Page 276 - Control Theory in Biomedical Engineering

248 Control theory in biomedical engineering

Comfort of wearing
As a patient wears the exoskeleton the entire time during the rehab session,
which ranges from a half-an hour to 2 hours, any kind of discomfort is unex-
pected. A HEI needs to ensure the patient’s comfort. Exoskeleton joints cor-
respond to human joints and as such there are reactive forces and torques
between exoskeleton joints and human joints. The weight of the exoskel-
eton contributes to producing reactive forces and torques in the human
joint. Therefore, the lesser the weight, the more comfortable for the wearer.
When it comes to wearing, an open-type structure (e.g., CADEN-7 (Perry
et al., 2007), ARMIN (Nef et al., 2009a), SUFUL-7 (Gopura et al., 2009),
MARSE-7 (Rahman et al., 2014)) is always preferred in exoskeleton design.
Open-type structure is advantageous because of easy don/doff, comfortable
fitting, and better compliance. In addition, it is expected that the exoskel-
eton is connected to its wearer with flexible straps/links in between.

Alignment of exoskeleton joints with human joints

To perform exercises during rehabilitation, forces and torques generated in
exoskeleton joints must successfully be transferred to human joints. This
transfer does not happen properly if the exoskeleton joints are not aligned
with the human joints. In addition, misalignment might cause the exoskel-
eton wearer pain and/or discomfort during rehabilitation (Schiele and van
der Helm, 2006; Stienen et al., 2009). Therefore, to provide better compli-
ance and for successful transfer of forces and torques, exoskeleton joints need
to be aligned with corresponding human joints. The human shoulder joint,
also known as the GH joint, is the most biomechanically complex joint and
has many articulations that eventually leads to three general motions (i.e.,
vertical flexion/extension, abduction/adduction, and upper-arm internal/
external rotation) (Schenkman and de Cartaya, 1987). The CR of the shoul-
der joint does not remain fixed during shoulder movements and has two
additional movements: elevation/depression in the frontal plane and pro-
traction and retraction in the sagittal plane (Halder et al., 2000). In the lit-
erature, we found many exoskeletons where the shoulder joint is simplified
and modeled as a three-DOF ball-and-socket joint by ignoring motion of
CR joint (Rahman et al., 2014; Tang et al., 2014; Mahdavian et al.,
2015; Liu et al., 2016; Stroppa et al., 2017; Madani et al., 2017). However,
a few exoskeletons have added extra DOFs in modeling the shoulder joint to
realize motion of CR (Kiguchi et al., 2003, 2008; Nef et al., 2009b; Gopura
et al., 2009; Kim and Deshpande, 2017). These adjustments come with the