본문 바로가기 주메뉴 바로가기
22대 대한민국 국회 국회정보길라잡이 국회의원검색
회원가입 로그인
HOME

정보검색

소장정보 검색

국회도서관 홈으로 정보검색 소장정보 검색

결과 내 검색

동의어 포함

고급검색

인명/단체명

저자정보 상세정보
인명/단체명을 입력하세요.

키워드

  • 대표어

  • 외국어

-

목차보기

Title Page 2

Contents 5

List of Abbreviations 8

Abstract 14

Chapter 1. Background 16

1.1. Stroke 16

1.2. Neuroplasticity 17

1.3. Motor Imagery 19

1.4. Motor Imagery-based Brain-Computer Interface (MI-BCI) 22

Chapter 2. Introduction 26

2.1. MI-BCI training in stroke rehabilitation 26

2.2. Previous attempts to guide MI 27

2.2.1. MI guidance system 28

2.2.2. Neurofeedback system 30

2.3. Objective of the thesis 33

Chapter 3. Proposed MI guidance system 34

3.1. VHI-based MI guidance system 34

3.1.1. Development 34

3.2. Validation in healthy people 37

3.2.1. Participants and experimental setup 37

3.2.2. Data Acquisition and Analysis 39

3.2.3. Results 41

3.2.4. Discussion 44

3.2.5. Conclusions 49

3.3. Validation in stroke patients 49

3.3.1. Participants and experimental setup 50

3.3.2. Data Acquisition and Analysis 52

3.3.3. Results 56

3.4. Conclusion 67

Chapter 4. Neurofeedback system 68

4.1. Proposed Neurofeedback system 68

4.2. Validation of the Proposed System 80

4.2.1. Implementation Details 80

4.2.2. Offline Validation of the Scoring Algorithm 82

4.2.3. Online Neurofeedback Application 94

Chapter 5. General Discussion 105

5.1. Revisiting the Research Hypothesis 105

5.2. Interpretation of Findings 105

5.3. Limitations and Considerations 107

5.4. Implications for BCI and Rehabilitation Fields 109

Chapter 6. Conclusion 112

6.1. Summary of Contributions 112

6.2. Contributions 112

6.3. Closing Remarks 113

References 114

논문요약 132

List of Tables 13

Table 3.1. Statistical analysis for the contralateral and ipsilateral channel. Friedman test and Wilcoxon... 43

Table 3.2. Characteristics of participants with stroke (NIHSS : National Institutes of Health Stroke Scale;... 51

Table 3.3. Questionnaire score for body ownerhsip transfer 57

Table 3.4. Characteristics of clustered source for body ownership illusion; asterisk indicated statistical dif-... 59

Table 4.1. Comparison of neurofeedback training approaches in previous studies 70

Table 4.2. Statistical analysis results for peak values across motor tasks. Asterisks (*) indicate significance... 84

Table 4.3. Proportion of subjects exhibiting saturated scores in each group 88

Table 4.4. Comparison between EEGNet-based score and the proposed distance-based neurofeedback score 93

Table 4.5. Source-level α-band ERSP analysis in motor-related clusters. Post-hoc Wilcoxon signed-rank... 101

Table 4.6. Source-level α-band ERSP analysis for all dipole clusters. Statistical significance from Wilcoxon... 101

List of Figures 9

Figure 1.1. Structure of typical MI-BCI 22

Figure 2.1. Structure for two kinds of attempts to guid MI 28

Figure 3.1. Schematic diagram and apparatus for VHI. The gray hand denotes the subject's real hand, and... 36

Figure 3.2. Experimental protocol: (a) Setup of pure motor imagery (MI-P) and pure motor execution (ME-... 38

Figure 3.3. Averaged event-related spectral perturbation (ERSP) maps of each paradigm. The red dotted... 42

Figure 3.4. Topographic distribution of average ERSP for each paradigm, within a two-second window after... 42

Figure 3.5. Findings in the contralateral and ipsilateral channel: (a) peak event-related desynchronization... 43

Figure 3.6. Survey results for each paradigm: (a) Amplified MI guidance system vs. MI-P; (b) Most... 45

Figure 3.7. Experimental setup (a) schematic diagrams of VHI, VT and VM; (b) experimental photo and... 53

Figure 3.8. Galvanic skin response for unexpected threat (body ownership transfer index (a) grand average... 57

Figure 3.9. Beta band ERD for quantifying body ownership transfer of AO and VHI (a) grand average of... 58

Figure 3.10. Topographical distribution map (a) for beta band power (b) log scaled p-value of statistical... 59

