(e-book) Rehabilitation Robotics (by Sashi S _部分12

更新时间:2023-06-07 01:16:01 阅读量: 实用文档 文档下载

说明:文章内容仅供预览,部分内容可能不全。下载后的文档,内容与下面显示的完全一致。下载之前请确认下面内容是否您想要的,是否完整无缺。

康复机器人相关的论文,拆分成了13份文档,都很不错的。

The ALLADIN Diagnostic Device:

An Innovative Platform for Assessing Post-Stroke Functional Recovery 541 Since April 2004 a complete product des ign and development cycle which included the computer aided design, the development of three early prototypes and the feedback from the tes ting were implemented. Refinement and detailing of the conceptual des ign was a natural result of this cyclic process. The eight 6-axis F/T sensors are respectively installed behind the trunk, below the pos terior, at the affected lower arm, at the affected thumb, index and middle finger, at the affected foot and toe (Figure 2). They output detailed data on the ADL tas ks to be performed. Table 5 s hows the bas ic characteris tics of the 6-axis F/T s ens ors (50M31A-I25, 67M25A-I40, 90M40A-I50, 90M40A-I50, 90M40S-I50, 90M40A-I50, 50M31A-I25; JR3 Inc., Woodland, USA). The orthogonal reference frame for the force and torque vectors is located inside the sensor. The platform has three positional settings for the patient according to the tasks to be performed. The first operational pos ition is as s ociated to the ADL tas k #1, ADL tas k #2 and ADL tas k #3, a second position is selected for the ADL task #4 task and a third for the ADL task #5 and ADL task #6. All operating instructions are presented on a screen in front of the patient. A first instruction is the video presentation of the task to be initiated by the patient; the second instruction is an invitation to “memorize the task” and then to “execute it”. The measured behaviour is the combined output of 48 channels repres enting the x, y, z F/T data for all eight F/T sensors.

The diagnostic device includes the following main units (Figure 2):

1.Accessory storage board

2.Transit lying wheelchair

3.Monitor for the patient

4.Podium

5.Trunk Device

6.Foot Device

7.Arm Device

8.Finger Device

9.Seat Device

The Arm Device, the Finger Device and the Foot Device are s hown in Figure 3. A customized software has been developed in order to allow the recording of different types of data: F/T data, clinical cale and natural language de cription made by the phy s iotherapi s t s (See s ub s ection 3.5). Several young volunteer s participated in a preliminary testing that aimed at verifying the output of the proposed isometric procedure. Altogether 250 s ubjects (150 hemiplegic patients and 120 normal control) were recruited during the clinical trials at the three hospitals. The centres participating in the multi-centre clinical trials were:

x

Algemeen Ziekenhuis Maria Middelares Sint-Jozef Hos pital (AZMMSJ), Gent, Belgium, x

Adelaide & Meath Hospital (AMNCH), Tallaght, Dublin, Ireland,

x Szent János Hospital, Budapest, Hungary.

All the three clinical trial centres obtained the approval of the relevant ethics committees. An informed consent was obtained from the subjects participating the clinical trials.

Subjects were measured and assessed twice a week during the first two months period and once a week during four cons ecutive months . They were s eated in a s pecial des igned wheelchair and driven into an anthropometrical adaptive mea s

uring in s trument characterized by the above mentioned three discrete positions (Small, Medium and Large).

康复机器人相关的论文,拆分成了13份文档,都很不错的。

542Rehabilitation Robotics Appropriate size accessories and device settings were also used to ensure that the error in the anatomical angles is minimal, as well as to keep the handling complexity of the diagnos tic device on a tolerable level for the operating phys iotherapis t (See the acces s ory

storage board in Figure 2).

Fig. 2. The components of the Alladin diagnostic device. Description Quantity Lateral forces (Fx, Fy) Axial force (Fz)Torques (Tx,Ty, Tz)

Dimensions Type-H(and) 3 150 N 300 N 8 Nm Ø 50 x 31 mm Type-A(rm) 1 150 N 200 N 10 Nm Ø 67 x 35 mm Type-B(ack) 1 250 N 250 N 20 Nm Ø 90 x 40 mm Type-P(osterior) 1 550 N 1100 N 50 Nm Ø 114 x 40 mm Type-F(oot) 1 400 N 800 N 25 Nm Ø 90 x 40 mm Type-T(oe) 1 150 N 300 N 8 Nm Ø 50 x 31 mm Table 5. Basic characteristics of the 6-axis F/T sensors.

The po ture cho en for the mea urement repre ent a trade-off between a good approximation of the natural posture and the anthropometric characteristics of the subject. This choice assures sufficient conditions of repeatability to the measurements.

The clinical a s s e s s ment wa s performed through the Fugl-Meyer Scale (Lindmark adaptation), the Motor Assessment Scale and the Stroke Impact Scale. The physiotherapists us ed a Portable Digital As s is tant (PDA) in order to record the s core for each as s es s ment scale and patients’ functional recovery.

康复机器人相关的论文,拆分成了13份文档,都很不错的。

The ALLADIN Diagnostic Device:

An Innovative Platform for Assessing Post-Stroke Functional Recovery 543 The aim of the s tudy is to identify if there are s ignificant links between the recovery that occurs pos t s troke as meas ured by the diagnos tic device and this recovery as measured by clinical scales and natural language descriptions. As described in section

3.2,

s ix different ADL ta

s

k

s

with a varying complexity were u

s

ed for F/T

measurements.

