- Open Access
Performance adaptive training control strategy for recovering wrist movements in stroke patients: a preliminary, feasibility study
© Masia et al; licensee BioMed Central Ltd. 2009
Received: 24 March 2009
Accepted: 7 December 2009
Published: 7 December 2009
In the last two decades robot training in neuromotor rehabilitation was mainly focused on shoulder-elbow movements. Few devices were designed and clinically tested for training coordinated movements of the wrist, which are crucial for achieving even the basic level of motor competence that is necessary for carrying out ADLs (activities of daily life). Moreover, most systems of robot therapy use point-to-point reaching movements which tend to emphasize the pathological tendency of stroke patients to break down goal-directed movements into a number of jerky sub-movements. For this reason we designed a wrist robot with a range of motion comparable to that of normal subjects and implemented a self-adapting training protocol for tracking smoothly moving targets in order to facilitate the emergence of smoothness in the motor control patterns and maximize the recovery of the normal RoM (range of motion) of the different DoFs (degrees of Freedom).
The IIT-wrist robot is a 3 DoFs light exoskeleton device, with direct-drive of each DoF and a human-like range of motion for Flexion/Extension (FE), Abduction/Adduction (AA) and Pronation/Supination (PS). Subjects were asked to track a variable-frequency oscillating target using only one wrist DoF at time, in such a way to carry out a progressive splinting therapy. The RoM of each DoF was angularly scanned in a staircase-like fashion, from the "easier" to the "more difficult" angular position. An Adaptive Controller evaluated online performance parameters and modulated both the assistance and the difficulty of the task in order to facilitate smoother and more precise motor command patterns.
Three stroke subjects volunteered to participate in a preliminary test session aimed at verify the acceptability of the device and the feasibility of the designed protocol. All of them were able to perform the required task. The wrist active RoM of motion was evaluated for each patient at the beginning and at the end of the test therapy session and the results suggest a positive trend.
The positive outcomes of the preliminary tests motivate the planning of a clinical trial and provide experimental evidence for defining appropriate inclusion/exclusion criteria.
Decreased wrist range of motion (ROM) (flexion and/or extension, abduction/adduction or pronation/supination) after trauma or surgery can be a challenging problem. Physical therapy, orthoses, and additional surgical interventions may not restore the desired functionality even after an intensive rehabilitation program. Therapists spend a considerable amount of practice time in differential diagnosis of these losses and selecting appropriate intervention strategies to restore passive and active motion in concordance with the pathology and to prevent loss of range of motion after injury.
While the regular treatment for wrist stiffness is physical therapy or surgery, researchers are looking for an alternative and more efficient and automatic procedure by means of robotic applications.
Several systems for wrist rehabilitation have been developed in research centres and universities, for example RiceWrist ; MIME ; IMT3 , HWARD ; the Okayama University pneumatic manipulator , and the devices overviewed in [6–9]. The majority are also used for rehabilitation in health centres and hospitals, often coupled with MIT-MANUS , ARMIN , MIME, HapticMaster  and wire-based device from Rosati et. al.  for rehabilitation of proximal limb. Robot assisted therapy are primarily based on goal-directed point-to-point movement involving multiple DoFs ; main purpose is increasing the ROM of the paretic limb in order to regain motor abilities for the Activities of Daily Living (ADL). Contrarily regular physical therapy of wrist rehabilitation consists in a splinting treatment for each single DoF at time, and there have been many studies that look at the splints' effectiveness and what type of splint would be best [15, 16]. Static progressive splinting is a time-honored concept, for more than 20 years, clinicians have recognized the effectiveness of static progressive splints to improve passive range of motion (PROM). Splint designers then sought a means to improve the technique with components that offer infinitely adjustable joint torque control and are easy to apply, lightweight, low-profile, and reasonably priced.
Dynamic splints use some additional component (springs, wires, rubber bands) to mobilize contracted joints [17–19]. This dynamic pull functions to provide a controlled gentle force to the soft tissue over long periods of time, which encourages tissue remodeling without tearing. The issues that make dynamic or static progressive splinting technically difficult include determining how much force to use, how to apply the force, how long to apply the force, and how to prevent added injury to the area. Things could change if the dynamic splinting is delivered using devices which are able to modulate torque delivering and space the range of motion.
