Building a modified N-Back paradigm for an fNIRS study

Hi, I am new to PsychPy and I was wondering whether someone could give me some advice. I have red through some threads with regards to the N-Back paradigm. However, my approach differs from an traditional N-Back for the reason that it will be part of my fNIRS study and will be later expanded with an network interface for Turbo-Satori (analysis software for fNIRS). My approach also includes a Visual Analog Scale between trials.

I am looking forward to create an N-back (0, 1, 2, 3, 4, 5) task in which subjects respond by choosing an answer (1, 2, 3, 4, 5 or None). Subjects are instructed to count the “matches”. A match is when the presented stimulus is identical to n stimulus presented immediately before. Therefore, subjects would respond by indicating the number of matches they have counted mentally on the keyboard rather pressing space bar.

Moreover, my approach includes 6 runs which subjects will go through in one session. 4 training runs have 18 trials in total, while 2 test runs include 12 trials. Each run includes all n (0,1,2,3,4,5) and will be randomized.

My question is how I should structure my flow with above mentioned runs. And how should I present my stream of letters? Is it possible to write the stream of letters in an Excel file and assign them with correctness values?

Thank you in advance and I wish you nice day!