The data acquisition followed a detailed protocol (Van Vaerenbergh et al., 2004). For each tas k, the patient watches a video s howing the movement (recording #1). Secondly he is as ked to mentally imagine and reproducing it with open eyes (recording #2). Finally, for three times he actually tries to perform it, exerting the forces at a comfortable level (recording #3, recording #4 and recording #5).

Fig. 3. Devices for measurements on the upper limbs and lower limbs. Left: the Arm Device (7) and the Finger Device (8). Right: the Foot Device (6).

3.5 The ALLADIN soft w are

This s ection pres ents in detail the functional and technical s pecifications and the des ign approach of the software of the diagnostic device and describes the implementation of the different software modules as well as for their integration.

A general architecture of the diagnostic device software has been defined according to the functional s pecifications of the diagnos tic device defined reported in previous s ection, and als o taking into account the additional information provided through a close collaboration with end users (i.e., clinicians and physiotherapists) on this specific topic (Figure 4).

Specifications of the Databas e Module were given us ing the UML (Cantor, 1998) notation and diagrams in order to provide a definition of the functionality of this module which can be eas ily interpreted both by the s oftware developers and by the clinicians. UML notation was als o us ed to define the interface between the Cover Application (CA) module and the Databas e (DB) module. All other modules, i.e. Data acquisition (DAQ), Data viewer, Automatic Speech Recognition (ASR), were described by us ing s impler notations, s uch as flowcharts or direct pres entation of the low level functions definition.

康复机器人相关的论文,拆分成了13份文档,都很不错的。

544Rehabilitation Robotics

Fig. 4. The ALLADIN Software Architecture.

The s oftware has been bas ed on the general us er requirements and s pecifications which have been defined in a preliminary phas e. The s oftware allows to manage all the functionalities provided by the ADD, including the recording and exchange of different kind of data between the CA module and the other modules and between the CA module and the DB. The main data to be collected and managed are:

x Patient data and case history

x Standard Outcome Measure (SOM)

x Natural language descriptions of the patient's status

x Voice records of the descriptions

x F/T measurement records of the ADL tasks

All data, after having been collected, are uploaded to the Local Database (Figure 4).

The main s oftware requirements and s pecifications formed the framework for the Cover Application software design and development.

The CA was implemented using the Microsoft Visual Basic (VB) Environment Release 6.0, which allows the creation of friendly graphical user interfaces and simplifies the integration of modules developed with heterogeneous techniques. In particular it provides a means to connect the CA module with the ALLADIN database and with the other ADD modules, as for instance, the dynamic link libraries (DLLs) which implements the DAQ module and the Data Viewer Module. The DB was implemented in Micros oft Acces s 2000 and the other s oftware modules were developed according to the CA module s pecifications in terms of I/O interfaces and functionality. The CA module allows the us er to create, retrieve and modify records by queries on the patient information and clinical assessment in the DB.

The CA module was developed in order to allow the us er to record the different information related to each patient and the data which come out from the measurements.

康复机器人相关的论文,拆分成了13份文档,都很不错的。

The ALLADIN Diagnostic Device:

An Innovative Platform for Assessing Post-Stroke Functional Recovery 545 The main menu (Figure 5), offers different functionalities, s uch opening a patient record, starting a new session of measurements, creating a new patient record, editing and creating an user's profile, synchronizing with the global DB, system settings adjustment and remote assistance.

Four different types of us ers were identified (ADD phys iotherapis t, Natural language physiotherapist, Principal Investigator and System administrator); for each user profile an access rights policy was defined.

Fig. 5. The Cover Application main window.

4. Results and discussion

Some preliminary res ults from a normal control s ubject and a pathological s ubject are here presented. The choice of the task and the sensors is based on the preliminary results of data mining algorithms applied to the pre-processed data (Van Djick et al. 2006). Let’s consider the task“Drinking” in a normal control (AHS-028, male, 45 years old, right dominant hand, measurement of the left side) and in a pathological subject (AHS-064, male, 43 years old, right dominant hand, right s ide of hemipares is, date of s troke 15/12/2005, meas urement on the right side), 25 days and 131 days following the stroke onset. The number of samples from force measurements shown in Figure 6-8 is 5400: as already stated, the sample frequency for data acqui ition i 100 Hz, therefore the ta k la t 5.4 econd. Figure 6 how the force measurements from the thumb in the normal control (top plot), in the hemiplegic patient, 25 days following the stroke onset (middle plot) and 131 days following the stroke onset.

康复机器人相关的论文,拆分成了13份文档,都很不错的。

546Rehabilitation Robotics

Fig. 6. Force measurements from the thumb in a normal control subject (top), in a hemiplegic patient, 25 days following the stroke event (middle) and 131 days following the stroke event (bottom) for the task “Drinking“.

康复机器人相关的论文,拆分成了13份文档,都很不错的。

The ALLADIN Diagnostic Device:

An Innovative Platform for Assessing Post-Stroke Functional Recovery 547

Fig. 7. Force meas urements from the index finger in a normal control s ubject (top), in a hemiplegic patient, 25 days following the stroke event (middle) and 131 days following the stroke event (bottom) for the task “Drinking“.

康复机器人相关的论文,拆分成了13份文档,都很不错的。

548Rehabilitation Robotics

Fig. 8. Force meas urements from the middle finger in a normal control s ubject (top), in a hemiplegic patient, 25 days following the stroke event (middle) and 131 days following the stroke event (bottom) for the task “Drinking“.

