Background: Multiple fractures, particularly femoral fractures, are increasingly prevalent and associated with high mortality rates and significant functional impairments. This highlights the urgent need for effective rehabilitation strategies, such as robot-assisted training, to enhance recovery and improve quality of life. Objectives: This study aimed to evaluate the clinical effectiveness of robotassisted rehabilitation for multiple femoral fractures. Design: Single-subject design. Methods: A 15-day A-B-A' single-subject design was employed. A man in his 30s with multiple fractures underwent standard rehabilitation during the baseline (A) and regression baseline (A') phases, with robotic therapy introduced during the intervention phase (B). Daily assessments of mobility and balance were analyzed using the two-standard deviation method. Results: Robotic therapy led to significant improvements: the TUG test time decreased from 16.21±0.64 seconds (A) to 10.63±0.46 seconds (B) and 9.64±0.35 seconds (A'). The 10 MWT time improved from 6.31±0.64 seconds (A) to 5.41±0.17 seconds (B) and 5.01±0.12 seconds (A'). LOS increased from 364.01±35.83 cm² (A) to 484.67±29.97 cm² (B) and 518.03±18.82 cm² (A'). Plantar pressure imbalance (59.2% right, 40.8% left in A) was corrected to nearly equal distribution in B (49.4%/50.6%) and A' (50.8%/49.2%). Conclusion: Robotic rehabilitation therapy improves balance and weightbearing capacity in patients with multiple fractures, suggesting its effectiveness as an early intervention following bone union.

본 연구에서는Stable Diffusion 프레임워크를 활용하여 게임 스타일의 스케치, 특히 도시 장면을 생성하는 방법 을 소개한다. 확산 기반의 모델인Stable Diffusion은 쉬운 접근성과 뛰어난 성능으로 많은 연구자와 일반인들에 게 선호되며, 텍스트-스케치, 이미지-스케치의 생성이 가능하다. Stable Diffusion의 몇 가지 문제는 이미지의 국 소성 보존 문제 및 미세 조정인데, 이를ControlNet과DreamBooth를 사용하여 해결한다. 결과적으로, 본 연구를 통 해 게임 제작에 사용될 수 있는 텍스트-스케치, 이미지-스케치 생성이 가능하며, 더 나아가 아티스트를 돕는 툴 로도 활용될 수 있다.