You are on page 1of 3

1

Computational model of the rod pathway in


human retina
Salil Pendse


Abstract—the rod pathway in the human retina is active under
low light conditions. The model studies signal processing carried
out by the first three layers of this path. It is observed that these
layers primarily pool data to obtain a substantial signal that can
be transmitted to the primary visual areas via the optic nerves.

Keywords— Retina, modeling, MATLAB, amacrine, bipolar,


rod.

1. INTRODUCTION

T HE retina is the first step of visual processing in the human


visual system. It consists of photoreceptors which convert
photons to electric current. The retina is complex and is made
up of six different categories of cells each performing a Fig. 1: Basic module of the model with 6 rod cells (blue), 3 bipolar cells
separate function. (red), one amacrine cell (gray). The model is made up of three such basic
The model I have developed simulates the dim light rod modules.
pathway in humans. Rod cells are active in low light
2.1 PHOTORECEPTOR LAYER
conditions when cones cannot generate any significant
The photoreceptor layer consists of 24 rod cells. These cells
response. The rod pathway synapses on to the cone pathway to
are exposed to light and hyperpolarize in response to
carry its signals to the optic nerve.
increasing light intensity. Table 1 provides the parameters used
in the model. The photoreceptor model has 6 different
2. RETINA MODEL
conductance and calcium dynamics. The equation of the model
The model was developed in MATLAB and constituted is given as:
three different types of cells. Each cell was considered to be
single compartment cylindrical model. The dimensions of the
cell were taken in to consideration for purposes of maintaining
dV /dt=(−Ih−Ikx−Ikv−Ica−IKca−IClca+Iphoto)/Cm
dimension equality. The model parameters were decided using
previous studies [1, 2] as reference. 2.1.1 PHOTO-CURRENT
The model consists of three layers of cells as shown in Fig.1 Modeling the entire photo transduction is computationally
each layer is connected to the next by a chemical synapse. All very expensive. The photo transduction mechanism was hence
the synapses are excitatory glutametergic synapses. The simply modeled as a photo current. Earlier research [1, 2] has
photoreceptor cells are exposed to flashes of light lasting 5 shown that the photocurrent equation used simulates the actual
seconds each. The response of the amacrine cells to these light current flow due to photo transduction.
flashes is depolarization relative to average intensity of light
pooled by the layers above it. The model is modular and
additional modules can be added to study larger light patterns. Iphoto=Idark+A[(1−e−t/T1)−(1/e−(t−b)/T2)+(1−e−t/T3)]
Where:
Idark is the dark current with a value of 40 pA.
A is a 5 step current with amplitude equal to intensity of
incident light.
T1= 50ms; T2 = 450ms; T3= 800ms; b=3800ms
2

effective synaptic input to the bipolar cell is a weighted sum of


all the synapses. Higher synaptic weights are assigned to
synapses from rods responding to incident photons. This is
done to improve the SNR by reducing the effect of dark
current on the bipolar cell. In the network model however the
synaptic input is a simple sum of the currents due to individual
synapses.
The current equation for a rod bipolar cell is as follows:

dV /dt=(−Ih−Ikv−Ica−IKca−Ika−Isyn)/Cm
Where Isyn is the total synaptic input current.

