Ankle Injury Rehabilitation Robot (AIRR): Review of Strengths and Opportunities Based on a SWOT (Strengths, Weaknesses, Opportunities, Threats) Analysis

: Generally, severity, any additional damage to the joint surface, and the optimal rehabilitation inﬂuence the recovery of an ankle injury. Optimal rehabilitation is the only approach for a human to heal as soon as possible. Ankle injury rehabilitation robots (AIRRs) are designed to fulﬁl the ideal rehabilitation by providing the required accuracy, consistency, and repeatability, compared to conventional rehabilitation methods. This review is to explore the performance of the existing AIRR using a SWOT analysis with a focus on the strengths and opportunities of an AIRR. Sources from journals and conference papers are selected for review after several screenings, according to the search conditions set by the authors. The results have shown a large group of AIRRs could accomplish all basic ankle motions and select parallel mechanisms to drive the foot platform. Most AIRRs provides crucial feedback sensors, such as position, torque, and angle. These factors determine the accuracy of the foot platform. Both the electrical/pneumatic actuation and wearable/platform-based AIRRs have their purpose for rehabilitation and must be considered as equal contributions to ankle injury rehabilitation research using robots. Opportunities to provide innovation to the already established AIRR research still exist in the ability to accommodate complex motion ankle rehabilitation exercises and to establish teaching and playback into the rehabilitation procedures for AIRRs. In general, the existing strengths of AIRRs provide advantages to patients where they can enhance the rehabilitation procedures while opportunities and knowledge gaps for AIRR research are still open to improvement.


Introduction
The ankle (talocrural) is one of the most exceptional human joints where the functional stability and the mobility of the joints are compatible. This human joint feature is unique, compared to other human joints, such as the elbow and saddle joint, where its function is limited to either functional mobility or functional stability [1]. These unique features of the ankle joint have allowed people to perform physical motions, such as jumping, walking, running, etc. [2]. However, the ankle joint is prone to injury, attributable to domestic-related accidents or sports-related accidents [3]. There are three classifications of ankle injury:

Inclusion Criteria for the Review
Sources were accepted for review if the following acceptance criteria for review were met: (1) Sources were accessible to the authors. (2) The articles were written in the English Language.
(3) The existing AIRRs must be used to rehabilitate an ankle injury (this includes ankle sprains or ankle strains). (4) AIRRs must be developed with physical prototypes and not concept designs and analytical prototypes. (5) Any sources that were encompassing both the rehabilitation for an ankle disability and the rehabilitation for an ankle injury. (6) Sources after 1990 were included in this review.

Exclusion Criteria for Review
Sources were rejected for review if the following rejection criteria for review were identified: (1) Sources were not written in the English Language.
(2) The existing AIRRs specify the purpose of the rehabilitation are for an ankle disability only and not the rehabilitation for an ankle injury. (3) Studies with a lack of information regarding the existing AIRRs. (4) Studies that were still under the proposal stage.
(5) Sources before 1990 were excluded from this review.
These criteria were set in such a way because a study, based on a physical prototype can provide a lot of necessary information for the review, compared to an analytical prototype, especially for information regarding the hardware, such as sensors and actuators.