Therefore we intend to approach the robotic therapy for wrist rehabilitation using a continuous dynamic splinting of each single DoF but contrarily to the regular progressive splinting we want also to highlight the voluntary component of movement. A performance adaptive control strategy has been developed, with the purpose of providing variable assistance by means of a general training paradigm for stroke patients.
Apparatus: the wrist device
The Wrist-Robot , herewith reported, has been developed at the Italian Institute of Technology with three main requirements: 1) back-drivability of the 3 DoFs (Degree of Freedom), in order to assure a smooth haptic interaction between the robot and the patient; 2) mechanical and electronic modularity, in order to facilitate the future integration into a haptic bimanual arm-wrist-hand system with up to 12 DoFs; 3) scalable software architecture. The Wrist Robot is intended to provide kinesthetic feedback during the training of motor skills or rehabilitation of reaching movements. Motivations for application of robot therapy in rehabilitation of neurological patients come from experimental studies about the practice-induced plastic reorganization of the brain in humans and animal models [21, 22].
The chosen class of mechanical solutions is based on a serial structure, with direct drive by the motors: one motor for pronation/supination, one motor for flexion/extension and two parallel coupled motors for abduction/adduction that allow to balance the pronosupination rotation during motion.
The problem of measurement of arm position is thus reduced to the solution of the device kinematics, with no further transformations required, allowing to actuate the robot to control feedback to a specific human joint, for example to constrain the forearm rotation during wrist rehabilitation, without affecting other joints.
The corresponding rotation axes meet at a single point as shown in figure 1.
The subjects hold a handle connected to the robot and their forearms are constrained by velcros® to a rigid holder in such a way that the biomechanical rotation axes are as close as possible to the robot ones. Unavoidable small joints misalignments are partially reduced by means of a sliding connection between the handle and the robot and the forearm can be moved vertically in order to fit the rotation axis of the pronation/supination DoF. In order to minimize the effect of occasional compensatory shoulder/trunk movements during training exercises, the body is firmly strapped to a robust chair and the chair is positioned in such a way to have the elbow flexed about 90 deg and the hand pointing to the centre of a 21" CD screen, in correspondence with the neutral anatomical orientation of the hand.
sufficient level of the torque at the handle (tab. I)
ROM of the Robot and the Human wrist
Human joint range of motion [deg]
Wrist Device Workspace Capability [deg]
Human Isometric Torque [Nm]
Wrist Device Continuous torque [Nm]
Each DOF is measured by means of a high-resolution encoder (2048 bits/rev) and is actuated by one or two brushless motors, in a direct-drive, back-drivable connection, providing the continuous torque values reported in table 1. The control architecture integrates the wrist controller with a bi-dimensional visual virtual reality environment (VR) for showing to the subjects the actual joint rotation transformation of the hand, the corresponding target direction and two performance indicators defined in the following. The software environment is based on Simulink® and RT-Lab®. The control architecture includes three nested control loops: 1) an inner loop, running at 7 kHz, used by the motor servos; 2) an intermediate loop, running at 1 kHz, for the low level control; 3) a slower loop, running at 100 Hz, for implementing the VR environment and the user interface. The mechanical structure of the wrist robot was designed in such a way to allow a simple and immediate mounting for patients' forearm.
The task is mono-dimensional tracking of a sinusoidally moving target, using one DOF at a time: F/E, Ad/Ab or P/S, respectively; this approach is consistent with the dynamic splinting paradigm which is primarily used to regain the passive ROM after trauma or surgical intervention; the subject aims to move the handle to track the harmonic motion of the target using his/her active ROM; the robot gently intervenes if the subject is not able to actively cover the required angular displacement. Three different experiments were then carried out for the three different DoFs of the wrist. For each experiment, there was one active DoF, which received controlled assistance by the robot, while the two other DoFs were hold by the robot in a small neighbourhood of the neutral position [24–26].
In order to make the task interesting and challenging at the same time, the level of difficulty was managed by the controller modulating two parameters as a function of the performance: a) frequency of the target motion; b) level of the robot assistance. The controller implementation is discussed and illustrated in the next section.
The general control architecture consists of three blocks: 1) target motion generator; 2) force filed generator; 3) performance evaluator.
