Algorithm 2. Computational implementation of the neurite growth algorithm.
1: Initialize: user chooses the number of neurites simulated (N) and, assuming the starting time is (t = 0), the duration (tend).
2: All N neurites at t = 0 are distributed randomly across the distal cross section of the NRC
3: while t < tend do
4:  Build the array LN of nodes (m) at a distance < 10± δ μm.
5:  Remove all the nodes that are occupied by neurites as well as the ones belonging to matrix $\mathrm{O}$, which are outside the geometry of the NRC.
6: if ɛ neurites detected the same node do
7:   Generate random variable X between 0 and 1 with each neurite having probability ${\displaystyle \frac{1}{\mathit{\epsilon }}}$ of being chosen.
8:   The first neurite that has $[0,\hspace{0.17em}1/\epsilon ,\hspace{0.17em}2/\epsilon \dots \mathit{\epsilon }/\mathit{\epsilon }]-\mathit{X}>0$ is chosen.
9:  end if
10:  Calculate θ between neurite consecutive segments.
11:  Get the value of the cues at the nodes detected $({\mathit{P}}_{{\mathit{w}}_{\mathit{i}}}(\mathit{t})$.)
12:  Normalize both $\mathit{M}({\mathit{\theta }}_{{\mathit{w}}_{\mathit{i}}})$ and ${\mathit{P}}_{{\mathit{w}}_{\mathit{i}}}(\mathit{t})$, respectively, and calculate F(wi) = M(θw) + Pw(t).
13:  Create the cumulative array ${\mathit{C}}_{\mathit{N}}=[\mathit{F}(1),\hspace{0.17em}\hspace{0.17em}F(1)+\mathit{F}(2),\hspace{0.17em}\hspace{0.17em}F(1)+\hspace{0.17em}\mathit{F}(2)+\mathit{F}(3),\hspace{0.17em}\hspace{0.17em}\dots \hspace{0.17em}F({\mathit{w}}_1)+\cdots +\mathit{F}(\vartheta )]$,
14:  Generate random variables $0\le {\mathit{X}}_{\mathit{N}}<{\sum }_{\mathit{i}=1}^{\mathit{m}}\mathit{F}(\mathit{w})$
15:  Find the node at γ which is the index of the first true value for TN = CN − X > 0.
16:  Perform step 14 and 15 NR times.
17:  The next position of the neurite is the node that has been most frequently chosen.
18:  t = t + τ, where τ is the time step considered.
19:   end while