Ant Colony Clustering for ROI Identification in Functional Magnetic Resonance Imaging
Algorithm 1
HAnt algorithm for ROI identification in fMRI.
(1)
Let be an anatomical label from the set . Let be the set, of cardinality , composed by fMRI signal instances with label . Each fMRI signal is arranged on a column vector of T time points
(2)
Let be the HRF generated according to equation (1), where is the length of the HRF
(3)
Let be the HRF convolution matrix of size
(4)
for all do
(5)
Initialize the parameters
(6)
Compute the UMAP grid using as inputs the neuronal stimulus estimated from fMRI signals. Assign the ants to this UMAP grid
(7)
Estimate the neuronal stimulus by deconvolution using [26, 27]
(8)
for to do
(9)
for to do
(10)
for all do
(11)
if , then
(12)
Compute and according to equations (2) and (3), respectively
(13)
Generate a random number
(14)
if , then
(15)
Pick up the neuronal signal
(16)
end if
(17)
else
(18)
if ant carrying neuronal signal and grid empty
(19)
Compute and according to equations (2) and (4), respectively
(20)
Generate a random number
(21)
if , then
(22)
Drop the neuronal signal
(23)
end if
(24)
end if
(25)
end if
(26)
Move to a randomly selected neighbouring grid position not occupied by other ant