康复机器人相关的论文,拆分成了13份文档,都很不错的。

The ALLADIN Diagnostic Device:

An Innovative Platform for Assessing Post-Stroke Functional Recovery 549 In the normal control, pos itive values in the F x-direction can be obs erved. In normal circums tances, for a gras ping movement, the thumb will be brought to the point where the index and the middle finger touch each other. In the diagnostic device, the thumb is fixated on the same height as the index finger. This causes a downwards movement of the thumb when the s ubject gras ps the glas s to drink. The pos itive values in F y-direction means that the subject moves the thumb forwards when he positions the fingers around the glass to drink. The positive values for F z-direction points out that the subject grasps the glass to drink.

The force measurement from the thumb recorded 25 days following the stroke onset show negative values on the x-axis: the force is directed in the opposite direction than the motor performance in the normal control s ubject, pointing out that the s ubject moves the thumb upwards to bring the glas s to the mouth, ins tead of moving downwards. The negative values obs erved along the F y-direction mean that the s ubject pus hes the thumb forwards to bring the glas s to the mouth. The pos itive values along the z-axis allow to conclude that the subject tries and grasps the glass to drink. The force is exerted in advance than the normal control subject and it lasts till to the end of the attempt.

The force meas urement from the thumb recorded 131 days following the s troke ons et s how the pos itive values on the x-axis, s ame direction as the motor performance in the normal control s ubject, meaning that the s ubject moves the thumb downwards to bring the glas s to the mouth, even if the force reaches the zero at about the half of the meas urement, before s till ris ing to pos itive values. The negative values obs erved along the y-axis point out that the subject pushes the thumb forwards to bring the glass to the mouth. The pos itive values along the z-axis mean that the s ubject tries and gras ps the glass to drink. The force is exerted in advance than the normal control subject and it lasts till to the end of the attempt.

Figure 7 s how the force meas urements from the index finger in the normal control (top plot), in the hemiplegic patient, 25 days following the s troke ons et (middle plot) and 131 days following the stroke onset.

In the normal control subject, the negative values along with the x-direction point out that the s ubject moves the index upwards to bring the glas s to the mouth, while the pos itive values along with the y-axis points out that the subject moves the index backwards to bring the glas s to the mouth. The pos itive values along with the z-axis means that the s ubject grasps the glass to drink.

The force measurement from the thumb recorded 25 days following the stroke onset show negative values along with the x-direction: the s ubject moves the index in the s ame direction as the normal control subject (upwards) to bring the glass to the mouth, even if with lower values and for a larger range of time. The pos itive, negative and pos itive values along with the y-axis points out that the s ubject firs t moves the index forwards, backwards and then forwards to bring the glass to the mouth. The positive values along with the z-axis means that the subject grasps the glass to drink, even if the exerted force doesn’t shows a bell shaped form, as in the normal control subject. It shows a peak and then a falling to a lower value.

The force measurement from the thumb recorded 131 days following the stroke onset show negative values along with the x-direction: the subject moves the index in the same direction as the normal control s ubject (upwards) to bring the glas s to the mouth, approaching the

康复机器人相关的论文,拆分成了13份文档,都很不错的。

550Rehabilitation Robotics force values of the normal control s ubject even if for a larger range of time. The rather pos itive values along with the y-axis points out that the s ubject firs t moves the index forwards to bring the glas s to the mouth, with a les s fragmented trend, compared to the measurement recorded 25 days following the stroke onset. The positive values along with the z-axis means that the subject grasps the glass to drink, with a quite bell shaped form, as in the normal control subject.

Figure 8 show the force measurements from the middle finger in the normal control (top plot), in the hemiplegic patient, 25 days following the stroke onset (middle plot) and 131 days following the s troke ons et. In the normal control, the negative values on the x-axis means that the s ubject moves the middle finger upwards to bring the glas s to the mouth. The pos itive values obs erved along the F y-direction point out that the s ubject pulls the middle finger backwards to bring the glas s to the mouth. The positive values along the F z-direction means that the subject grasps the glass to drink.

The negative values on the x-axis recorded from the hemiplegic patient, 25 days following the stroke onset means that the subject moves the middle finger upwards to bring the glass to the mouth, showing lower amplitude values than the normal control. The negative values observed along the F y-direction and the positive values along the F z-direction mean that the s ubject pulls the middle finger backwards to bring the glas s to the mouth and gras ps the glass to drink respectively, but showing a fragmented trend. The negative values on the x-axis from the middle finger force meas urements recorded 131 days following the s troke ons et means that the s ubject moves the middle finger upwards to bring the glas s to the mouth, s howing lower amplitude values and a more fragmented trend than the normal control.

The negative values obs erved along the F y-direction and the pos itive values along the F z-direction point out that the subject pulls the middle finger backwards to bring the glass to the mouth and grasps the glass to drink respectively, with a more regular trend if compared to the force measurements recorded 25 days following the stroke onset.

5. Future directions and conclusions

A large set of features characterizing the clinical recovery have been extracted from the data according to the preliminary results from data mining techniques (Van Djick et al., 2006b) in order to track the recovery proces s through miles tones and to fores ee the rehabilitation outcome through predictive markers.