Each step of the staircase has a duration of 40s plus a 4s rest interval, during which the harmonic motion of the target is stopped as well as the attractive force. For each DoF, the ROM is scanned by the staircase starting from the "easier" to the "more difficult" angular position, taking into account the specific pathological conditions of the treated subjects. In this feasibility study the sequence was, for all the patients, from Flexion to Extension, from Adduction to Abduction, and from Pronation to Supination, respectively. The sequence is ordered "from easy to difficult" considering the hypertonic trend in the range of motion for each trained DoF: 1) the offset angle steps from the easy (more natural and less hypertonic) to the difficult (less natural) joint configuration; 2) the oscillation is modulated from slow (easy) to quick (difficult) frequency.
Growth and decay coefficients of Eq. 9 for each DOF and amplitude oscillation and max/min ROM for each Dof
The different contribution of the force field generator is shown in figure 2 (right).
The assistive control law τ m consists of a non linear elastic field with a parabolic profile (eq. 5). This non linear characteristic was chosen according to the principle of minimal assistance  or also assist as needed : assistance forces/torques should be kept as low as possible in order to promote the emergence of voluntary control. In fact, the chosen pattern of assistance has a less-than-linear increase for small errors, thus facilitating the emergence of active un-aided control at the end of training; for large errors, which are likely to occur at the beginning of training, the assistance grows more than linearly in order to speed up the learning process. The same concept of minimal assistance is used for selecting, in an individual-specific manner, the gain K: it is chosen as the minimum value capable to induce the initiation of movements of the paretic wrist and it was chosen by experimentally observing the active voluntary movements of the participating subjects before starting the rehabilitation protocol.
The equation contains two terms: a raising term with a coefficient α and a decaying term depending on the average angular error F e multiplied by the decay coefficient β. For clarity sake figure 2 shown the entire controller scheme highlighting the different blocks of the controller. There are also two saturation levels that keep the task in a suitable range of difficulty: we chose the range 0.1-1.0 Hz empirically, looking at the performance of the unimpaired subjects. Also the values of α and β for each DoF were experimentally chosen, in order to balance the conflicting requirements of readiness and smoothness and provide a symmetric counterbalance of decaying and raising contributions: these values are listed in table 2.
During the performance of an exercise, when eq. 2 switches the offset ϑ o from one step to the next one, the initial value of eq. 10 is reset to the minimum value of frequency (0.1 Hz). Therefore, the initial target oscillation will be very slow and will smoothly speed-up as a function of the tracking accuracy e = ϑ T - ϑ W , until the end of the step (40s).
Virtual Reality environment
We wanted to strengthen the effectiveness of the system in monitoring wrist use while providing encouragement and reminders throughout a therapy session .
Hence we also display, on the left side of the screen, the instantaneous levels of the two performance indicators by means of height-modulated bars: 1) the level of assistance and 2) the frequency of oscillation. The patients were instructed to minimize the height of the former one while maximizing the height of the latter. This kind of intuitive performance feedback was easily understood by the patients and well appreciated by them.
Max frequency: the maximal frequency that the subject is able to reach, in the possible range 0.1-1 Hz;
Mean assistive torque: the average torque delivered to the patient during the rehabilitation protocol for each DoF
ROM achieved in the single step;
The ROM in the whole session (minimum-maximum degree of movement in the entire exercise).
The active voluntary ROM of the subject holding the passive inactivated device, before and after the exercise in order to compare if the rehabilitation protocol would provide fast benefits even after one therapy session.
Active Range of motion of the subjects pre and post treatment
Maximal frequency reached and average assistive torque
MAXIMAL FREQUENCY REACHED
AVERAGE ASSISTIVE TORQUE
We can also observe that minimal frequency values correspond to the position in which subjects have a reduced range of motion. Moreover, table 5 shows that maximal assistive joint torque is generally provided on the side of the movement of each DoF where the subject is more defective.
The performance of the subjects can also be investigated by comparing the mean speed of the two opposite movements for each DoF in relation with each offset step of the staircase (Figure 5B: F vs. E, Ad vs. Ab, and P vs. S).
exercising movements that are more difficult for him/her, given his specific pathological condition, for example Extension vs. Flexion;
moderating the predominance of pathology-aided behaviours that would enhance Flexion vs. Extension etc.
