Since the first prototype (~ early 2014) was made to verify the hypothesis that external abdominal compression can assist human respiratory functions, the system has been developed to cover more diverse user groups. The size and volume were reduced to grant more mobility of its users while maintaining its mechatronic performances.
Exo-Abs prototype version 1
Exo-Abs prototype version 3
Exo-Abs prototype version 2
Exo-Abs @ T-RO
Exo-Abs prototype version 2.5
In order to assist various human respiratory functions (0.3 Hz ~ 3 Hz), the electromagnetic system should be able to generate repetitive impulse-like actuation. Owing to a simulation-based electromagnetic performance estimation protocol, an optimal combination of mechatronic components could be achieved after reduced numbers of trials.
Mechatronic components Exo-Abs
SImulation of the Exo-Abs electromagnetic system
30+ cases of in-depth & one-on-one interviews were conducted, and 5+ IRBs were proceeded.
Spinal cord injury patient (ex-singer)
Spinal cord injury patients
Spinal cord injury patient (ex-singer)
Students of vocal music
Spinal cord injury patient (ex-singer)
Patient after stroke (dysarthria)
Spinal cord injury patient (singer)
Spinal cord injury patient (intubation)
Based on fluid dynamics and pulmonary biomechanics, the customized sensors and their combination were proposed to optimally compromise between usability and accuracy.
Spirometer for precisely measuring the user's air usage
Chest-belt plethysmograph for monitoring the user's thoracic cavity motion
Sensor combination logic for assisting various respiratory functions
The developed sensor system was validated with medical-grade instruments. Especially, a hand-held sonograph was adopted to verify the phase-sensitiveness and precision of the Exo-Abs sensor system by spontaneously measuring the lung volume.
Schematic layout for validating the Exo-Abs sensor system
Image processing procedure for obtaining lung volume deviation measured with a sonograph
Until now, there were few studies that tried to analyze the human respiratory system subject to external abdominal compression. The overall model was reconstructed, and real-time sequential system identification process was proposed for the control of Exo-Abs.
Prof. Jin-Oh Hahn (the university of Maryland) advised me significantly in this process.
Abdominal cavity deformation model
Identification process for the abdominal cavity deformation model
parameters
Lung dynamics model
Identification process for the lung dynamics model parameters
Based on the proposed model, various respiratory functions of a typical user could be simulated via SIMULINK/MATLAB.
Simulated spontaneous/assisted breathing of a user (time plot)
Simulated spontaneous/assisted breathing of a user (phase plot)
SIMULINK/MATLAB based real-time simulation of biomechanical interaction between Exo-Abs and its user
For practicality, a reactive model-based impedance control was designed to provide assistance in an explainable manner.
Schematic for the overall control procedure
The applied impedance control alleviates the lung model parameters
Based on the proposed model, various respiratory functions of a typical user could be simulated via SIMULINK/MATLAB.
Analyzing voice sample via Praat
Validating the speaking assistance performance of Exo-Abs