Use of the SWOT Analysis in the Review
The main aim of using a SWOT analysis is to characterize the strengths, weaknesses, opportunities, and threats, to decide the strategic position of a company [21,22]. Although a SWOT analysis is used to assess a person, place, product, or even an industry, a SWOT analysis has also often been used in the field of healthcare, compared to agriculture, according to Benzaghta et al. (2021) [22]. This study was significant as healthcare rehabilitation requires decision-making to determine the practicality of the listed options [23,24]. Therefore, in this technical review, the SWOT analysis, with the exception of weaknesses and threats indicators, is used to evaluate the research trend of AIRR developments, based on the obtainable sources presented through the technical knowledge of the reviewers. Figure 1 presents the structure of the SWOT analysis with examples of the components relating to AIRRs. The main reason the negative factors from the SWOT analysis, such as weaknesses and threats, are discarded from the review is that these indicators provide a biased evaluation of the existing robot as some of the weaknesses are intentional to fulfil a specific purpose or scope of a previous project. For example, one of the researchers might opt to develop a 1 DOF (degree of freedom) AIRR only, from the start, as part of his or her research scope. However, if the weaknesses and threats criteria from the SWOT analysis are applied, his or her works would be unfairly undermined in this review. For instance, a robot that can fulfil the whole 3 DOFs (representing the whole ankle motion) is better than fulfilling only 1 DOF. Therefore, to prevent the unfair judgment on the established AIRR, only the strengths and opportunities of the existing robots are selected as the main indicators for this article, from the SWOT analysis, to highlight the accomplishment of the existing AIRRs (through strengths) and to provide inspiration to the incoming researchers to provide new knowledge and improvements to the already established AIRR research (through opportunities).
The main benefit of the SWOT analysis is that this analysis can present the internal and external factors that were beneficial for the reviewed AIRR, by looking at the established AIRR from two perspectives-present and future. Additionally, this review is not only celebrating the achievements of the previous AIRRs but also provides suggestions for new contributions and knowledge gaps that have been identified by the reviewers, which can inspire future researchers of AIRRs to create a breakthrough in the established AIRR research.

Selection Review Process
Based on the selection reviewing process in Figure 2, the sources that focused on an ankle disability and related to neurological impairments only, were excluded from the review . Sources that contained only the analytical prototyping, analysis, design, and simulation stages were also removed . Sources were gathered until August 2022. Based on the selection process, 27 AIRRs were selected for review. Table 1 presents a summary comparative table on ankle injury rehabilitation robots (AIRRs), according to the features of AIRRs, such as the DOF, actuation, mechanism, covered ankle motion, sensors, platform or wearable, and training mode.

Strengths
The considerations for strengths of AIRRs are described as follows.

Capability
There will be three major perspectives in reviewing the existing AIRRs, in terms of capability, where the authors will study how much the AIRR is capable to fulfil most ankle motions and ankle exercises with the current structure. This can be determined through the number of the DOFs, the covered ankle motion, and the training mode. A higher number of DOFs, the covered ankle motion, and training mode for the AIRR, means the robot is of the highest capability to provide most of the ankle rehabilitation training by minimizing the amount of rehabilitation equipment needed to complete the rehabilitation.
Secondly, this category investigates the feedback capability of the AIRRs, where the implementation of the sensors was crucial for the AIRR system to ensure the safety and accuracy of the user and the sensors could provide a closed-loop feedback for the AIRR control system. This is important to ensure the AIRRs can be operated accurately and safely. The users and the system would know whether the motion of the foot platform (end effector) had exceeded beyond the limit of the ankle motion, or the robot was applying more than the necessary amount of force and torque to the injured ankle. This factor is crucial to ensure the effectiveness of AIRRs in quick recovery, while reducing the dependency on the strenuous conventional approach to physiotherapy and preventing re-injury. Additionally, the simple use of the mechanisms can determine the ease of control.
Thirdly, one of the features of AIRRs, such as platform-based or wearable-based, can also become an important indicator of a capability. With a platform based AIRR, it had the potential to perform various ankle rehabilitation exercises, such as muscle strength training and motion therapy. Furthermore, the ROM of the ankle-foot complex was restricted as they were bound in the limited workspace of the robot [10]. This allowed the AIRR to hold onto the actuated ankle completely, to make sure the injured ankle could experience the same motion or force that is usually carried out through the conventional treatment. With a wearable based AIRR, the patients could also apply the robot to improve the gait for stroke patients who experienced an ankle disability. This widens the robot's capabilities. The actuation features are ignored for this consideration as the selection of the electrical and pneumatic actuation, by the previous study, is based on the scope of the researchers. However, this feature is only limited to the discussion about the portability.

