一般口演
視覚 / 眼球運動
Vision / Eye Movements
座長：菅生-宮本 康子（産業技術総合研究所） |
|---|
| 2022年7月2日 14:00～14:15 沖縄コンベンションセンター 会議場A2 第7会場
3O07a1-01
Saccadic eye movements and visual saliency in pharmacological model of schizophrenia during free-viewing in common marmosets
*Zlata Polyakova(1), Masatoshi Yoshida(1)
1. Ctr for Hum Nat, AI and Neurosci, Hokkaido Univ, Sapporo, Japan
Keyword: Attention, Natural images, Saliency map, Ketamine
Various eye movement abnormalities and impairments in visual information processing have been reported in patients with schizophrenia. In a line with the aberrant salience hypothesis (Kapur, 2003), subjects with schizophrenia have a deficit of exploratory eye movements and tend to fix gazes to the salient positions during free-viewing. Therefore, dysfunction of saccadic eye movements and visual saliency is a potential biological marker for schizophrenia. Human data demonstrated the emergence of clinical symptoms of schizophrenia after the application of NMDA antagonists such as ketamine (Krystal et al., 1994). We examined whether injection of ketamine reproduces schizophrenia-like symptoms in common marmosets.
In five animals, eye movements during free-viewing under head-fixed condition were binocularly recorded at 500Hz with the EyeLink 1000 Plus eye-tracker. The natural and/or complex images identical to those used for the human subjects were randomly presented for 8 seconds. As a pharmacological model of schizophrenia, we injected a subanesthetic dose of ketamine (0.25, 0.5, or 1.0 mg/kg) intramuscularly for transient and reversible manipulation. The saline injections were performed for each animal and used as a control for further comparison. By recording gaze parameters, such as eye fixations (coordinates and times), saccades, and pupil size, eye movement patterns in control and modeled schizophrenic states were compared in the experiment.
Saccades were reliably detected in all animals before and after ketamine injection. The analysis of eye movements revealed that the mean saccade amplitudes were reduced and scanpath length decreased after injection of ketamine. The slope of the main sequence relationship between the peak velocities and the amplitudes of saccades was not altered. The visual salience of the presented images was calculated based on Itti’s model (Itti et al., 1998) and used for the evaluation of salience values at the gazes during free-viewing. We found that salience values averaged across all images and saccades were higher in the ketamine condition. These results suggest that ketamine affected some attentional processes but did not affect the motor activity of marmosets.
We conclude that subanesthetic dose of ketamine is a promising pharmacological model of schizophrenia in common marmosets which can be used in combination with free-viewing paradigms in order to establish “translatable markers” for schizophrenia in primates. | | 2022年7月2日 14:15～14:30 沖縄コンベンションセンター 会議場A2 第7会場
3O07a1-02
サル上側頭溝中間部における自己と他者の動作情報表現
Representation of social action and prediction error in the macaque mid-superior temporal sulcus region
*二宮 太平(1)、則武 厚(1,2)、磯田 昌岐(1,2)
1. 生理研認知行動発達、2. 総研大院生命科学生理
*Taihei Ninomiya(1), Atsushi Noritake(1,2), Masaki Isoda(1,2)
1. Div Behav Devel, NIPS, Aichi, Japan, 2. Dep Physiol Sci, Grad Univ Adv Stud (SOKENDAI), Hayama, Japan
Keyword: Macaque mid-STS, Mentalizing system, Predictive coding
Inferring the mental states of others, or mentalizing, is of crucial importance for successful social interactions. Human neuroimaging studies have revealed that this ability is implemented in diverse brain regions including the temporoparietal junction (TPJ). While accumulating evidence has pointed out the evolution and expansion of this cortical area in the human brain, its homologous area in nonhuman primates remains unclear. It has recently been shown that the middle part of the macaque superior temporal sulcus (mid-STS) region has an anatomical fingerprint most similar to the human TPJ. In the present study, we show that the mid-STS region is functionally similar to the human TPJ in aspects of a preference for live social interactions and in a theoretical framework of predictive coding. We trained macaque monkeys to perform a turn-taking choice task with a real monkey sitting face-to-face or a filmed monkey displayed on a monitor. Of 531 mid-STS neurons recorded, 328 neurons showed significantly modulated activities during a peri-action period. Approximately 75% of the task-related neurons exhibited agent-dependent activity, most responding selectively or preferentially to the partner's action. At the population level, activities of these partner-type neurons were significantly greater with the real partner than with the filmed partner. Furthermore, a subset of the partner-type neurons responded proactively when predictions about the partner's action were violated. This prediction error coding was specific to the action domain; almost none of the neurons signaled error in the prediction of reward. The present findings indicate that the macaque mid-STS region has unique roles in social cognitive processes and is possibly an evolutionary signature associated with the human TPJ. | | 2022年7月2日 14:30～14:45 沖縄コンベンションセンター 会議場A2 第7会場
