Load packages

library(RCon3D)
library(plotly)

Find the images

img <- findIMG("/Data")

Image details

pwidth is the pixel width and zstep is the distance between the layers:

# In microns:
pwidth <- 0.75
zstep <- 0.25

Quantify pixels for each layer for each channel

myq <- quant(img,channels=c("xan","pan","mic","ste"),naming=list(Time=c("24h")))

The naming argument is optional but can be used to look through the names of the images and add corresponding variables Here it looks for “24h” in the image name, and makes a variable called Time. This is of course only useful when there are several images with different metadata: eg. Time=c(“12h”,“24h”).

myq2 <- myq
myq2$Layer <- myq2$Layer * zstep
myq2$Layer <- max(myq2$Layer) - myq2$Layer

p <- ggplot(data = myq2, aes(x = Layer, y = Count, colour = Channel, group = Channel)) +
  theme_classic() +
  geom_freqpoly(stat = "identity", position = position_dodge(width = 0), size = 1) +
  coord_flip() +
  xlab("microns from the bottom")
ggplotly(p)

fullq <- aggregate(Count ~ ., data = myq[, colnames(myq) != "Layer"], FUN = sum)

# In cubic microns
fullq$Biovolume <- fullq$Count * pwidth^2 * zstep

fullq

##                  Img Channel Time    Count Biovolume
## 1 FourSpecies24h_xan     xan  24h  3552673  499594.6
## 2 FourSpecies24h_pan     pan  24h 26685719 3752679.2
## 3 FourSpecies24h_mic     mic  24h  1291929  181677.5
## 4 FourSpecies24h_ste     ste  24h  7192105 1011389.8

Section the quantification

We can split the image in top, middle and bottom based on various criteria. Here we define the Top as all layers from the very top until 75 percent of the image is filled (pt = 0.75). The bottom is the bottom 20 layers (add.b = 20). The middle is the rest. Note that the top of the specimen is in the layers with the low numbers, hence layer.start = "Top". add.t is how many layers to add to the top, after pt calculation (e.g. pt = 0 and add.t = 10 will specify the top as the top 10 layers)

myq.split <- layer_split(myq, layer.start = "Top", side = 512, pt = 0.75, add.b = 20, add.t = 0)

myq.split$Layer <- myq.split$Layer * zstep
myq.split$Layer <- max(myq.split$Layer) - myq.split$Layer

p <- ggplot(data = myq.split, aes(x = Layer, y = Count, colour = Split, group = Split)) +
  theme_classic() +
  geom_freqpoly(stat = "identity", position = position_dodge(width = 0), size = 1) +
  coord_flip() +
  facet_grid(.~Channel) +
  xlab("microns from the bottom")
ggplotly(p)