The pos itive res ults obtained s o far through the extens ive us e of the propos ed diagnos tic device during the clinical trials for functional as s es s ment of pos t-s troke patients allow to foresee new possible scenarios for the neurorehabilitation domain. The use of the diagnostic device together with systems for brain imaging (PET, fMRI, MEG) and techniques for monitoring brain activity (EEG) will allow to monitor the degree of learning and the change in motor performance induced by the rehabilitative treatments through traditional and robotic therapies. Alternative applications for the proposed platform are:

x isometric motor exercise. Many clinical protocols for motor therapy of different type of patients pres cribe the execution of is ometric exercis es. The propos ed

y tem could allow to accurately tune, monitor, mea ure and record the

康复机器人相关的论文,拆分成了13份文档,都很不错的。

The ALLADIN Diagnostic Device:

An Innovative Platform for Assessing Post-Stroke Functional Recovery 551 forces/torques exerted by the patient during such exercises. To this aim, a self-

calibration routine will be added to the system, such that forces/torques due to

the body’s weight will be automatically s et to zero at the s tart of the motor therapy;

x human-machine interface. The propos ed s ys tem can be as s ociated to a virtual reality environment for motor rehabilitation, as recently implemented with similar

devices for isometric measurements in the upper limb (Kurillo et al., 2005) or it can

be used as novel human-machine interface for many different applications where

the us e of the hand and the foot is required, e.g. pedal and handle interfaces for

game, s urgical robots, vehicles for enabling independent living to citizens with residual abilities.

Further possible developments comprise also:

x a migration to a mechatronic platform embedding actuator

s to produce

perturbations and assisted constrained motion of the affected limbs;

x the application of the propos ed platform for res earch in neuros cience, e.g. by comparing is ometric performance of healthy controls and different patients, and

for studying anticipative and high-level planning capabilities based on the study of

whole-body dynamics in is ometric conditions at the inception of voluntary movements.

The proposed platform is the first device which acquires a great deal of different data (F/T data, clinical scales, natural language descriptions made by the physiotherapists) till now. It is a versatile research tool, which records heterogeneous fields in a complete and detailed way. It has been us ed in clinical trials in order to verify the clinical hypotheses.

The propos ed platform, which has been validated in three different clinical centres in Europe, proved to be effective as a tool for experimental use in novel functional assessment procedures of post-stroke patients, according to the original specifications provided by the medical doctor and therapi t. The platform ha al o a range of other potential applications, from motor therapy to human-machine interface.

The reduction of its level of complexity and the development of an optimized version for clinical uses are the next steps, after the completion of the analysis of data collected during the clinical trials. The use of actuators in the platform can be also considered as a further development of the platform as tool both for functional as s es s ment and rehabilitative treatment.

6. Ackno w ledgment

This work was partly supported by the European Commission - 6th Framework Programme under the grant N. 507424 (ALLADIN – Natural Language Bas ed Decis ion Support in Neuro-rehabilitation).

The ALLADIN project is co-ordinated by Jo Van Vaerenbergh, Arteveldehogeschool (Gent, Belgium). The other partners of the ALLADIN project are: Language and Computing NV (Belgium), Budape t Univer ity of Technology and Economic (Hungary), School of Electrical Engineering of the Univers ity of Ljubljana (Slovenia), Zenon SA Robotics and Informatics (Greece), Multitel ASBL (Belgium), Trinity College Dublin (Ireland), National

康复机器人相关的论文,拆分成了13份文档,都很不错的。

552Rehabilitation Robotics Ins titute for Medical Rehabilitation (Hungary), Scuola Superiore Sant’Anna (Italy) and Campus Bio-Medico University (Italy).

The authors greatly acknowledge the inputs received from all the people participating in the ALLADIN project in many project meetings and internal working groups activities.

7. References

Aisen, M. L.; Krebs, H. I.; Hogan, N.; McDowell, F. & Volpe, B. T. (1997). The effect of robot as s is ted therapy and rehabilitative training on motor recovery following s troke.

Arch Neurol, Vol. 54, No. 4, April 2000, pp. 443-446.

Bobath, B. (1978). Adult hemiplegia: ev aluation and treatment, William Heinemann Medical Books, London.

Bozec, S. L.; Goutal, L. & Bouisset, S. (1997), Dynamic postural adjustments associated with the development of is ometric forces in s itting s ubjects, C R Acad Sci III, Vol. 320,

No. 9, pp. 715-720.

Bruns trom, S. (1970). Mov ement therapy in hemiplegia: a neurophysiological approach, Harper and Row, New York.

Buchanan, T. S. & Shreeve, D. A. (1996). An evaluation of optimization techniques for the prediction of mus cle activation patterns during is ometric tas ks. J. Biomech. Eng,

Vol. 118, No. 4, November 1996, pp. 565-574.

Cantor, M. R. (1998), Object-Oriented Project Management with UML, John Wiley & Sons, New York.

Carr, J. & Shepard, R. (1998). Neurological rehabilitation, Butterworth-Heineman, Oxford. Clark, S.; Tremblay, F. & Ste-Marie, D. (2004). Differential modulation of corticos pinal excitability during obs ervation, mental imagery and imitation of hand actions.

Neuropsychologia, Vol. 42, No. 1, pp. 105-112.

Dechent, P.; Merboldt K. D. & Frahm J. (2004). Is the human primary motor cortex involved in motor imagery? Brain Res Cogn Brain Res, Vol. 19, No. 2, April 2004,

pp. 138-144.

Deuts ch, J. E.; Merians, A. S.; Adamovich, S.; Poizner, H. & Burdea, G. C. (2004), Development and application of virtual reality technology to improve hand us e

and gait of individuals post-stroke, Restor Neurol Neurosci, Vol. 22, No. 3-5, pp. 371-

386.

Dewald, J. P.; Pope, P. S.; Given, J. D.; Buchanan, T. S. & Rymer, W. Z. (1995). Abnormal mus cle coactivation patterns during is ometric torque generation at the elbow

and shoulder in hemiparetic subjects. Brain, Vol. 118, Pt 2, , April 1995, pp. 495-

510.

