Synapses, predictions, and prediction errors: a neocortical computational study of MDD using the temporal memory algorithm of HTM

AbstractBackgroundSynapses and spines are central in major depressive disorder (MDD) pathophysiology, recently highlighted by ketamine’s and psilocybin’s rapid antidepressant effects. The Bayesian brain and interoception perspectives formalize MDD as being “stuck” in affective states constantly predicting negative energy balance. We examined how synaptic atrophy relates to the predictive function of the neocortex and thus to symptoms, using temporal memory (TM), an unsupervised machine-learning algorithm. TM represents a single neocortical layer, learns in real-time using local Hebbian-learning rules, and extracts and predicts temporal sequences.MethodsWe trained a TM model on random sequences of upper-case alphabetical letters, representing sequences of affective states. To model depression, we progressively destroyed synapses in the TM model and examined how that affected the predictive capacity of the network.ResultsDestroying 50% of the synapses slightly reduced the number of predictions, followed by a marked drop with further destruction. However, reducing the synapses by 25% dropped the confidence in the predictions distinctly. So even though the network was making accurate predictions, the network was no longer confident about these predictions.ConclusionsThese findings explain how interoceptive cortices could be stuck in limited affective states with high prediction error. Growth of new synapses, e.g., with ketamine and psilocybin, would allow representing more futuristic predictions with higher confidence. To our knowledge, this is the first study to use the TM model to connect changes happening at synaptic levels to the Bayesian formulation of psychiatric symptomatology, making it possible to understand treatment mechanisms and possibly, develop new treatments.Graphical abstract