The bipolar cell ideally responds to flashes of light by


Fig. 2: Photocurrent corresponding to 12 different light intensities for depolarizing. The membrane remains depolarized during the
duration of 5 seconds.
entire duration of flash and quickly returns to its resting
2.1.2 Hyperpolarization activated Current (Ih) membrane potential once the flash is switched off.
The hyperpolarization activated current is responsible for
shaping the Rod membrane voltage in response to light flashes. Current Maximum Conductance Reversal Potential
It is modeled as a simple two state channel with a reversal Type ( S/cm^2) (mv)
potential of -32mV. I 0.001989 -75
Kiva 0.001989 -58
2.1.3 Noninactivating K Current (Ikx) Ica 0.01094 +40
The slow dynamics of these channels sets the resting Inca 0.0845 -58
membrane voltage of the rods. The channel does not play a IA 0.2486 -58
significant role in the hyperpolarization observed to flashes of
Table 2: Parameters for Bipolar Cells
light.
2.3 AII Amacrine cells.
2.1.4 Delayed Rectifying K Current (Ikv) The AII amacrine cells form the last cell layer of this model.
Electrophysiological recordings show the presence of this They have a peculiar firing pattern. They spike in response to
type of current. The delayed rectifier is a voltage gated flash and remain hyperpolarized during the remainder of time.
channel. These cells predominantly express only voltage gated Na
and K channels. Thus these cells are modeled with only those
2.1.5 Calcium Current (Ica) two ionic currents.
The calcium channels expressed in the rod cells is an L-type The current equation for AII amacrine cells is as follows.
calcium channel. The calcium current drives two other calcium
dependent ionic channels.
dV /dt=(−Ina−Ik−Isyn)/Cm
Current Maximum Conductance Reversal Potential Current Maximum Conductance Reversal Potential
Type ( S/cm^2) (mv) Type ( S/cm^2) (mv)
Ih 0.0025 -32 Ina 0.0026 +40
IKx 0.00104 -74 Ik 0.00026 -77
Ikv 0.01 -80 Table 3: Parameters for Amacrine Cells
Ica 0.004 +40
Ikca 0.005 -80 2.4 Synapses
IClca 0.0013 -20 The model has chemical synapses between rod and bipolar
cells and between bipolar cells and amacrine cells. They are
Table 1: Parameters for Rod Cells
graded glutametergic ribbon syanapses. To avoid
computational cost of accurately modeled ribbon synapse, a
2.2 BIPOLAR CELL LAYER simpler graded synapse developed earlier [1, 2] has been used.
The Bipolar cell layer receives synaptic input from the The synapse follows the following equation:
photoreceptor layer above. The rod bipolar cells are not
specific and may pool input from 2 to 10 rod cells. The Isyn=Ḡ ∗s∗(V−Esyn)
number of input synapses on the bipolar cells depends on the
location of the cell in the retina. In biological situations, the
3

Time Maximum Conductance Reversal Potential


constant ( S/cm^2) (mv)
(msec) The synaptic release of the rod cell actually increases due to
10 0.0255 0 hyperpolarization. This leads to depolarization of the bipolar
cell.
Table 1: Parameters for Ribbon Synapse The inabilities of my model to mimic the bipolar cells lead
to no further results being generated.
3. METHOD AND STIMULUS
Each cell layer was first simulated for a period to 1 sec each 5. DISCUSSION
without any input stimulus. This allowed me to find out the The model behaves accurately for till the photoreceptor
rest conditions of all the neurons. These conditions were then layers. The rod cells hyperpolarize as seen in physiological
compared with experimental values [1] to check for accuracy. recordings. If the model were working as expected, we would
The stimulus paradigm consisted of a single flash of light have seen depolarization of the bipolar cells in response to
lasting for 5 seconds. 12 different light intensities were to light flashes. The AII amacrine cells finally form a gap
determine their effect on the model. Figure 3 shows the light junction type synapse on the cone bipolar cells to transmit Rod
intensity pattern to which the cells were exposed. information out to the ganglionic cells,

REFERENCES
[1] Publio R, Oliveira RF, Roque AC (2009)” A Computational Study on
the Role of Gap Junctions and Rod Ih Conductance in the Enhancement
of the Dynamic Range of the Retina”. PLoS ONE 4(9): e6970.
doi:10.1371/journal.pone.0006970W.-
[2] Publio R, Oliveira RF, Roque AC (2006)” A realistic model of rod
photoreceptor for use in large scale retinal simulations’.

Fig.3: Stimulus for the model. 24 Rod cells are arranged in a single layer and exposed
to the above intensities for 5 seconds.

4. RESULTS

The photoreceptor layer responded to flashes of light with


graded hyperpolarization for the duration of stimulus. As soon
as the stimulus was switched off, the cells rapidly returned to
their resting potential. The response of a single rod cell to all
the intensities of light is shown in Figure 4.

Fig. 4: Rod membrane potential in response to different light intensities.


Highest light intensity causes least hyperpolarization.

You might also like