Ehrs s on, H. H.; Geyer, S. & Naito, E. (2003). Imagery of voluntary movement of fingers, toes, and tongue activates corresponding body-part-specific motor representations.

J Neurophysiol, Vol. 90, No. 5, November 2003, pp. 3304-3316.

Jacks on, P. L.; Lafleur, M. F.; Malouin, F.; Richards, C. L. & Doyon J. (2003). Functional cerebral reorganization following motor equence learning through mental practice with motor imagery. Neuroimage, Vol. 20, No. 2, October 2003, pp. 1171-

1180.

康复机器人相关的论文,拆分成了13份文档,都很不错的。

The ALLADIN Diagnostic Device:

An Innovative Platform for Assessing Post-Stroke Functional Recovery 553 Johnson-Frey, S. H. (2004). Stimulation through simulation? Motor imagery and functional reorganization in hemiplegic stroke patients. Brain Cogn, Vol. 55, No. 2, July 2004,

pp. 328-331.

Kilner J. M.; Paulignan, Y. & Boussaoud, D. (2004). Functional connectivity during real vs imagined visuomotor tasks: an EEG study. Neuroreport, Vol. 15, No. 4, March 2004,

pp. 637-642.

Koo, T. K.; Mak, A. F.; Hung, L. & Dewald, J. P. (2003). Joint pos ition dependence of weaknes s during maximum is ometric voluntary contractions in s ubjects with

hemiparesis. Arch Phys Med Rehabil, Vol. 84, No. 9, September 2003, pp. 1380-

1386.

Krebs H. I.; Hogan N.; Aisen M. L. & Volpe B. T. (1998). Robot-aided neurorehabilitation.

IEEE Trans Rehabil Eng, Vol. 6, No. 1, March 1998, pp. 75-87.

Kurillo, G.; Mihelj, M.; Munih, M. & Bajd, T. (2005). Gras ping and manipulation in virtual environment us ing 3by6 finger device, Proceedings of the 9th IEEE

International Conference on Rehabilitation Robotics, pp. 131-134, Chicago, IL, USA,

2005.

Lehericy, S.; Gerardin, E.; Poline, J. B.; Meunier, S.; Van de Moortele, P. F.; Le Bihan, D. & Vidailhet, M. (2004). Motor execution and imagination networks in post-stroke dystonia. Neuroreport, Vol. 15, No. 12, August 2004,

pp. 1887-1890.

Perfetti C. (1997). Der hemiplegische Patient. Kognitiv-therapeutische Ubungen. Richard Pflaum Verlag GmbH & Co, München.

Peebles, L. & Norris, B. J. (1998) Adultdata. The handbook of adult anthropometrical and strength measurements - Data for design safety, Department of Trade and Indus try, London,

UK.

Reinkens meyer, D. J.; Hogan, N.; Krebs, H. I.; Lehman, S. L. & Lum, P. S. (2000).

Rehabilitators, robots, and guides: new tools for neurological rehabilitation, biomechanics

and neural control of posture and movement, J. Winters and P. Crago, Springer-Verlag,

pp. 516-533.

Sejnowski, T. J. (1998). Making smooth moves. Nature, Vol. 394, No. 6695, August 1998, pp.

725-26.

Van Dijck, G.; Van Hulle, M. M. & Van Vaerenbergh, J. (2006). Statistically Rigorous Human Movement Onset Detection with the Maximal Information Redundancy Criterion,

Proceedings of 28th IEEE EMBS Annual International Conference, pp. 2474-2477, New

York City, USA, August 2006.

Van Dijck, G.; Van Hulle, M. M. & Van Vaerenbergh, J. (2006). Hybrid feature s ubs et s election for the quantitative as s es s ment of s kills of s troke patients in activity of

daily living tas ks, Proceedings of 28th IEEE EMBS Annual International Conference,

pp. 5699-5703, New York City, USA, August 2006.

Van Vaerenbergh, J. et al. (2004), Deliverable D1.1: Methodology for multi centre trial, ALLADIN Project, IST-2002-507424.

Walsh, S.; Saltzman, C.; Talbot, K.; Aper, R. & Brown, T. (1996), In vivo validation of in vitro tes ting of hallucal flexor mechanics, Clin Biomech, Vol. 11, No. 6, pp. 328-

332.

康复机器人相关的论文,拆分成了13份文档,都很不错的。

554Rehabilitation Robotics Woldag, H. & Hummels heim, H. (2002). Evidence-bas ed phys iotherapeutic concepts for improving arm and hand function in stroke patients: a review. J.Neurol, Vol. 249, No. 5, May 2002, pp. 518-528.

Wolpert, D. M.; Ghahramani, Z. & Flanagan, J. R. (2001). Pers pectives and problems in motor learning. Trends Cognitive Sciences, Vol. 5, No. 11, November 2001, pp. 487-

494.

康复机器人相关的论文,拆分成了13份文档,都很不错的。

30 Synthesis of Prosthesis Architectures and

Design of Prosthetic Devices for

Upper Limb Amputees

Marco Troncossi & Vincenzo Parenti-Castelli DIEM – Dept. of Mechanical Engineering of the University of Bologna

Italy 1.Introduction

This chapter pres ents a procedure for the Determination of the Optimal Pros thes is Architecture for upper limb amputees (DOPA). The pres ented approach can cons is tently manage both the clinical as pects and the technical is ues involved in the des ign of electromechanically actuated pros thes es. The procedure is compos ed on one hand of algorithms us eful for analyzing the patients’ requirements and on the other hand of algorithms that perform kinematic and kinetos tatic s imulations of s everal architectures of artificial arms attempting to fulfil important activities of daily living. The s ys tematic evaluation of the prosthesis models’ performance can methodically guide designers in the synthesis of the optimal prosthesis that best suits the patients’ requirements.

