Research Articles: Behavioral/Cognitive Modulations of insular projections by prior belief mediate the precision of prediction error during tactile learning https://doi.org/10.1523/JNEUROSCI.2904-19.2020 Cite as: J. Neurosci 2020; 10.1523/JNEUROSCI.2904-19.2020 Received: 5 December 2019 Revised: 26 February 2020 Accepted: 27 February 2020 This Early Release article has been peer-reviewed and accepted, but has not been through the composition and copyediting processes. The final version may differ slightly in style or formatting and will contain links to any extended data. Alerts: Sign up at www.jneurosci.org/alerts to receive customized email alerts when the fully formatted version of this article is published. Copyright © 2020 Wang (##) et al. 1 Modulations of insular projections by prior belief mediate the 2 precision of prediction error during tactile learning 3 Bin A. Wang (⦻ᮼ)1,2, Lara Schlaffke1,2, Burkhard Pleger1-3 4 1Department of Neurology, BG University Hospital Bergmannsheil, Ruhr-University Bochum, 5 Bürkle-de-la-Camp Place 1, 44789 Bochum, Germany 2 6 Collaborative Research Centre 874 "Integration and Representation of Sensory Processes", Ruhr 7 University Bochum, Universitätsstraße 150, 44780 Bochum, Germany 8 3Department of Neurology, Max Planck Institute for Human Cognitive and Brain Sciences, 9 Stephanstr. 1a, 04103 Leipzig, Germany Abbreviated title: Insular projections encode precision of prediction Correspondence to: Burkhard Pleger, MD Department of Neurology, BG University Hospital Bergmannsheil Ruhr-University Bochum Bürkle-de-la-Camp Place 1, 44789, Bochum, Germany Phone: +49 (0) 234-302-3551 Fax: +49 (0) 234-302-6888 Email: [email protected] & Bin A. Wang, PhD Department of Neurology, BG University Hospital Bergmannsheil Ruhr-University Bochum Bürkle-de-la-Camp Place 1, 44789, Bochum, Germany Phone: +49 (0) 234-302-3857 Email: [email protected] 10 1 11 Number of pages: 35 12 Number of figures: 7 13 Number of tables: 1 14 Number of words for abstract: 151 15 Number of words for introduction: 558 16 Number of words for discussion: 1301 17 18 19 20 Acknowledgements: 21 This work was supported by the Deutsche Forschungsgemeinschaft (DFG, German 22 Research Foundation): Project number 122679504 - SFB 874 ‘Integration and 23 Representation of Sensory Processes’. 24 Conflict of Interest: 25 The authors declare no competing financial interests. 2 26 Abstract 27 Awareness for surprising sensory events is shaped by their prior belief inferred from past 28 experience. Here, we combined hierarchical Bayesian modeling with fMRI on an 29 associative learning task in 28 male human participants to characterize the effect of the 30 prior belief of tactile events on connections mediating the outcome of perceptual 31 decisions. Activity in anterior insula (AIC), premotor cortex (PMd) and inferior parietal 32 lobule (IPL) were modulated by prior belief on unexpected targets as compared to 33 expected targets. On expected targets, prior belief decreased the connection strength from 34 AIC to IPL, whereas it increased the connection strength from AIC to PMd when targets 35 were unexpected. Individual differences in the modulatory strength of prior belief on 36 insular projections correlated with the precision that increases the influence of prediction 37 errors on belief updating. These results suggest complementary effects of prior belief on 38 insular-frontoparietal projections mediating the precision of prediction during 39 probabilistic tactile learning. 40 41 Key words: hierarchical Bayesian modeling; dynamic causal modeling; functional 42 magnetic resonance imaging; tactile learning; prediction 43 3 44 Significance Statement 45 In a probabilistic environment, the prior belief of sensory events can be inferred from 46 past experiences. How this prior belief modulates effective brain connectivity for 47 updating expectations for future decision-making remains unexplored. Combining 48 hierarchical Bayesian modeling with fMRI, we show that during tactile associative 49 learning, prior expectations modulate connections originating in the anterior insula cortex 50 and targeting salience and attention related frontoparietal areas (i.e., parietal and 51 premotor cortex). These connections seem to be involved in updating evidence based on 52 the precision of ascending inputs to guide future decision-making. 