A.O.L. Atkin,F.G. Garvan
Abstract:New identities and congruences involving the ranks and cranks of partitions are proved. The proof depends on a new partial differential equation connecting their generating functions.
What problem does this paper attempt to address?
The problem that this paper attempts to solve is to understand the complex interactions between the visual cortex and the thalamus, especially how the visual cortex controls the spatiotemporal response characteristics of the lateral geniculate nucleus (LGN) of the thalamus through feedback mechanisms. Specifically, the author proposes a new hypothesis: **The visual cortex controls the spatiotemporal structure of the cortical receptive field by feeding back to the LGN**. In addition, the author also constructs and analyzes a model, demonstrating the ability of this control mechanism to achieve object segmentation through statistical motion analysis in the visual field.
### Summary of the core issues in the paper:
1. **Understanding the cortical - thalamic interactions in complex sensory processing**:
- Although it is known that the cortical feedback to the thalamus is the main anatomical input, its impact on thalamic processing is difficult to explain.
- In order to understand complex sensory processing, a detailed concept of cortical - thalamic interactions needs to be established.
2. **Proposing a new hypothesis**:
- The visual cortex controls the spatiotemporal structure of the cortical receptive field by feeding back to the LGN.
- This control mechanism can be achieved by adjusting the response times of different types of cells in the LGN (such as lagged and nonlagged cells).
3. **Model demonstration**:
- A cortical - geniculate loop model was constructed, demonstrating the ability of this control mechanism to achieve object segmentation through statistical motion analysis in the visual field.
### Key background information:
- **Two types of X - type relay cells in the LGN**: There are two types of X - type relay cells (lagged and nonlagged) in the LGN, and they have significant differences in their temporal response characteristics. Among them, the peak response of lagged cells is delayed by about 100 milliseconds compared to nonlagged cells.
- **Feedback from cortical layer 6**: Feedback from cortical layer 6 can locally modulate the response patterns and times of LGN relay cells, thereby affecting the speed tuning of cortical simple cells.
- **Speed tuning control**: By adjusting the membrane potential of LGN relay cells, cortical feedback can change the speed tuning of cortical simple cells, and further affect their responses to moving visual features.
### Key points of the model:
- **Membrane potential modulation of GRCs**: The feedback signal from cortical layer 6 modulates the membrane potential of LGN relay cells through NMDA and metabotropic glutamate receptors (causing depolarization) or GABA_B receptors (causing hyperpolarization).
- **Response characteristics of lagged and nonlagged cells**: By adjusting the relative strength of feed - forward excitatory and inhibitory, typical lagged and nonlagged responses can be produced.
- **Total geniculate input rate**: By comprehensively analyzing the responses of lagged and nonlagged cells, the total geniculate input rate can be obtained and its correlation with the activity of cortical simple cells can be observed.
In conclusion, this paper aims to reveal how the visual cortex controls the spatiotemporal response characteristics of the LGN through feedback mechanisms by in - depth study of cortical - thalamic interactions, thereby achieving complex visual processing tasks such as object segmentation.