Toward Robust and Scalable Brain-Computer Interfaces: Innovations in Neural Tracking, Signal Modalities, and Multi-Scale Analysis
Kaydedildi:
| Yayımlandı: | ProQuest Dissertations and Theses (2025) |
|---|---|
| Yazar: | |
| Baskı/Yayın Bilgisi: |
ProQuest Dissertations & Theses
|
| Konular: | |
| Online Erişim: | Citation/Abstract Full Text - PDF |
| Etiketler: |
Etiket eklenmemiş, İlk siz ekleyin!
|
| Özet: | Brain-computer interfaces (BCIs) hold great promise for restoring lost motor and sensory functions by enabling direct brain-device communication. Among various BCI paradigms, spike-based BCIs offer high temporal and spatial resolution but face challenges such as signal instability and difficulty in tracking neuronal identities across sessions, limiting their long-term reliability. This thesis addresses these limitations by developing novel neuron-tracking algorithms and validating alternative neural signals for BCI control. A key challenge in intracortical BCIs is the loss of temporal continuity in recorded neurons. Traditionally, spikes have been treated as anonymous across sessions, preventing the study of long-term neural dynamics related to learning, adaptation, and memory. To address this, I developed a longitudinal neuron-tracking algorithm for Utah arrays, enabling robust assessment of neural stability and plasticity over time. This work provides a framework for chronic BCI stability and long-term neural representation studies. Beyond spikes, I validated local field potential (LFP)-based neurofeedback in the ventral tegmental area (VTA), demonstrating volitional control over deep brain activity and the transferability of learned strategies, suggesting applications for neuromodulation therapies. Additionally, I optimized one-photon calcium imaging algorithms to shorten training periods, improving feasibility for calcium imaging-based BCIs. While my research primarily focused on single-modality BCIs, I also reviewed multi-scale neural analysis, highlighting the benefits and challenges of integrating spikes, LFPs, and functional imaging. These contributions lay the foundation for multi-modal BCI systems that improve neural decoding accuracy and robustness across behavioral contexts. By addressing neural stability, alternative signal modalities, and multi-scale integration, this thesis advances BCI research toward scalable, adaptable, and clinically viable neural interfaces, bringing us closer to stable, long-term BCI applications for assistive and therapeutic use. |
|---|---|
| ISBN: | 9798270232702 |
| Kaynak: | ProQuest Dissertations & Theses Global |