1.1 Prosthetic rehabilitation of upper limb amputees

The loss or the congenital deficiency of an upper limb part represents a serious physical and ps ychological trauma, apart from having an evident and cons iderable res triction on pers onal autonomy in everyday living. Rehabilitating an amputee with a proper device allows the patient to recover (part of) the los t autonomy and the s ens e of ps ychophys ical integrity, and thus to enable his/her reintegration in domes tic, working and s ocial environments.

The pros thetic intervention is a complex proces s which involves technical as pects and clinical is s ues s trictly dependent on the amputee to be treated; pros thetic rehabilitation is therefore carried out by a multidi ciplinary team including phy ician, technician, therapists and psychologists which operates with the aim of providing the amputee with the device and the s ervices that bes t match his/her different requirements. The firs t s tep the rehabilitation team must face is to investigate the individual needs of every amputee. The choice of the bes t pros thes is for a given patient depends on s everal as pects, all of which must be taken into account (Atkins & Meyer, 1989):

-amputation level

-mono- or bi-laterality of the amputation

-patient’s age

-patient’s gender

-stump conditions (shape, muscle strength, skin conditions, pain…)

-range of motion of the residual limb

康复机器人相关的论文,拆分成了13份文档,都很不错的。

556Rehabilitation Robotics -presence of other diseases

-personal motivations for rehabilitating and expectations regarding the prosthesis -psychological status

-home environment and family support

-subject’s particular characteristics

-…

Even if it is glaringly obvious that the evaluation of thes e as pects is s trictly patient-dependent, it is generally possible to state that for mono-lateral amputees the sound limb becomes dominant and the pros thes is works mainly as an auxiliary device for bimanual activities. On the contrary, for bilateral amputees the pros thes es are s trictly neces s ary to perform those activities of daily living that allow the patient not to be completely dependent on others’ help thanks to the acquis ition of a certain level of functional autonomy. Obviously, the higher the level of amputation the greater the importance of the prosthetic devices. The right s election of the proper pros thes is for a given patient relies on the assessment of the patient’s characteristics and must be made by experienced personnel.

In order to satisfy the patient’s needs the features that a prosthesis must have are:

1.the highest possible dexterity

2.good performance (in terms of velocity and forces/torques)

3.appropriate robustness

4.efficient control

5. a humanlike appearance

6. a light weight

7.proper size and proportions

8.good comfort for the wearer

9.easy control for the amputee

10.extremely reliable components of the artificial system

11. a low noise level

12.sufficient autonomy of the energy source to allow the prosthesis to work all day

It is possible to summarize the features required of a prosthesis as good functionality of the artificial arm on one hand (features 1–4) and good wearability for the patient on the other (features 5–9). The last specifications, 10 to 12, concern technological issues and the level of their obs ervance depends bas ically on the component des ign, the materials us ed and the s tate of the art of both the electronic and the mechanical fields. It is worth noting that functionality and wearability are basically contrasting features; for instance, a device which has to provide high forces and speed must supply an appropriate power, implying a size of actuators far from lightweight. When s electing the appropriate pros thes is for a given amputee, the importance to be allocated to every s ingle factor s trictly depends on the evaluation of the patient’s characteristics and requirements.

1.2 Upper limb prostheses

There are various kinds of prosthesis to be evaluated and chosen from. The “externally powered pros thes is”, i.e. a robotic arm where the artificial limb s egments are driven by electromechanical joints, is the mos t advanced. The joint motors are directly activated by the amputee by means of input commands that are collected by s pecific sensors located in the socket of the prosthesis, the socket being the interface by means of which the prosthesis is suspended on the patient’s stump. The command signals are

康复机器人相关的论文,拆分成了13份文档,都很不错的。

Synthesis of Prosthesis Architectures and Design of Prosthetic Devices for Upper Limb Amputees 557 processed by a programmable electronic circuit which carries out the control strategy to operate the device. Rechargeable batteries power all thes e components . Some passive friction joints are sometimes included in the system and are useful to give the pros thetic limb an optimal pre-determined configuration when performing certain tas ks. The pas s ive joints are operated by applying external forces by means of the sound limb or resting the artificial segments on fixed points, before or after the action of the active joints.

Currently, the mos t common electromechanical pros thetic components available on the market are many kinds of terminal devices (each with one degree of freedom – DoF – for gras ping), the elbow joint, the wris t prono-s upination unit (which allows the terminal device to rotate around the longitudinal forearm axis ), and the wris t flexion unit. Many other active articulations have been studied and proposed as prototypes but they have not yet been dis tributed commercially. Among the recent interes ting res earch res ults , the authors would like to mention: the multi-fingered pros thetic hands (Kyberd et al., 2001; Pons et al., 2004; Yang et al., 2004), which provide the terminal device with more than one DoF, thus making different grip patterns available (one of thes e s eems ready to be commercialized 1); a powered humeral rotator (Weir & Grahn, 2005) which allows the forearm segment to rotate around the longitudinal humeral axis (this is a novelty, because most above-elbow prostheses are equipped with passive humeral rotators); a shoulder joint with one DoF for upper arm elevation (Gow et al., 2001); a s houlder joint with two DoFs which is bas ed on a differential mechanis m and provides upper arm elevation and abduction (Cattaneo et al., 2001).