Portability
Portability is one of the considerations for the AIRR review, as portable AIRRs can help the therapist to move the equipment from one place to another and encourage a home-based rehabilitation. This will help patients to rehabilitate more frequently, as a visit to the hospital can be minimized. In this review, the type of actuation and mechanism was chosen to define the portability of the robot. For example, an electrical actuation system requires less additional equipment, compared to a pneumatic-based system, whereas a pneumatic system requires additional equipment, such as an air compressor to run the system. Nevertheless, the introduction of the compact air compressors available on the market has helped pneumatic-based machines to be more portable. In terms of the mechanism, the parallel mechanism and the linear actuator have helped AIRRs to be more compact and much easier to carry.
Features, such as the DOF, the covered ankle motion, sensors, platform/wearable, and training mode are ignored as these factors are not influencing the portability of AIRRs. The main reasons are: (1) the ability to fulfil the ankle motion and training mode, are a must for AIRRs, regardless of the portability, (2) due to its size, the application of the sensors would barely affect the weight of AIRRs, and (3) platform-based and wearable-based AIRRs have their purpose to fulfil certain ankle rehabilitation treatments, based on their structure. Thus, these features cannot be considered for portability.

Opportunities
The consideration for opportunities of AIRRs are described as follows.

Innovation
Features to determine the potential future innovation for the established AIRRs are the number of DOFs, the actuation, mechanism, ankle motion, sensors, wearable/platform, and training mode. These features focus on the hardware components that can be introduced for a future project. Furthermore, the opportunities in ankle motion are already covered under the scope of usage, therefore this feature will be ignored in the innovation consideration.

Scope of Usage
For the scope of usage, the covered ankle motion and training mode were chosen as these features were focused more on the potential additional exercises and training that can be introduced or need to be stressed for the potential AIRR project or upgrade. Other features, such as the DOFs, mechanism, and sensors are ignored as these features emphasize the hardware component of AIRRs instead of the exercise and training. Table 2 presents summary comparative table on the ankle injury rehabilitation robot (AIRR)'s features with the strengths (Capability and Portability) and opportunities (Innovation and Scope of Usage).    The strength of using a pneumatic actuator is that the pneumatic muscle is compliant and lightweight, that allowing the natural interaction with the patients.

N/A
Innovation -The combination of electrical and pneumatic actuations can be one new thing to explore where future researchers can adapt both the pneumatic advantage of the intrinsic softness and the electrical advantage of the controlled precision. The strength of using a parallel mechanism was because the payload or weight was carried by several links in parallel, which reduced the payload-to-weight ratio. Thus, reducing the wear and tear of the manipulator is suitable for the ankle when its workspace is very limited.

Innovation
-Continuation of the parallel mechanisms in future research is still relevant as the parallel mechanism is seen as the best approach to actuate the injured ankle. It is suggested that the of new parallel mechanisms would be strongly encouraged. -Implementation of the soft robotic mechanisms in AIRRs. - The cable-driven mechanism can be explored more but needs to address the ability to provide both the push and pull motions/forces on the actuated ankle.  Torque is an important parameter in measuring the effectiveness of the ankle rehabilitation treatment, where the AIRR needs to apply the right torque on the actuated ankle, according to the procedures.

N/A
Innovation -Continuation of the installation and combination of the position, torque, and angle sensors would be beneficial for the accuracy of AIRRs, as the sensors can provide the necessary torque, position, and rotation angle of the foot platform. -Implementation of the mechanical sensor where the motion of the foot platform can be limited, by being independent from the system. -Continuous exploration using sensors to implement virtual reality treatment, where interaction between patients and AIRRs encourages recovery.  Wearable-based AIRRs have one main advantage, in that they can provide walking aid rehabilitation, compared to the platform-based AIRRs. Contrasted with the platform based AIRR, this AIRR can be expanded in rehabilitating the lower limbs and not just limited to the ankle. The application for an improved gait for stroke patients can also be encouraged with wearable based AIRRs.

