TY - GEN
T1 - ACCURACY OF MEDIAPIPE VISUAL HAND TRACKING FOR USE IN MEDICAL TRAINING PROCEDURES
AU - Budzinski, Cynthia
AU - Wu, Hang Ling
AU - Sarraf, Elie
AU - Miller, Scarlett Rae
AU - Moore, Jason Zachary
N1 - Publisher Copyright:
© 2024 by ASME.
PY - 2024
Y1 - 2024
N2 - This paper explores the need for individualized feedback in medical simulation training for procedures such as endotracheal intubations and colonoscopies. Current pass-fail assessments lack the specificity required for skill refinement, leading to potential complications in critical procedures. This experiment investigates the application of visual hand tracking technology, specifically Google's MediaPipe, to offer quantitative feedback. Using a wooden hand model on a linear motor with a depth camera, this experiment explores tracking hand movements in real time. Google's MediaPipe software tracks 3D hand motions, generating data points for analysis. Results demonstrated consistent patterns in hand movements corresponding to linear motor actions. Total distance traveled by each point of interest and error analysis, averaging 3.1 mm, provided insights into the accuracy of visually measured hand movements. Visual hand tracking, such as MediaPipe, may prove to be a promising tool for refining operator techniques in critical medical procedures. The experiment explores the technology's accuracy and underscores its potential application in medical training, addressing a gap in current practices and emphasizing the importance of tailored feedback for improved patient outcomes.
AB - This paper explores the need for individualized feedback in medical simulation training for procedures such as endotracheal intubations and colonoscopies. Current pass-fail assessments lack the specificity required for skill refinement, leading to potential complications in critical procedures. This experiment investigates the application of visual hand tracking technology, specifically Google's MediaPipe, to offer quantitative feedback. Using a wooden hand model on a linear motor with a depth camera, this experiment explores tracking hand movements in real time. Google's MediaPipe software tracks 3D hand motions, generating data points for analysis. Results demonstrated consistent patterns in hand movements corresponding to linear motor actions. Total distance traveled by each point of interest and error analysis, averaging 3.1 mm, provided insights into the accuracy of visually measured hand movements. Visual hand tracking, such as MediaPipe, may prove to be a promising tool for refining operator techniques in critical medical procedures. The experiment explores the technology's accuracy and underscores its potential application in medical training, addressing a gap in current practices and emphasizing the importance of tailored feedback for improved patient outcomes.
UR - http://www.scopus.com/inward/record.url?scp=85205993768&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=85205993768&partnerID=8YFLogxK
U2 - 10.1115/DMD2024-1039
DO - 10.1115/DMD2024-1039
M3 - Conference contribution
AN - SCOPUS:85205993768
T3 - Proceedings of the 2024 Design of Medical Devices Conference, DMD 2024
BT - Proceedings of the 2024 Design of Medical Devices Conference, DMD 2024
PB - American Society of Mechanical Engineers (ASME)
T2 - 2024 Design of Medical Devices Conference, DMD 2024
Y2 - 8 April 2024 through 10 April 2024
ER -