At last, figure 5C compares, for each DoF, the ROM of the robot target motions (shaded grey band is the amplitude of the target oscillation at different starting position on each DoF workspace) with the actual ROM (bold lines with markers for the two directions of each Dof) exhibited by patient S3 in relation with each offset position. It appears that generally the maximal joint rotation achieved by the patient is asymmetric in the two opposing directions of each DoF (P vs S, F vs. E, Ad vs. Ab) and this is reflected in the pattern of values stored in table 4 of the active range of motion measured by the uncontrolled device at the beginning of protocol. i.e. In spite of the assistance, the subject S3 does not succeed in following the harmonic motion of the target represented by the shaded grey band; he systematically undershoots extension (blue line) and overshoots flexion (red line), whereas the performance is closer to physiological conditions for the two other DoFs.
On the other hand, table 4 reports the active range of motion (uncontrolled device) measured at the end of the training session and the comparison between the part of the table 4 shows a clear increase and symmetrisation before and after the threatment; this result suggests that using robot to generate mobilising splints might be useful to modify the joint stiffness, and reducing hypetonia; even if the total ROM is reduced the symmetry noticeably increases; it is possible the passive component due to hyper tonicity before the splinting added a bias to each joint drifting from the anatomical neutral position.
In the lights of these considerations however we present a preliminary study on the feasibility of using a performance adaptive control strategy combined with a dynamic splinting; in order to strengthen the effectiveness of the proposed approach a wider clinical protocol with higher number of subjects and therapy session is needed.
Although it has been shown in a number of studies that robots can decrease motor impairment after stroke with certain advantages, less emphasis to date has been put on robotic developments for the hand and on corresponding preliminary clinical studies. A notable exception is the work by Takahashi et al.  who reported the use of the pneumatic-actuated HWARD wrist robot with 13 patients. The main difference of HWARD with respect to the Wrist robot (here with reported) is related to the wrist movements: HWARD can only operate with F/E whereas Wrist Robot can operate equally well with Ab/Ad and P/S.
In this preliminary experiment investigating patients, only one joint DoF was exercised at a time. The procedure simulated as much as possible the use of splints widely used in clinical applications. However, there is no hardware or software limitation to design 2D and 3D experiments, which indeed are planned and will be carried out in the near future.
We wish to emphasize that our control system is based of a principle of minimal assistance that focuses on the initiation of the movement; on the contrary most of the other rehabilitation robots, focuses on the termination phase (goal directed movements), by forcing the patient to complete the movements if he/she is unable to achieve the target. We also plan to integrate in the robot an active finger F/E unit, by means of a motorized handle  to study the impact of single-DoF rehabilitation protocol on cylindrical grasping and compare the effectiveness of different rehabilitation strategies that include distal and/or proximal limb.
The results reported in this single-session study show that the proposed adaptive control strategy is robust, in terms of patient response, is well accepted by the subjects and the control architecture is capable to smoothly adapt to the specific impairments of the patients without needing a fine customization of the controller gains for each subject; this controller robustness allows to introduce the system in the clinical application providing a user friendly interface for users and patients, and to deliver an automatic execution of the therapy sessions.
The results of the presented preliminary work shows that robotic therapy may improve motivations in patients and provide tangible results even in a short term experience. The technological approach with the use of customized devices may strengthen the potentials of the regular physical therapy in delivering assistance and training. The proposed controller strategy is simply based on an automation of the well established methodology of dynamic splinting; this kind of approach can result familiar to the medical staff allowing technology to progressively take part to the emerging and increasing needs of rehabilitation, without shocking the entrenched application of regular therapy. It remains to be investigated, as we plan to do in a systematic clinical trial, to which extent a suitable protocol can induce permanent improvements in the neural control of wrist movements, necessary for any attempt to achieve functional gains in the activities of daily life.
Acknowledgements: these work was carried out at Human Behaviour Laboratory of Italian Institute of Technology and it was supported by a grant of Italian Ministry of Scientific Research and Ministry of Economy. This work is partly supported by the EU grants FP7-ICT-271724 HUMOUR and FP7-ICT-2007-3 VIACTORS.