-
The combination of platform-based and wearable-based mechanisms will provide more coverage on the ankle exercise, using only one AIRR.

Training Mode
Capability -All AIRRs will be involved in passive and isokinetic training or exercise where rehabilitation equipment will be used, during the passive rehabilitation stage and isokinetic is a type of exercise that uses a specialized machine, such as the AIRR.

Discussion
The purpose of the review is to evaluate the existing AIRRs, using strengths and opportunities from the SWOT analysis as an additional supplement to the already established guidelines, for future researchers to advance their research and development of AIRRs. Although the medical results were much more reliable in evaluating the feasibility of AIRRs, this approach is only useful if there is plentiful medical data available. However, in this case, not all sources that have been gathered are presented with medical data to demonstrate the feasibility of the AIRR that has been developed. Furthermore, most sources from the previous research were not undergoing clinical trials which could provide crucial medical data, to evaluate its effectiveness. In this review, the authors can only use the available critical technical information that can be gathered from the articles, to evaluate the AIRRs. Using this approach helps the reviewers to determine the features of the robot, according to its strengths, and help to predict by presenting potential opportunities that can be identified with the limited medical information available.

Strengths
The overall result, based on strengths, shows that the majority of the selected AIRRs had three DOFs and above, which should cover the minimum number of six basic ankle motions, although not all these three DOF AIRRs translate to a full six basic ankle motions. The actuation, the electrical and pneumatic actuators, and the platform-based, or wearable based indicators played their purpose, specifically in aiding the rehabilitation procedures in which the actuation and the selection of the structure for the AIRRs must be seen as equal in their respective advantages.
In terms of mechanism, the advantage of the parallel mechanism in AIRRs is the payload or weight being carried by several links, in parallel, which reduces the payloadto-weight ratio, thus reducing the wear and tear of the manipulator. Furthermore, its simple structure supported its end effectors in multiple areas, resulting in stifles. Similarly, the positioning errors had been 'averaged out' in each leg, thus the accuracy has been improved. In terms of control, the use of a parallel mechanism guaranteed an easier control and was easier to determine the movement of the robot, due to the non-cumulative joint error. Unlike the serial mechanism, the parallel mechanism can be highly efficient, due to the position of the motors close to the base, which means that the power from the motor was quickly transferred towards the base, resulting in fewer losses of power.
Additionally, the strength of using a cable-based AIRR, where the weight of the driving linkage can be ignored, allows this cable-driven AIRR to be lighter, compared to the rigid linkage driven AIRR, and using a linear actuation will allow the AIRR to be compact and easy to control. While the rigid linkages can provide additional support on the weight of the patients, thus allowing the need to support the weight-based exercise, such as balancing. In summary for strengths, in terms of the ankle motion, less covered ankle motion means less hassle to control the machine while the ability for the AIRR to fulfil all ankle motions would be a great advantage in reducing the dependency on a lot of ankle rehabilitation equipment required for the complete recuperation. The soft robotic mechanism can be a popular approach for the rehabilitation, as the soft robotic mechanism can adapt to patients' wearability and is safer for the human-machine interaction, as the mechanism is using soft materials to actuate the motion, instead of hard materials for the rigid links [201].
For the sensors, the use of load cells or a force sensor is one of the main strengths for AIRRs to apply the right amount of force to the injured ankle, based on the procedures. The addition of the position, angle, and torque sensors is also an advantage where these sensors can help to ensure that the foot platform position, rotation, and applied torque are correct.
Finally for training mode, apart from the passive and isokinetic exercise ankle rehabilitation, additional training modes, such as active exercise training, would widen the rehabilitation training scope and help to reduce the dependency on the rehabilitation equipment. The balance or proprioception exercise will be an added advantage to restore balance on the ankle.