3O07a1-03
顔選択的ニューロンにおける積計算
Multiplicative computation in face-selective neurons
*細谷 晴夫(1)、Minggui Chen(2)、Winrich Freiwald(2)
1. 国際電気通信基礎技術研究所、2. ロックフェラー大学
*Haruo Hosoya(1), Minggui Chen(2), Winrich Freiwald(2)
1. Advanced Telecommunication Research Institute International, 2. Rockefeller University
Keyword: Inferotemporal cortex, face-processing system, model-based data analysis, primate vision
Previous research has identified multiple patches of cortex specialized to face processing. Neurons in this system provide information both about faces as a category and face identity. How they accomplish this, is currently unknown. A recently proposed theory, called mixture of sparse coding models, predicts that a crucial computational component of those face-selective neurons may be multiplication between two units: facial feature detection (eye, mouth, etc.) and facial category detection (face or non-face). Such an operation is absent from other general models of information processing like sparse coding or deep learning models. It is thus crucial to test this prediction in order to make strides towards identifying the computational principles in the face-processing system. In this study, to test this prediction, we conducted electrophysiological recordings while presenting a large number of face images added with Gabor-shape noise. We estimated, in a data-driven manner, a form of multiplicative model, which starts with a simple basis model mimicking the early visual processing and then performs a multiplicative operation on the outputs of two parametrized linear filters with different nonlinearities. In a population of neurons in the anterior-medial face patch AM, we found significantly better prediction performance of this model for held-out test data, than a non-multiplicative single filter model. Moreover, for those neurons, one filter often had a pattern combining facial parts with the outline, whereas the other depended on the presence of the face as shown by strong suppression when the input face image was heavily corrupted by Gabor-shape noise. These results are consistent with the theoretical prediction of a multiplicative computation between a facial feature detector and a category detector, raising a new possibility for a general computational mechanism in the primate higher visual cortex. | | 2022年7月2日 14:45～15:00 沖縄コンベンションセンター 会議場A2 第7会場
3O07a1-04
高次視覚野の機能的な発達における一次視覚野及び高次視床核からの入力が果たす役割
The role of bottom-up inputs from the primary visual cortex and higher visual thalamus in functional development of higher visual areas
*加藤 利樹(1)、上村 允人(4)、村上 知成(1,2,3)、大木 研一(1,2,3)
1. 東京大学大学院医学系研究科機能生物学専攻統合生理学教室、2. ニューロインテリジェンス国際研究機構、3. Beyond AI研究所、4. 関西医科大学生物学教室
*Toshiki Kato(1), Masato Uemura(4), Tomonari Murakami(1,2,3), Kenichi Ohki(1,2,3)
1. Department of Physiology, Graduate School of Medicine, The University of Tokyo, 2. International Research Center for Neuroscience (WPI-IRCN), 3. Institute for AI and Beyond, The University of Tokyo, 4. Department of Biology, Kansai Medical University
Keyword: Visual, Function, Development, Axon
In the mouse visual cortex, several higher visual areas (HVAs) process different visual information such as different spatial and temporal frequencies. They receive functionally specific inputs from the primary visual cortex (V1) and higher visual thalamus (Lateral Posterior Nucleus, LPN) in adults (Glickfeld et al., 2013; Matsui et al., 2013; Blot et al., 2020). We previously reported that at eye opening (EO: P14), HVAs already show functional differences although in a lesser extent than in adults (Murakami et al., 2017). Anatomically, HVAs already receive axonal projections from V1 and LPN at EO. However, their functional role in the functional differences among HVAs at EO is unclear.
Here, we examined spatiotemporal frequency selectivity of the axon terminals projecting from V1 and LPN to HVAs and their role in functional differences among HVAs at EO. We found that, at EO, V1 axons projecting to any HVA were functionally similar to each other and responded preferentially to low spatial and low temporal frequencies using in vivo wide-field and two-photon calcium imaging. The selectivity of V1 axons did not necessarily match the selectivity of neurons in each target HVA. On the other hand, we found that LPN axons projecting to HVAs were functionally different and matched the selectivity of neurons in each target HVA. In addition, the response amplitude and properties of HVAs were not changed by V1 inactivation at EO. These results suggest that V1-HVAs pathway is not critical for the functional diversity of HVAs at EO.
In adults, we confirmed that HVAs received specific visual information from V1 and LPN. However, inconsistent with the previous report (Blot et al., 2020), the response properties of LPN axons were similar to those of V1 axons and matched those of neurons in target HVA. The response amplitude of HVAs was decreased, and the response properties were changed by V1 inactivation in adults. These results suggest that V1-HVAs pathway is critical in adults. Now, we are testing the effect of LPN inactivation on functional differences among HVAs. | |
|
|