53 4 54 Introduction 55 The expectation for the occurrence of tactile events is modulated by prior experiences 56 (Lovero et al. 2009; van Ede et al. 2010; Van Ede et al. 2014). In a probabilistic 57 environment, prior belief about the causes of relevant stimuli, such as imminent tactile 58 inputs, is continuously updated based on past experiences following the principle of 59 predictive coding (Rao and Ballard 1999) and free energy (Friston and Kiebel 2009). In 60 this framework, bottom-up sensations and top-down predictions are integrated by 61 precision-dependent computational processes, where the precision or confidence 62 increases the influence of ascending prediction error (PE) signals on perceptual inference 63 via post-synaptic gain (i.e., cortical gain control or excitation-inhibition balance) 64 (Feldman and Friston 2010). 65 Awareness of surprising or unexpected tactile sensations is thought to depend on 66 interactions within a hierarchically organized somatosensory system. Somatosensory 67 mismatch negativity, as a PE equivalent, elicited by an unexpected deviant has been 68 shown to follow Bayesian rules (Ostwald et al. 2012; Allen et al. 2016; Fardo et al. 2017). 69 These rules seem to be encoded by recurrent processing loops in a network encompassing 70 primary/secondary somatosensory cortices (S1/S2), inferior frontal cortex and anterior 71 cingulate cortex (Ostwald et al. 2012). Unexpected tactile events particularly increase 72 recurrent connectivity between S1/S2, inferior/middle frontal and inferior parietal regions 73 (Allen et al. 2016; Fardo et al. 2017), while the anterior insular cortex (AIC) seems 74 crucial for the integration of afferent sensory inputs with top-down control arising in the 75 frontal and cingulate cortex (Allen et al. 2016). 5 76 Behavioral performance and neural responses to expected and unexpected outcomes are 77 modulated by prior beliefs about future states of the tactile environment. The degree of 78 anticipation has been shown to improve the precision of tactile sensations (Van Ede et al. 79 2014), whereas uncertainty (Rossi-Pool et al. 2016; Schröder et al. 2019) or the 80 constitutive elements of perceptual inference, i.e., prediction error and precision 81 weighting (Ostwald et al. 2012; Allen et al. 2016; Fardo et al. 2017), are well reflected by 82 neural responses in the insular, cingulate, and premotor cortex. Despite all this evidence, 83 how prior belief regulates effective connectivity in the somatosensory network for future 84 decision-making remains unknown. 85 In an associative learning task, the association strength of the decision and the 86 corresponding feedback is continuously updated on the basis of recent observations. 87 Considerable evidence has shown that this inference can be conceptualized by 88 hierarchical Bayesian modeling (Iglesias et al. 2013; Vossel et al. 2015; Kuhns et al. 89 2017; Weilnhammer et al. 2018). Here, we applied an adapted tactile associative learning 90 task where participants had to learn the changing predictive strength of a sample stimulus 91 in forecasting a subsequently presented target. Trial-wise prior belief about the sample- 92 target contingency were used as modulators of effective connectivity for expected and 93 unexpected targets. 94 We hypothesized to identify a hierarchically organized somatosensory network, 95 consisting of the insular cortex and frontoparietal regions, in which predicted outcomes, 96 i.e., the awareness for expected and unexpected targets, are differentially modulated by 97 prior belief. Since the insular cortex is considered as a core hub regulating the interaction 98 of bodily, attentional, and anticipatory tactile signals (Sridharan et al. 2008; Lovero et al. 6 99 2009; Menon and Uddin 2010; Allen et al. 2016), we further expected that efferent 100 projections originating in the insula may change as a function of prior belief for expected 101 and unexpected targets, which, in turn may signal the precision weighting for belief 102 updating. 103 Materials and Methods 104 Participants 105 We recruited thirty-three healthy male participants (mean age ± SD: 25.1 ± 3.8 years). 106 Only male participants were included to avoid the influences of hormonal fluctuations 107 over the menstrual cycle on learning and associated blood-oxygen-level-dependent 108 (BOLD) signals (Dreher et al. 2007; Sacher et al. 2013; Wetherill et al. 2016). Five 109 participants were excluded due to more than 20% invalid trials (i.e., missed or late 110 response >1300ms) or less than 60% correct responses. Consequently, twenty-eight 111 participants were included for further date analysis (mean age ± SD: 25.3 ± 3.9 years). 112 All participants were right-handed as assessed by the Edinburgh Handedness Inventory 113 (Oldfield 1971) and had normal or corrected to normal vision, no history of 114 psychiatric/neurological disorders or regular medication. The study was approved by the 115 ethics committee of the Ruhr-University Bochum. 116 Tactile stimuli 117 The tactile stimuli were generated and delivered using an MRI-compatible piezoelectric 118 Braille stimulator (Metec, Stuttgart, Germany).