Browsing by Subject "Monkey"
Now showing 1 - 2 of 2
Results Per Page
Sort Options
Item Neural basis of context processing dysfunction in schizophrenia: a monkey model(2013-08) Blackman, Rachael KeirRanking among the top ten causes of years lost due to disability worldwide, schizophrenia is a psychiatric disease whose pathophysiology has not been fully characterized to-date. The objective of my dissertation is to characterize the change in neuronal information processing that leads to cognitive dysfunction in the disease. To this end, I trained monkeys to perform a translational cognitive task that measures context processing deficits in schizophrenia patients. Context processing is the ability to use prior contextual information maintained in working memory to conditionally respond to subsequent stimuli. I then recorded neural activity from the prefrontal (PFC) and posterior parietal (PAR) cortex after administering N-methyl-D-aspartate (NMDA) receptor antagonists that 1) are known to mimic symptoms of schizophrenia in human control subjects, and 2) block the NMDA receptor which is thought to be dysfunctional in the disease. I found that after drug administration, monkeys produced essentially the same pattern of behavioral errors on this task that schizophrenia patients commit. Further, by recording neural activity in PFC and PAR during the period of cognitive impairment, I was able to determine that the maintenance of contextual information in PFC was selectively diminished. In addition, I was able to use trial-by-trial changes in neural activity in both cortical areas to predict errors on the task, linking neuronal activity to behavioral performance. Overall, I have been able to characterize for the first time the change in cortical information processing at a cellular level that could account for context processing dysfunction in schizophrenia.Item Segmentation and Dense Keypoints Estimation of Monkeys(2021-12) Yu, HaozhengAnimal tracking and pose estimation are core topics in neuroscience. However, for monkeys, current deep learning based algorithms often fail to perform well on segmentation and dense keypoints estimation due to the lack of annotated training data. In this thesis, we address this challenge by developing transfer learning based deep learning algorithms without using fully-annotated monkey data. We develop a bootstrapping strategy to refine the pretrained segmentation model on monkey data annotated with 2D sparse landmarks. In addition, we implement a voxel-based visual hull reconstruction approach to recover the 3D monkey pose from the silhouettes. For dense keypoints estimation, we follow a similar bootstrapping strategy to refine a pretrained HRNet, which is then used to learn a dense keypoint detector by leveraging multiview consistency. Our methods outperform the baseline methods on in-the-cage and in-the-wild monkey data.