Luke Sy

Luke Wicent Sy, Nigel H. Lovell, Stephen J. Redmond

Goal: This paper presents an algorithm for estimating pelvis, thigh, shank, and foot kinematics during walking using only two or three wearable inertial sensors. Methods: The algorithm makes novel use of a Lie-group-based extended Kalman filter. The algorithm iterates through the prediction (kinematic equation), measurement (pelvis position pseudo-measurements, zero-velocity update, and flat-floor assumption), and constraint update (hinged knee and ankle joints, constant leg lengths). Results: The inertial motion capture algorithm was extensively evaluated on two datasets showing its performance against two standard benchmark approaches in optical motion capture (i.e., plug-in gait (commonly used in gait analysis) and a kinematic fit (commonly used in animation, robotics, and musculoskeleton simulation)), giving insight into the similarity and differences between the said approaches used in different application areas. The overall mean body segment position (relative to mid-pelvis origin) and orientation error magnitude of our algorithm ($n=14$ participants) for free walking was $5.93 \pm 1.33$ cm and $13.43 \pm 1.89^\circ$ when using three IMUs placed on the feet and pelvis, and $6.35 \pm 1.20$ cm and $12.71 \pm 1.60^\circ$ when using only two IMUs placed on the feet. Conclusion: The algorithm was able to track the joint angles in the sagittal plane for straight walking well, but requires improvement for unscripted movements (e.g., turning around, side steps), especially for dynamic movements or when considering clinical applications. Significance: This work has brought us closer to comprehensive remote gait monitoring using IMUs on the shoes. The low computational cost also suggests that it can be used in real-time with gait assistive devices.

Luke Sy, Nigel H. Lovell, Stephen J. Redmond

This paper presents an algorithm that makes novel use of a Lie group representation of position and orientation alongside a constrained extended Kalman filter (CEKF) to accurately estimate pelvis, thigh, and shank kinematics during walking using only three wearable inertial sensors. The algorithm iterates through the prediction update (kinematic equation), measurement update (pelvis height, zero velocity update, flat-floor assumption, and covariance limiter), and constraint update (formulation of hinged knee joints and ball-and-socket hip joints). The paper also describes a novel Lie group formulation of the assumptions implemented in the said measurement and constraint updates. Evaluation of the algorithm on nine healthy subjects who walked freely within a $4 \times 4$ m$^2$ room shows that the knee and hip joint angle root-mean-square errors (RMSEs) in the sagittal plane for free walking were $10.5 \pm 2.8^\circ$ and $9.7 \pm 3.3^\circ$, respectively, while the correlation coefficients (CCs) were $0.89 \pm 0.06$ and $0.78 \pm 0.09$, respectively. The evaluation demonstrates a promising application of Lie group representation to inertial motion capture under reduced-sensor-count configuration, improving the estimates (i.e., joint angle RMSEs and CCs) for dynamic motion, and enabling better convergence for our non-linear biomechanical constraints. To further improve performance, additional information relating the pelvis and ankle kinematics is needed.

Luke Sy, Michael Raitor, Michael Del Rosario et al.

Goal: This paper presents an algorithm for accurately estimating pelvis, thigh, and shank kinematics during walking using only three wearable inertial sensors. Methods: The algorithm makes novel use of a constrained Kalman filter (CKF). The algorithm iterates through the prediction (kinematic equation), measurement (pelvis position pseudo-measurements, zero velocity update, flat-floor assumption, and covariance limiter), and constraint update (formulation of hinged knee joints and ball-and-socket hip joints). Results: Evaluation of the algorithm using an optical motion capture-based sensor-to-segment calibration on nine participants ($7$ men and $2$ women, weight $63.0 \pm 6.8$ kg, height $1.70 \pm 0.06$ m, age $24.6 \pm 3.9$ years old), with no known gait or lower body biomechanical abnormalities, who walked within a $4 \times 4$ m$^2$ capture area shows that it can track motion relative to the mid-pelvis origin with mean position and orientation (no bias) root-mean-square error (RMSE) of $5.21 \pm 1.3$ cm and $16.1 \pm 3.2^\circ$, respectively. The sagittal knee and hip joint angle RMSEs (no bias) were $10.0 \pm 2.9^\circ$ and $9.9 \pm 3.2^\circ$, respectively, while the corresponding correlation coefficient (CC) values were $0.87 \pm 0.08$ and $0.74 \pm 0.12$. Conclusion: The CKF-based algorithm was able to track the 3D pose of the pelvis, thigh, and shanks using only three inertial sensors worn on the pelvis and shanks. Significance: Due to the Kalman-filter-based algorithm's low computation cost and the relative convenience of using only three wearable sensors, gait parameters can be computed in real-time and remotely for long-term gait monitoring. Furthermore, the system can be used to inform real-time gait assistive devices.

Jinmiao Huang, Cesar Osorio, Luke Wicent Sy

Background and Objective: Code assignment is of paramount importance in many levels in modern hospitals, from ensuring accurate billing process to creating a valid record of patient care history. However, the coding process is tedious and subjective, and it requires medical coders with extensive training. This study aims to evaluate the performance of deep-learning-based systems to automatically map clinical notes to ICD-9 medical codes. Methods: The evaluations of this research are focused on end-to-end learning methods without manually defined rules. Traditional machine learning algorithms, as well as state-of-the-art deep learning methods such as Recurrent Neural Networks and Convolution Neural Networks, were applied to the Medical Information Mart for Intensive Care (MIMIC-III) dataset. An extensive number of experiments was applied to different settings of the tested algorithm. Results: Findings showed that the deep learning-based methods outperformed other conventional machine learning methods. From our assessment, the best models could predict the top 10 ICD-9 codes with 0.6957 F1 and 0.8967 accuracy and could estimate the top 10 ICD-9 categories with 0.7233 F1 and 0.8588 accuracy. Our implementation also outperformed existing work under certain evaluation metrics. Conclusion: A set of standard metrics was utilized in assessing the performance of ICD-9 code assignment on MIMIC-III dataset. All the developed evaluation tools and resources are available online, which can be used as a baseline for further research.