Figure 3.11. alpha ERD for MI enhancement assessment (a) Relative potentials for alpha band power across... 60

Figure 3.12. Topographical distribution map (a) for alpha band power (b) log-scaled p-value of statisti-... 61

Figure 3.13. Characteristics of clustered source for MI enhancement in motor-related area (a) for right-... 62

Figure 3.14. Subgroup analysis for stroke patients; (a) motor function (b) cognitive function 65

Figure 3.15. Results for correlation between body ownership transfer and MI enhancement (BOT: body... 66

Figure 4.1. Overall system diagram of neurofeedback system 69

Figure 4.2. System architecture of the proposed neurofeedback system 71

Figure 4.3. System architecture of the proposed feature selection process, including individualized channel... 73

Figure 4.4. Feature Space Transformation Before and After CSP Projection 73

Figure 4.5. Illustration of CSP-based channel selection. The CSP algorithm identifies spatial filters that... 74

Figure 4.6. Anomaly rejection using DBSCAN. (Left) Clustered alpha-band features show separation of... 76

Figure 4.7. Structure of the score calculation algorithm using Euclidean distance to reference and rest maps 77

Figure 4.8. Example of 1D visual neurofeedback interface based on peak distance score within each MI task... 79

Figure 4.9. Schematic diagram of index calculation across three candidate strategies 84

Figure 4.10. Temporal profiles and statistical differences of motor performance indices across three... 85

Figure 4.11. Boxplots of peak values for each index across motor paradigms 85

Figure 4.12. Comparison between ERD amplitude in ROI channel (C3), proposed distance-based score, and... 87

Figure 4.13. Subject-wise comparison of proposed distance-based score (red) and CSP + LDA score (blue).... 87

Figure 4.14. Discriminability analysis based on AUC values of the proposed score (blue) and CSP+LDA... 89

Figure 4.15. Comparison of spatial channel importance derived from EEGNet (left: without regularization;... 91

Figure 4.16. Discriminability analysis (AUC) of EEGNet-based scores across paradigms for healthy (a) and... 91

Figure 4.17. Time-series visualization of EEGNet-based scores across different imagery conditions. Left:... 92

Figure 4.18. Experimental photo for neurofeedback session 95

Figure 4.19. Median neurofeedback scores for task and rest trials across all sessions. Asterisks (*) indicate... 99

Figure 4.20. Channel analysis (a) ERSP time-frequency maps and (b) topographical distributions (c) Aver-... 100

Figure 4.21. Clustering evaluation metrics used to determine the optimal number of dipole clusters (n＝11) 100

초록보기

 뇌졸중 환자의 효과적인 재활을 위해, 실제 움직임 없이 뇌의 운동 네트워크를 활성화할 수 있는 운동 상상이 널리 활용되고 있다. 그러나 운동 상상의 본질적인 모호성과 뇌졸중 환자의 운동 및 인지 기능 저하로 인해, 그 효과를 안정적으로 끌어내는 데에는 여전히 어려움이 존재한다.

본 연구는 이러한 문제를 해결하기 위해, 몰입형 안내와 실시간 뇌 상태 기반의 정량적 피드백으로 구성된 새로운 운동 상상 기반 재활 프레임워크를 제안한다. 먼저, 가상손 착각을 활용해 신체 소유감을 유도하여, 상상 수행을 보다 생생하고 일관되게 만들어주는 운동 상상 안내 시스템을 개발하였다. 이 시스템은 뇌졸중 환자를 포함한 대상자에게 적용되어, 기존 운동상상 수행의 한계를 보완할 수 있음을 확인하였다. 또한, 개별 사용자의 뇌 활성 패턴을 기준으로 상상 상태를 평가하고 실시간 피드백을 제공하는 뉴로피드백 시스템을 구성하고, 건강한 참가자를 대상으로 한 기초 실험을 통해 운동 상상 수행의 점진적 향상과 뇌 반응의 변화를 확인하였다.

두 시스템은 폐루프 구조로 통합되어, 향상된 훈련 신호와 정량적 피드백을 통한 실시간 학습을 연결하는 새로운 재활 훈련 체계를 구성한다. 이는 기존 운동 상상 기반 재활 시스템의 한계를 극복할 수 있는 실용적이고 임상 적용 가능한 대안으로 기대된다.

추천서가 (다양한 추천 자료를 만나보세요)

권호기사보기

권호기사 목록 테이블로 기사명, 저자명, 페이지, 원문, 기사목차 순으로 되어있습니다.
기사명 저자명 페이지 원문 기사목차

권호

기사명 원문보기 기사목차
닫기