There are several ways of controlling electrically-powered prostheses, the most popular being myoelectric control: electromyographic ignal (EMG, i.e. electrical ignal a ociated with the mu cle contraction ), mea ured on the kin by myoelectric electrodes located in the socket, are properly amplified and filtered, and then processed by the controller that s witches the motors on or off in the active joints to produce movement s and function s . Although theoretically po s s ible, the simultaneous and independent contraction of distinct bundles of muscles, that would generate EMG signals able to operate different functions at the same time is very difficult and stressing for the patient. The myoelectric control s cheme is therefore generally bas ed on the s equential activation of the pros thetic articulations one at a time, res ulting in a not very natural control. In this context, s ome recent res ults s eem to be promis ing to overcome this limitation (Kuiken et al., 2004).

The good qualities of this pros thes is are s ufficient functionality, good performance and a pleas ant appearance. The critical as pects are the weight and the volume of the phys ical structure, and the intricate control, due to the sequential activation of both the active and passive joints. Finally, it is proven that electrically-powered prostheses provide a high level of technology but at a high cost.

1.3 Ne w prosthesis design and overvie w of the presented method

In order to provide high level disarticulated patients with a comfortable, humanlike and user-friendly prosthesis, not all the physiological joint movements can be replicated, thus limiting the functionality of the artificial arm. When compared to the human arm, the 1 /page.php?pageid=12&section=3.

康复机器人相关的论文,拆分成了13份文档,都很不错的。

558Rehabilitation Robotics dexterity of current commercial prostheses is very poor and amputees generally have to resort to compensatory movements of the residual limb, or even of other parts of the body as well as to auxiliary aids to execute many motor tasks. For patients with very high level amputations (bilateral above all), who have an extremely res tricted res idual movement ability, current pros thes es could be inadequate to guarantee the functionality needed to reach a s atis factory level of autonomy. In order to s olve this deficiency and to improve the quality of life of this amputee population, the development of new electrically powered pros thes es with greater mobility, advanced control and good “wearability” is thus required.

It is authors’ firm belief that the same observations that guide the rehabilitation team in the selection of the proper device for a given patient must also be considered by engineers and technicians when designing new prosthetic devices. Computer-based simulations represent a us eful tool, in this pers pective, for the development of new mechanical s ys tems. In particular, the work presented in (Romilly et al., 1994) shows how the kinematic simulations of artificial arm models attempting to execute given trajectories can guide the development of powered orthoses with less than six DoF.

This chapter pres ents a methodology for the s ynthes is of new pros thetic architectures for patients with high level amputation, based on a procedure for the determination of the best compromise between the contrasting features required of a prosthesis, taking into account the different needs of divers e patient profiles. “Architecture” is intended here as the geometry and the topology of a robotic arm model, i.e. the number and type (active/passive) of its joints and their arrangement.

With the proposed procedure, the characteristics, the needs and the goals of a generic patient are formalised and organised in such a way as to be systematically analysed by means of s pecifically developed algorithms. Bas ed on the collected data, further algorithms then perform the kinematic and kinetos tatic s imulations of s everal robotic arm architectures with one up to six active joints differently arranged (and considering the pos s ible pres ence of pas s ive joints as well) when carrying out s ome activities of daily living cons idered important for a given amputee. The models with les s than s ix DoF corres pond to s impler robot architectures and are thus appreciated from the wearability viewpoint; on the other hand, their performance is poorer than those of the six DoF models, because they normally carry out the manipulation tasks with an error which increases as the number of active joints decreases. The structure simplification of the e robot and the corre ponding wor ening of their global functionality are evaluated with res pect to the quality of life as s igned to the patient profile. For this purpose, some indices have been specifically developed to properly weigh up both the clinical aspects (depending on the patient) and the technical factors (depending on the robotic models).

The approach is subject-oriented, foreseeing a single patient as input and providing his/her optimal prosthesis as output. However, the final application of the methodology can supply more general des ign guidelines, s uitable to determine a limited number of pros thes is architectures able to match the requirements of many different amputees.

Finally, the results of the kinematic and kinetostatic simulations can provide the mechanical design specifications (e.g. the joint range of motion, the power of the actuators) of the new device that will prove fundamental in overcoming the limitation of the exi ting prostheses.

康复机器人相关的论文,拆分成了13份文档,都很不错的。

Synthesis of Prosthesis Architectures and Design of Prosthetic Devices for Upper Limb Amputees 559

2. The DOPA procedure

The procedure receives a given patient’s information as input and provides his/her corresponding optimal prosthesis model as output (Fig. 1). It is based on a database and a number of algorithms which make it possible to choose the optimal robot arm architecture able to fit patient’s specific needs and limitations. On one hand the database (database DB-T) collects upper limb activities of daily living and the corresponding trajectories which model them – hereinafter “Reference Trajectories”, normally requiring six DoF for positioning and orienting tasks – and on the other hand (database DB-R) several kinematic models of upper limb prostheses with one up to six active revolute joints differently arranged and possible revolute or spherical passive joints (Fig. 3). The procedure can be considered as a process of three sequential steps, running automatically once that the appropriate amputee data have been entered.

