) tac_pvc <- split_pvc(tac, TRUE) tac_nc <- split_pvc(tac, FALSE)
Calculate the standardized uptake value (SUV) from a tac object, the participant's weight, and the tracer dose. These values may be in the tac object or manually supplied. The weight must be in kg, and the tracer units must be specified. The dose is converted to MBq, the tac is converted to kBq/cc, and the final SUV units are thus in g/cc. Aside from the tac object, the remaining parameters should be left NULL if the required data is in the tac object attributes (as can be done with batch_load()).
suv(tac, SUV_def, dose = NULL, dose_unit = NULL, weight_kg = NULL, ...)
suv(tac, SUV_def, dose = NULL, dose_unit = NULL, weight_kg = NULL, ...)
tac |
time-activity curve object (decay-corrected) |
SUV_def |
vector of start times for window for SUV weighted average, or alternatively, "max" for the maximum ROI SUV value |
dose |
the injected tracer dose |
dose_unit |
unit of tracer dose (e.g. "MBq", "kBq", "mCi"...) |
weight_kg |
the participant's weight in kg |
... |
When called from tm_batch, unused parameters may be supplied |
table of SUV values
Other SUV functions: tac_suv
f <- system.file("extdata", "AD06.tac", package="tacmagic") fv <- system.file("extdata", "AD06_TAC.voistat", package="tacmagic") AD06_tac <- load_tac(f, format="PMOD") AD06_volume <- load_vol(fv, format="voistat") AD06 <- tac_roi(tac=AD06_tac, volumes=AD06_volume, ROI_def=roi_ham_pib(), merge=FALSE, PVC=FALSE) # dose and weight are fabricated for the example AD06_suvmax <- suv(AD06, "max", dose = 9.0, dose_unit = "mCi", weight_kg = 70) AD06_suv <- suv(AD06, c(3000, 3300, 3600), dose = 9.0, dose_unit = "mCi", weight_kg = 70)
f <- system.file("extdata", "AD06.tac", package="tacmagic") fv <- system.file("extdata", "AD06_TAC.voistat", package="tacmagic") AD06_tac <- load_tac(f, format="PMOD") AD06_volume <- load_vol(fv, format="voistat") AD06 <- tac_roi(tac=AD06_tac, volumes=AD06_volume, ROI_def=roi_ham_pib(), merge=FALSE, PVC=FALSE) # dose and weight are fabricated for the example AD06_suvmax <- suv(AD06, "max", dose = 9.0, dose_unit = "mCi", weight_kg = 70) AD06_suv <- suv(AD06, c(3000, 3300, 3600), dose = 9.0, dose_unit = "mCi", weight_kg = 70)
Calculate the standardized uptake value ratio (SUVR) for all ROIs in the provided tac data, using the specified reference region.
suvr(tac, SUVR_def, ref, ...)
suvr(tac, SUVR_def, ref, ...)
tac |
The time-activity curve data from tac_roi() |
SUVR_def |
a vector of start times for window to be used in SUVR |
ref |
a string, e.g. "cerebellum", to specify reference region |
... |
When called from tm_batch, unused parameters may be supplied |
A data.frame of SUVR values for the specified ROIs
Other SUVR functions: suvr_auc
f <- system.file("extdata", "AD06.tac", package="tacmagic") fv <- system.file("extdata", "AD06_TAC.voistat", package="tacmagic") AD06_tac <- load_tac(f, format="PMOD") AD06_volume <- load_vol(fv, format="voistat") AD06 <- tac_roi(tac=AD06_tac, volumes=AD06_volume, ROI_def=roi_ham_pib(), merge=FALSE, PVC=FALSE) AD06_SUVR <- suvr(AD06, SUVR_def=c(3000,3300,3600), ref="cerebellum")
f <- system.file("extdata", "AD06.tac", package="tacmagic") fv <- system.file("extdata", "AD06_TAC.voistat", package="tacmagic") AD06_tac <- load_tac(f, format="PMOD") AD06_volume <- load_vol(fv, format="voistat") AD06 <- tac_roi(tac=AD06_tac, volumes=AD06_volume, ROI_def=roi_ham_pib(), merge=FALSE, PVC=FALSE) AD06_SUVR <- suvr(AD06, SUVR_def=c(3000,3300,3600), ref="cerebellum")
Calculate the standardized uptake value ratio (SUVR) for all ROIs in the provided tac data, using the specified reference region. This is an alternate to suvr() which should provide very similar values.
suvr_auc(tac, SUVR_def, ref, ...)
suvr_auc(tac, SUVR_def, ref, ...)
