Two chronic, nonfluent aphasia patients participated in overt naming fMRI scans, pre-and post-a series of repetitive transcranial magnetic stimulation (rTMS) treatments as part of a TMS study to improve naming. Each patient received ten, 1-Hz rTMS treatments to suppress a part of R pars triangularis. P1 was a 'good responder' with improved naming and phrase length; P2 was a 'poor responder' without improved naming.Pre-TMS (10 yr. poststroke), P1 had significant activation in R and L sensorimotor cortex, R IFG, and in both L and R SMA during overt naming fMRI (28% pictures named. At 3 mo. post-TMS (42% named), P1 showed continued activation in R and L sensorimotor cortex, R IFG, and in R and L SMA. At 16 mo. post-TMS (58% named), he also showed significant activation in R and L sensorimotor cortex mouth and R IFG. He now showed a significant increase in activation in the L SMA compared to pre-TMS and at 3 mo. post-TMS (p<.02; p<.05, respectively). At 16 mo. there was also greater activation in L than R SMA (p<.08). At 46 mo. post-TMS (42% named), this new LH pattern of activation continued. He improved on the Boston Naming Test from 11 pictures named pre-TMS, to scores ranging from 14-18 pictures, post-TMS (2 mo. to 43 mo. post-TMS). His longest phrase length (Cookie Theft picture) improved from 3 words pre-TMS, to 5-6 words post-TMS.Pre-TMS (1.5 yr. poststroke), P2 had significant activation in R IFG (3% pictures named). At 3 and 6 mo. post-TMS, there was no longer significant activation in R IFG, but significant activation was present in R sensorimotor cortex. On all three fMRI scans, P2 had significant activation in both the L and R SMA. There was no new, lasting perilesional LH activation across sessions for this patient. Over time, there was little or no change in his activation. His naming remained only at 1-2 pictures during all three fMRI scans. His BNT score and longest phrase length remained at 1 word, post-TMS.Lesion site may play a role in each patient's fMRI activation pattern and response to TMS treatment. P2, the poor responder, had an atypical frontal lesion in the L motor and premotor cortex that extended Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain. The fMRI data of our patient who had good response following TMS support the notion that restoration of the LH language network is linked in part, to better recovery of naming and phrase length in nonfluent aphasia. NIH Public Access

The arcuate fasciculus (AF) is a white matter pathway traditionally considered to connect left Broca's area with posterior language zones. We utilized diffusion tensor imaging (DTI) in eight healthy subjects (5M) to track pathways in the horizontal mid-portion of the AF (hAF) to subregions of Broca's area -pars triangularis (PTr) and pars opercularis (POp); and to ventral premotor cortex (vPMC) in the right and left hemispheres (RH, LH). These pathways have previously been studied in the LH, but not in the RH. Only 1/8 subjects showed fiber tracts between PTr and hAF in the RH (also, only 1/8 in the LH). In contrast to PTr, 5/8 subjects showed fiber tracts between POp and hAF in the RH (8/8 in the LH). Fiber tracts for vPMC were similar to those of POp, where 7/8 subjects showed fiber tracts between vPMC and hAF in the RH (8/8 in the LH). Our designated hAF could have included some of the superior longitudinal fasciculus (SLF) III, because it is difficult to separate the two fiber bundles. The SLF III has been previously reported to connect supramarginal gyrus with POp and vPMC in the LH. Thus, although the present DTI study showed almost no pathways between PTr and hAF in the RH (and in the LH), robust pathways were observed between POp and/or vPMC with hAF in the RH (and in LH). These results replicate previous studies for the LH, but are new, for the RH. They could contribute to better understanding of recovery in aphasia.

The physical arrangement of receptive fields (RFs) within neural structures is important for local computations. Nonuniform distribution of tuning within populations of neurons can influence emergent tuning properties, causing bias in local processing. This issue was studied in the auditory system of barn owls. The owl’s external nucleus of the inferior colliculus (ICx) contains a map of auditory space where the frontal region is overrepresented. We measured spatiotemporal RFs of ICx neurons using spatial white noise. We found a population-wide bias in surround suppression such that suppression from frontal space was stronger. This asymmetry increased with laterality in spatial tuning. The bias could be explained by a model of lateral inhibition based on the overrepresentation of frontal space observed in ICx. The model predicted trends in surround suppression across ICx that matched the data. Thus, the uneven distribution of spatial tuning within the map could explain the topography of time-dependent tuning properties. This mechanism may have significant implications for the analysis of natural scenes by sensory systems.