Opportunities
The overall result, based on strengths shows that minimizing the number of mechanisms or reducing the number of DOFs, and yet being able to fulfil all basic ankle motions, would be encouraging in future research. Hopefully, the future mechanisms for AIRRs would be much simpler and easier to control. Secondly, the concept of reconfigurable AIRRs will allow AIRRs to provide additional DOFs or training modes to their overall operation, as the open structures of the robots (if designed) allow any possibility of additional structures to be added although reconfiguration, based on the size of the patient's foot, which has already been explored by the previous AIRRs. For the mechanism, the continuation of the parallel mechanism is still relevant in future research, with the introduction of the new mechanism. Additionally, the cable mechanism in AIRR research can be explored but must address issues on the pull/push motions on the actuated ankle.
The combination of the pneumatic and electrical actuation would be an excellent opportunity, where the advantages of electrical and pneumatic actuations can be combined into one AIRR. Additionally, if an AIRR can have both wearable and platform-based mechanisms, through a structure reconfiguration or structure integration, this will widen the ankle exercise treatment and reduce the amount of equipment needed to fully rehabilitate the ankle injury.
In terms of the covered ankle motion, a combination of two or more basic ankle motions has a great potential for future AIRRs, as some of the ankle rehabilitation exercises or training require more than two or more basic ankle motions. This provides good opportunities for future researchers who want to explore complicated ankle exercises but prefer to do them with as few ankle motions as possible. With this approach, time and effort will be committed to setting and programming much more complicated exercises, instead of fulfilling all basic ankle motions through simple exercises. For the sensor, a combination of the position, torque, and angle sensors will help the foot platform of the AIRR to be operated effectively, according to the rehabilitation procedures where the applied motion and force are crucial. The addition of a mechanical-based sensor would be an added advantage to ensure the foot platform will not exceed the required limit independently from the control system.
For the training mode of the AIRR, the addition of the assistive walking rehabilitation training exercises, resistance, and balance would provide a lot of choices in the rehabilitation stages for the patients. Additionally, with teaching and playback, AIRRs can overcome the problem of difficulty to actuate the platform, using offline programming, especially dealing with different sizes of feet. This is important as different sizes of feet patients require a slight adjustment of the movement to make sure their foot is actuating accurately. This has been accomplished as the robot will try to imitate the movement of the ankle of the patient while every movement will be recorded into the system.

Conclusions
The existing AIRRs can provide good alternatives for patients to improve ankle injury rehabilitation, thus encouraging a faster and more effective recovery. Some of the reviewed AIRRs that also provide active ankle rehabilitation exercises and expanded the scope of their ankle rehabilitation procedures using a robot. Likewise, the role of existing AIRRs is crucial to improve the range of motion, balancing, and muscle strengthening of the ankle, although the introduction of the robots will not replace the human touch from the physiotherapist altogether. Instead, this will strengthen the ankle rehabilitation procedures, where the users can have the opportunity to increase the frequency of the treatment.
The implementation of strengths and opportunities from a SWOT analysis, based on the technical point of view can help to elevate the benefits and expectations of AIRRs, from the perspective of internal and external factors. The factors of weaknesses and threats are a challenge to review from the existing bibliography, as some weaknesses and threats are part of the previous research scope. Adding weaknesses and threats will provide an unfair judgment for the researchers who devoted their time, cost, and efforts to make their mark in the AIRR research.
In addition, home rehabilitation plays an important role to increase the frequency of rehabilitation and reduce the role of doctors/therapists to a supervisory role, without the need for constant visits between patients and doctors/therapists. As the review was currently written during the COVID-19 pandemic, the dependency on getting to public buildings, such as clinics and hospitals, can be minimized with the introduction of homebased rehabilitation activities, that can be provided by rehabilitation robots. With this review, the potential benefits of the existing work can help future researchers to improve or to develop a much better AIRR.