Abbreviation

M eaning

T-labels Markers of motor tasks

P-labels Markers of the patient profile

DB-T Upper Limb Activities Database

TaS Task Selection Algorithm

DB-PT Upper Limb Activities selected for the patient

Char Characterization Algorithm

DB-R Robotic Models Database

KiS Kinematic Simulation Algorithm

DB-RT1 Robotic models kinematic performance

Promotion Filter Model kinematic performance evaluation

DB-RT2 Promoted models kinematic performance

KiA Kinetostatic Analysis

DB-RT3 Promoted models dynamic performance

C Complexity Index

KP Kinematic performance Index

FL Flexibility Index

I Overall Index Fig. 1. The DOPA procedure schematic layout.

康复机器人相关的论文,拆分成了13份文档,都很不错的。

560Rehabilitation Robotics Step 1 –The compilation of the patient’s Clinical Identification Form is the s tarting point of the procedure. It collects all the information neces ary to clas ify the amputee’s needs; different as pects are inves tigated in order to identify a patient’s profile us eful to portray a pers onalized level of life quality to be achieved after the pros thetic rehabilitation plan. An algorithm (TaS) bas ed on the proces s ing of this profile determines which upper limb activities are mos t s ignificant for the given patient from the viewpoint of reaching a ati factory functional autonomy in everyday living. The prosthesis will be designed aiming at performing these selected functional tas ks. A further algorithm (Char)determines the values of the parameters involved in the s election of the optimal architecture by balancing the relative importance of the different factors which contribute to define the amputee’s quality of life (e.g. expected level of functional autonomy, s implicity of the s tructure, easiness of control, etc.).

Step 2 – Kinematic s imulations (KiS), performed for all the models in DB-R, generate the trajectories performed by the robot when attempting to fulfil the tasks assigned by TaS. The models with less than six DoF (hereinafter “Deficient Robots”), corresponding to simpler, less massive robot architectures (thus appreciated from the wearability viewpoint), execute the Reference Trajectories with a certain error which increases as the number of active joints decreases.

If the error overcomes the acceptable value fixed for every given task, then the robot model is considered not adequate to perform that activity. Only the robotic models which fulfil a given number of tas ks, dependent on the functionality required by the s ubject (Thresholds t1,…, t4), will move on to the next phas es, whereas the others will be dis carded. The structural simplification of the Deficient Robots and the corresponding worsening of their global functionality have to be evaluated with res pect to the quality of life as s igned (by means of the Char algorithm) to the given patient. A further kinetos tatic analys is (KiA), performed for the “promoted” robots, provides the values of torque and power that all the actuated joints must generate to perform the successfully-executed tasks, defining the size of the actuators in a first approximation.

Step 3 – The artificial arm models are as s es s ed in the las t s tep of the procedure: their performance and the complexity of their architecture are evaluated by means of three purpos e-built indices, all ranging from 0 to 1, named as KP,FL and C which, properly combined in an overall index I, univocally determine the optimal prosthesis architecture, i.e. the robotic arm with the simplest and lightest structure possible which can best satisfy the patient’s personal requirements:

I

w

C

w

FL

w

KP

(1)

2

3

1

where w1,w2,w3 [0,1] are weighting factors which depend on the patient’s profile.

The model with the maximum value for I des ignates the optimal architecture of the prosthesis associated with the given patient.

The main elements and algorithms of the DOPA procedure will be explained, in s uch a way to make their comprehens ion eas y for the reader. In particular they will be outlined as if the procedure must guide the design of an ad-hoc prosthesis for a single patient; the intended implementation is actually different (s ee Section 3).

康复机器人相关的论文,拆分成了13份文档,都很不错的。

Synthesis of Prosthesis Architectures and Design of Prosthetic Devices for Upper Limb Amputees 561 CLINICAL IDENTIFICATION FORM

Label Definition Value Definition

P1 Gender P1.1 Female P1.2 Male

P2 Age P2.1 0 – 15 years old

P2.2 16 – 35 P2.3 36 – 65 P2.4 > 65

P3 Body-build P3.1 Small

P3.2 Medium

P3.3 Large

P4 Non-dominant limb P4.1 Distal upper arm amputation

level of amputation *P4.2 Medial upper arm amputation

P4.3 Proximal upper arm amputation

P4.4 Shoulder di s articulation

P4.5 Forequarter amputation

P5 Dominant limb P5.1 Healthy (single extremity amputation) level of amputation *P5.2 Distal upper arm amputation

… …

P5.6 Forequarter amputation

P6 Patient preferences P6.1 High predilection for comfort and appearance about the prosthesis P6.2 Moderate predilection for comfort and appearance P6.3 No preference

P6.4 Moderate predilection for device functionality P6.5 High predilection for device functionality

P7 Living situation He/she lives with someone else who can aid him/her He/she lives alone, but someone is often present He/she lives alone, someone is occasionally present

P7.1P7.2

P7.3P7.4 He/she lives alone in complete autonomy

P8 Work P8.1 None

P8.2 Hou s eman/hou s

ewife

P8.3 Admini s trative employment

P8.4 Technical employment

P9 P9.1 None

P9.2 Cooking (and kitchen-related activities) P9.3 Hou s ework Other activities (All the activities not

related to work )

P9.4 Doing the shopping

P9.5 Driving the car

P9.6 Using home appliances (stereo, computer…) P9.7 Home maintenance and workshop activities * “Non-dominant” (ND) is the injured upper limb of a monolateral amputee or the limb with the severest injury for a bilateral amputee; “dominant” (D) is the other limb. ND and

D prostheses are the corresponding artificial arms that will replace the missing limb(s). An ND pros thes is s hould perform functions of s upport (s impler) to the D limb (natural or artificial).

Table I. A schematic layout of the Clinical Identification Form with the P-labels meaning.

本文来源:https://www.bwwdw.com/article/yaj1.html

Top