Lesions of the basolateral amygdala (BLA) have long been associated with abnormalities of taste-related behaviors and with failure in a variety of taste-and odor-related learning paradigms, including taste-potentiated odor aversion, conditioned taste preference, and conditioned taste aversion. Still, the general role of the amygdala in chemosensory learning remains somewhat controversial. In particular, it has been suggested that the amygdala may not be involved in a form of chemosensory learning that has recently received a substantial amount of study-socially transmitted food preference (STFP). Here, we provide evidence for this involvement by pharmacologically inactivating the basolateral amygdala bilaterally during STFP training. The same inactivation sites that impaired taste aversion learning eliminated the normally conditioned preference for a food smelled on a conspecific's breath. Impairments of learned preference persisted even in testing sessions in which BLA was not inactivated, and learning was normal when the BLA was inactivated only during testing sessions; thus, the impairment was a true acquisition deficit. In conjunction with previous results from other paradigms, therefore, our data suggest that the amygdala is vital for learning procedures involving pairings of potent and arbitrary chemosensory stimuli.In rodents, the amygdala is one of the major recipients of gustatory and olfactory information. The basolateral amygdala (BLA) receives olfactory projections from the olfactory piriform cortex and gustatory projections from the parabrachial nuclei and insular cortex (Pare 2003). It is unsurprising, therefore, that singleunit and field potential electrophysiology studies of BLA in awake rats demonstrate responses to both gustatory (Yasoshima et al. 1995;Nishijo et al. 1998) and olfactory stimuli (Rosenkranz and Grace 2002;Sevelinges et al. 2004). These findings make the amygdala a probable neural substrate for sensory processes involving gustation and/or olfaction.Permanent lesion studies support this prediction, providing evidence that the BLA plays an important role in taste-and olfaction-related classical learning behaviors. Excitotoxic BLA lesions significantly alter taste palatability Jeganathan 1999, 2000;Touzani et al. 1997) and attenuate taste preference learning (Gilbert et al. 2003). BLA ablation also impairs learning of conditioned taste aversions (CTA) (Rollins et al. 2001;Mickley et al. 2004;Reilly 2005) and taste-potentiated odor aversions (Hatfield et al. 1992;Ferry et al. 1999;Ferry and Di Scala 2000;Sakai and Yamamoto 2001).Despite these findings, however, the involvement of BLA in basic chemosensory aversion and preference learning remains a topic of some controversy (Yamamoto et al. 1994;Reilly 2005). This is partially because the BLA dependence of CTA may vary as a function of exact experimental protocol; there is some evidence that the amygdala may be vital for the formation of aversions to tastes delivered to a passive rat (i.e., via intra-oral cannulae, or IOC), but not f...

Motion direction is a crucial cue for predicting future states in natural scenes. In the auditory system, the mechanisms that confer direction selectivity to neurons are not well understood. Neither is it known whether sound motion is encoded independently of stationary sound location. Here we investigated these questions in neurons of the owl's external nucleus of the inferior colliculus, where auditory space is represented in a map. Using a high-density speaker array, we show that the preferred direction and the degree of direction selectivity can be predicted by response adaptation to sounds moving over asymmetric spatial receptive fields. At the population level, we found that preference for sounds moving toward frontal space increased with eccentricity in spatial tuning. This distribution was consistent with larger receptive-field asymmetry in neurons tuned to more peripheral auditory space. A model of suppression based on spatiotemporal summation predicted the observations. Thus, response adaptation and receptive-field shape can explain direction selectivity to acoustic motion and an orderly distribution of preferred direction.

Bayesian models are often successful in describing perception and behavior, but the neural representation of probabilities remains in question. There are several distinct proposals for the neural representation of probabilities, but they have not been directly compared in an example system. Here we consider three models: a non-uniform population code where the stimulus-driven activity and distribution of preferred stimuli in the population represent a likelihood function and a prior, respectively; the sampling hypothesis which proposes that the stimulus-driven activity over time represents a posterior probability and that the spontaneous activity represents a prior; and the class of models which propose that a population of neurons represents a posterior probability in a distributed code. It has been shown that the non-uniform population code model matches the representation of auditory space generated in the owl's external nucleus of the inferior colliculus (ICx). However, the alternative models have not been tested, nor have the three models been directly compared in any system. Here we tested the three models in the owl's ICx. We found that spontaneous firing rate and the average stimulus-driven response of these neurons were not consistent with predictions of the sampling hypothesis. We also found that neural activity in ICx under varying levels of sensory noise did not reflect a posterior probability. On the other hand, the responses of ICx neurons were consistent with the non-uniform population code model. We further show that Bayesian inference can be implemented in the non uniform population code model using one spike per neuron when the population is large and is thus able to support the rapid inference that is necessary for sound localization.

Summary of Defense Advanced Research Projects Agency's (DARPA) explainable artificial intelligence (XAI) program from the program managers' and evaluator's perspective.

DARPA formulated the Explainable Artificial Intelligence (XAI) program in 2015 with the goal to enable end users to better understand, trust, and effectively manage artificially intelligent systems. In 2017, the four-year XAI research program began. Now, as XAI comes to an end in 2021, it is time to reflect on what succeeded, what failed, and what was learned. This article summarizes the goals, organization, and research progress of the XAI Program.