Dropout in Neural Networks Simulates the Paradoxical Effects of Deep Brain Stimulation on Memory

Abstract

Neuromodulation techniques such as deep brain stimulation (DBS) are a promising treatment for memory-related disorders including anxiety, addiction, and dementia. However, the outcomes of such treatments appear to be somewhat paradoxical, in that these techniques can both disrupt and enhance memory even when applied to the same brain target. In this article, we hypothesize that disruption and enhancement of memory through neuromodulation can be explained by the dropout of engram nodes. We used a convolutional neural network (CNN) to classify handwritten digits and letters and applied dropout at different stages to simulate DBS effects on engrams. We showed that dropout applied during training improved the accuracy of prediction, whereas dropout applied during testing dramatically decreased the accuracy of prediction, which mimics enhancement and disruption of memory, respectively. We further showed that transfer learning of neural networks with dropout had increased the accuracy and rate of learning. Dropout during training provided a more robust "skeleton" network and, together with transfer learning, mimicked the effects of chronic DBS on memory. Overall, we showed that the dropout of engram nodes is a possible mechanism by which neuromodulation techniques such as DBS can both disrupt and enhance memory, providing a unique perspective on this paradox.

Prelimbic cortical stimulation disrupts fear memory consolidation through ventral hippocampal dopamine D2 receptors

“…Here, we employed a systematic approach to examine the temporal specificity of prelimbic cortex DBS administration and isolate the effects of DBS on the consolidation of memory by stimulating post acquisition. We also showed consistent results across multiple models of conditioned fear and in repeat experiments, which is important given the paradoxical ability of DBS to both enhance and disrupt memories (Tan, Du, et al, 2020;Tan, Fung, et al, 2020;Tan, Neoh, et al, 2020). Despite the limitations of using conditioned fear to model anxiety disorders (LeDoux, 2015), it is one of the most wellestablished and translatable models currently available (Ganella & Kim, 2014).…”

“…However, few studies have systematically investigated the effects of DBS on fear memory. We previously hypothesized that DBS would be able to disrupt memories through the disruption of the engram process (Tan, Du, et al, 2020;Tan, Fung, et al, 2020;Tan et al, 2019). In this study, we examined the effects of DBS on the prelimbic cortex, which is a structure that is considered to be an ideal target (Lim et al, 2011;Tan et al, 2019) for the disruption of fear memories, as it is implicated in the expression of learned but not innate fear (Corcoran & Quirk, 2007).…”

Prelimbic Cortical Stimulation Improves Spatial Memory Through Distinct Patterns of Hippocampal Gene Expression in Aged Rats

“…The stimulation parameters (chronic high-frequency stimulation) used in this study were based on our previous experiments [7]. We previously showed chronic DBS could enhance memory in middle-aged animals [7], and hypothesised that the effects of chronic DBS on memory circuitry would enhance rather than disrupt memories [13,26]. Furthermore, the use of high-frequency stimulation rather than lowfrequency stimulation was based on our previous data that suggested possible anxiogenic effects (decreased time spent in the centre zone of the OFT) with low-frequency ventromedial prefrontal cortex (vmPFC) stimulation [14].…”

“…Briefly, individual animals were placed in the middle of the open field (100 × 100 × 40 cm made of black plexiglass) under a low-light setting and allowed to explore for 300 s. The behaviour of each animal was video-recorded and the distance travelled and duration of time spent in the different zones was analysed using Anymaze 5.0. The MWM test was conducted on week 4 (days [22][23][24][25][26][27][28]. Animals (PrL DBS, n = 12; sham, n = 13) received 30 min of electrical stimulation (or sham stimulation) prior to MWM test during 8:00-14:00 h (Fig.…”

Therapeutic Potential of Human Stem Cell Implantation in Alzheimer’s Disease

“…Of particular interest is a single-case report in which electrical stimulation was used to treat a patient with morbid obesity, in which the electrical brain stimulation unexpectedly evoked autobiographical memory episodes in the patient [101]. In animal studies, experimental data showed that memory functions could be enhanced by stimulating the medial prefrontal cortex [102][103][104][105], entorhinal cortex, and perifornical region [106][107][108]. It has also been shown to induce antidepressant-like effects in animal studies [109,110].…”