- Gupta A, O'Malley MK, Patoglu V, Burgar C: Design, Control and Performance of RiceWist: A Force Feedback Exoskeleton for Wrist Rehabilitation and Training. The Intl J of Robotics Res 2008,27(1):233-251. 10.1177/0278364907084261View ArticleGoogle Scholar
- Shor PC, Lum PS, Burgar CG, Loos HFM, Majmundar M, Yap R: The Effect of Robotic-Aided Therapy on Upper Extremity Joint Passive Range of Motion Pain", Proc of Intl Conf on Rehab Robotics, Integration of Assisted Technol in the Information Age. Edited by: Mounir Mokhtari. IOS Press; 2001:79-83.Google Scholar
- Krebs HI, Volpe BT, Williams D, Celestino J, Charles SK, Lynch D, Hogan N: Robot-Aided Rehabilitation: A Robot for Wrist Rehabilitation. IEEE Trans on Neural Systems and Rehab Engineering 2007, 5: 327-335. 10.1109/TNSRE.2007.903899View ArticleGoogle Scholar
- Takahashi CD, Der-Yeghiaian L, Le V, Motiwala RR, Cramer SC: Robot-Based Hand Motor Therapy After Stokes. Brain 2008, 131: 425-437. 10.1093/brain/awm311View ArticlePubMedGoogle Scholar
- Takaiwa M, Noritsugu T: Development of Wrist Rehabilitation Equipment Using Pneumatic Parallel Manipulator. Proceeding of the 2005 IEEE International Conference of Robotics and Automation, Barcelona 2005, 2302-2307. full_textView ArticleGoogle Scholar
- Lambercy O, Dovat L, Gassert R, Burdet E, Teo CL, Milner TE: A Haptic Knob for Rehabilitation of Hand Function. IEEE Transactions on Neural Systems and Rehabilitation Engineering (TNSRE) 2007,15(3):356-366. 10.1109/TNSRE.2007.903913View ArticleGoogle Scholar
- Hesse S, Schulte-Tigges G, Konrad M, Bardeleben A, Werner C: Robot-assisted arm trainer for the passive and active practice of bilateral forearm and wrist movements in hemiparetic subjects. Arch Phys Med Rehabil 2003,84(6):915-20. 10.1016/S0003-9993(02)04954-7View ArticlePubMedGoogle Scholar
- Loureiro RCV, Collin CF, Harwin WS: Robot Aided Therapy: Challenges Ahead for Upper Limb Stoke Rehabilitation. Proceed of Intl Conf on Disability, Virtual Reality and Assoc Tech 2004, 33-39.Google Scholar
- Jack D, Boian R, Merians AS, Tremaine M, Burdea GC, Adamovich S, Recce M, Poizner H: Virtual Reality-Enhanced Stroke Rehabilitation. IEEE Trans on Neural Syst and Rehab Engineer 2001, 9: 308-318. 10.1109/7333.948460View ArticleGoogle Scholar
- Krebs HI, Ferraro M, Buerger SP, Newbery MJ, Makiyama A, Sandmann M, Lynch D, Volpe BT, Hogan N: Rehabilitation robotics: pilot trial of a spatial extension for MIT-Manus. Journal of NeuroEngineering and Rehabilitation 2004, 1: 5. 10.1186/1743-0003-1-5PubMed CentralView ArticlePubMedGoogle Scholar
- Nef T, Mihelj M, Kiefer G, Perndl C, Muller R, Riener R: ARMin - Exoskeleton for Arm Therapy in Stroke Patients. Proceedings of the IEEE 10th International Conference on Rehabilitation Robotics, 13-15 June 2007, Noordwijk, The Netherlands 68-74.Google Scholar
- Loureiro R, Harwin W: Reach & Grasp therapy: Design and control of a 9 DoF Robotic neuro-rehabilitation system. IEEE 10th Int Conf Rehab Robot., 13-15 June 2007, Noordwijk, The Netherlands 2007, 68-74.Google Scholar
- Rosati G, Gallina P, Masiero S: Design, Implementation and Clinical Tests of a Wire-Based Robot for Neurorehabilitation. IEEE Transactions on Neural Systems and Rehabilitation Engineering 2007,15(4):560-569. 10.1109/TNSRE.2007.908560View ArticlePubMedGoogle Scholar
- Timmermans AAA, Seelen HAM, Willmann RD, Kingma H: Technology-assisted training of arm-hand skills in stroke: concepts on reacquisition of motor control and therapist guidelines for rehabilitation technology design. Journal of NeuroEngineering and Rehabilitation 2009, 6: 1. 10.1186/1743-0003-6-1PubMed CentralView ArticlePubMedGoogle Scholar