tac |
The time-activity curve data from tac_roi() |
SUVR_def |
a vector of start times for window to be used in SUVR |
ref |
is a string, e.g. "cerebellum", to specify reference region |
... |
When called from tm_batch, unused parameters may be supplied |
A data.frame of SUVR values for the specified ROIs #' f <- system.file("extdata", "AD06.tac", package="tacmagic") fv <- system.file("extdata", "AD06_TAC.voistat", package="tacmagic") AD06_tac <- load_tac(f, format="PMOD") AD06_volume <- load_vol(fv, format="voistat") AD06 <- tac_roi(tac=AD06_tac, volumes=AD06_volume, ROI_def=roi_ham_pib(), merge=FALSE, PVC=FALSE)
AD06_SUVR <- suvr_auc(AD06, SUVR_def=c(3000,3300,3600), ref="cerebellum")
Other SUVR functions: suvr
Calculate weighted time-activity curves for specified regions of interest
tac_roi(tac, volumes, ROI_def, merge, PVC)
tac_roi(tac, volumes, ROI_def, merge, PVC)
tac |
The time-activity curve data from loading function |
volumes |
The ROI volume data from loading function |
ROI_def |
The definition of ROIs by combining smaller ROIs from TAC file |
merge |
If TRUE, includes the original ROIs in the output data |
PVC |
If TRUE, appends "_C" to ROI name header (as in PMOD TAC files) |
Time-activity curves for the specified ROIs
Other tac functions: plot.tac
,
save_tac
, split_pvc
# f_raw_tac and f_raw_vol are the filenames of PMOD-generated files f_raw_tac <- system.file("extdata", "AD06.tac", package="tacmagic") f_raw_vol <- system.file("extdata", "AD06_TAC.voistat", package="tacmagic") tac <- load_tac(f_raw_tac) vol <- load_vol(f_raw_vol) AD06_tac_nc <- tac_roi(tac, vol, roi_ham_full(), merge=FALSE, PVC=FALSE)
# f_raw_tac and f_raw_vol are the filenames of PMOD-generated files f_raw_tac <- system.file("extdata", "AD06.tac", package="tacmagic") f_raw_vol <- system.file("extdata", "AD06_TAC.voistat", package="tacmagic") tac <- load_tac(f_raw_tac) vol <- load_vol(f_raw_vol) AD06_tac_nc <- tac_roi(tac, vol, roi_ham_full(), merge=FALSE, PVC=FALSE)
Calculate the standardized uptake value (SUV) time-activity curve from a tac object, the participant's weight, and the tracer dose. The weight must be in kg, and the tracer dose must be specified. The dose is converted to MBq, the tac is converted to kBq/cc, and the final SUV units are thus in g/cc. Aside from the tac object, the remaining parameters should be left NULL if the required data is in the tac object attributes (as can be done with batch_load().
tac_suv(tac, dose = NULL, dose_unit = NULL, weight_kg = NULL)
tac_suv(tac, dose = NULL, dose_unit = NULL, weight_kg = NULL)
tac |
time-activity curve object (decay-corrected) |
dose |
the injected tracer dose |
dose_unit |
unit of tracer dose (e.g. "MBq", "kBq", "mCi"...) |
weight_kg |
the participant's weight in kg |
tac object with SUV values
Other SUV functions: suv
f <- system.file("extdata", "AD06.tac", package="tacmagic") fv <- system.file("extdata", "AD06_TAC.voistat", package="tacmagic") AD06_tac <- load_tac(f, format="PMOD") AD06_volume <- load_vol(fv, format="voistat") AD06 <- tac_roi(tac=AD06_tac, volumes=AD06_volume, ROI_def=roi_ham_pib(), merge=FALSE, PVC=FALSE) # dose and weight are fabricated for the example AD06_suv <- tac_suv(AD06, dose = 9.0, dose_unit = "mCi", weight_kg = 70)
f <- system.file("extdata", "AD06.tac", package="tacmagic") fv <- system.file("extdata", "AD06_TAC.voistat", package="tacmagic") AD06_tac <- load_tac(f, format="PMOD") AD06_volume <- load_vol(fv, format="voistat") AD06 <- tac_roi(tac=AD06_tac, volumes=AD06_volume, ROI_def=roi_ham_pib(), merge=FALSE, PVC=FALSE) # dose and weight are fabricated for the example AD06_suv <- tac_suv(AD06, dose = 9.0, dose_unit = "mCi", weight_kg = 70)
The main features of tacmagic are to load PET time activity curve (tac) data from multiple formats, merge ROIs weighted for volume, calculate binding potential models including SUVR and DVR, basic plotting, and calculation of cut-off values. Please see the walkthrough vignette for a detailed overview.