- McGrath M, Ulrich S, Bonutti P, Smith J, Seyler T, Mont M: Evaluation of Static Progressive Stretch for the Treatment of Wrist Stiffness. The Journal of Hand Surgery 2008,33(9):1498-1504. 10.1016/j.jhsa.2008.05.018View ArticlePubMedGoogle Scholar
- Doornberg JN, Ring D, Jupiter JB: Static progressive splinting for posttraumatic elbow stiffness. J Orthop Trauma 2006, 20: 400-404. 10.1097/00005131-200607000-00006View ArticlePubMedGoogle Scholar
- McPherson JJ, Becker AH: Dynamic Splint to Reduce the Passive Component of Hypertonicity. Archives of Physical Medicine and Rehabilitation 1985, 66: 249-252. 10.1016/0003-9993(85)90162-5View ArticlePubMedGoogle Scholar
- Scheker LR, Chesher SP, Netscher DT, Julliard KN, O'neill WL: Functional Results of Dynamic Splinting after Transmetacarpal, Wrist, and Distal Forearm Replantation. Journal of Hand Surgery 1995, 20: 584-590.View ArticleGoogle Scholar
- Coverdale J: Does a Uni-directional Dynamic Splint Affect Bi-directional Wrist ROM? American Society of Hand Therapists Annual Meeting, September 21, 20002Google Scholar
- Masia L, Casadio M, Morasso P, Pozzo T, Sandini G: Using a Wrist Robot for Evaluating how Human Operators learn to perform Pointing Movements to a Rotating Frame of Reference. Proceed IEEE BioRob October 19-22 2008, Scottsdale, AZ, USA 2008.Google Scholar
- Nudo RJ: Mechanisms for recovery of motor function following cortical damage. Current Opinion in Neurobiology 2006, 16: 638-644. 10.1016/j.conb.2006.10.004View ArticlePubMedGoogle Scholar
- Tna H, Srinivan B, Eberman B, Cheng B: Human factors for the design of a force reflecting haptic interface. Dynamic Syst Control 1994,55(1):353-359.Google Scholar
- Wolbrecht ET, Chan V, Reinkensmeyer DJ, Bobrow JE: Optimizing Compliant, Model-Based Robotic Assistance to Promote Neurorehabilitation. IEEE Trans Neural Syst Rehabil Engineer 2008, 16: 286-97. 10.1109/TNSRE.2008.918389View ArticleGoogle Scholar
- Basics of wrist rehabilitation protocols[http://rehabworks.ksc.nasa.gov/education/protocols/basicwristelbow.php]
- Bridger Robert: Introduction to Ergonomics. CRC 3rd edition. 2008.Google Scholar
- Gavriel Salvendy: Handbook of Human Factors and Ergonomics. Third edition. J Wiley; 1997.Google Scholar
- Casadio M, Giannoni P, Morasso P, Sanguineti V: A proof of concept study for the integration of robot therapy with physiotherapy in the treatment of stroke patients. Clinical Rehab 2009, 23: 217-228. 10.1177/0269215508096759View ArticleGoogle Scholar
- Emken JL, Benitez R, Sideris A, Bobrow JE, Reinkensmeyer DJ: Motor adaptation as a greedy optimization of error and effort. J Neurophysiol 2007,97(5):3997-4006. 10.1152/jn.01095.2006View ArticlePubMedGoogle Scholar
- Colombo R, Pisano F, Mazzone A, Delconte C, Micera S, Carrozza MC, Dario P, Minuco G: Design strategies to improve patient motivation during robot- aided rehabilitation. Journal of NeuroEngineering and Rehabilitation 2007, 4: 3. 10.1186/1743-0003-4-3PubMed CentralView ArticlePubMedGoogle Scholar
- Masia L, Krebs HI, Cappa P, Hogan N: Design and characterization of hand module for whole-arm rehabilitation following stroke. IEEE/ASME Trans Mechatronics 2007,12(4):399-407. 10.1109/TMECH.2007.901928View ArticlePubMedGoogle Scholar
This article is published under license to BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.