Skip to main content

Table 2 Description of studies on upper limb rehabilitation (Second part)

From: Action observation training for rehabilitation in brain injuries: a systematic review and meta-analysis

AuthorExperimental groupControl groupStudy OutcomeResults: differences between experimental and control group
Type of AOTVideos (perspectives)Type of intervention
Buccino G. et al.; 2018 [20]15 videos of task of daily action subdivided in ¾ motor segmentsDifferent perspectivesVideos with no specific motor content + execution of same actions at the EGMUUL, AHAAfter treatment, the functional score gain was significantly different in the EG and CG
Fu J et al.; 2017 [21]Observation (10 min) + imitating (10 min)
Actions: Many different movements in different directions with different complexity
Each action was filmed from 2 different anglesObserve videos of different geometric patterns and symbols + performed action as the EGFMA, WMFT, MBI, MEPFMA, WMFT, MBI increased significantly compared with that before therapy in both groups. The indexes were significantly higher in the EG group compare to the CG
Kirkpatrick E et al.; 2016 [22]watch a parent perform a movement + execution(no video) egocentric viewpointPurposeful action observation program (without the observation)AHA, MA2, ABILHAND-Kidsno between-group differences in AHA, MA2, or ABILHAND-Kids at 3 and 6 months vs baseline
Kim CH et al.; 2016 [23]Observation (9 min), followed by a break (1 min to organize + practicing (30 min)
Actions: Task-oriented training consisted of performing task based on ADLs
Each video provided 3 views simultaneously: front, side and topThe same tasks during a 30 min period, without watching the videoFMA, BBT, MBI, MASThe mean change of FMA, BBT, and MBI in the AOT was significantly different between groups. No differences at MAS
Zhu M-H et al.; 2015 [1]Observation followed by the action
Actions: total of 30 actions showing many different movements or more complex actions
Straight on (20s), right above (15 s) and right inside (15 s)Routine rehabilitation treatment and nursingFMA, MBI, MASFMA, BI and MAS scores were significantly better after treatment in the EG compared to the CG
Kim E et al.; 2015 [24]Observation of 2 from a variety of activities per session selected by patients, repeated over 1 weekn.s.Perform the purposeful AO program without observing actionsWMFTThe EG showed significantly greater improvement compared with the CG
Kim E et al.; 2015 [25]Observation of 2 daily life activities per session selected by patients repeated over 1 weekn.s.Perform the AOT assignments, without the observational part3D motion analysis systemEG showed improvement than the CG (no significant). Both groups showed improvements in average velocity, trajectory ratio, and movement degree, but not statistically significant
Sale P et al.;2014 [26]Observation followed by performing the same tasks (2 min) from the easiest to the most complex action
Actions: 20 daily activities composed by 3 different meaningful motor sequences displayed in order of ascending difficulty
first-personControl Treatment: 5 static images displaying objects, without any animal or human being, for 3 min + to perform the same tasks of the EGBBT, FMASignificant higher gain for EG than CG, with respect to functional measures taken at T1 and T2. Left hemiparetic subjects achieved significantly greater benefits compared to the right ones. FMA and BBT between groups, statistically significant differences only for left hemiparetic.
Sgandurra G et al.; 2013 [5]15 sets of daily life, un- or bi-manual goal-related actions of increasing complexityFirst-person perspectiveTo watch computer games + verbally instructions to perform same actions as AOT groupAHA, MUUL, ABILHAND-KidsSignificant AHA within-group differences at all follow-up assessments. At T1 significant between-group difference and at T2 and T3 at the limits of significance. No differences at MUUL and ABILHAND-Kids
Lee D et al.; 2013 [27]AOT group: observation of task video of drinking behaviour (5 min) followed by the actions (5 min)From the front of the modelObservation group: observation of a task video (20 times);
APG: repeatedly practiced actions performed during the preliminary test for 10 min; CG: neither watched the video nor practiced the actions
Number of times the full drinking action was performed in 1 minAll groups showed statistically significant improvements compared to CG. Combined group had a significant higher number of drinking behaviors than Observation group, immediately after and 1 week after the experiment. No statistical differences between the Combined and the AOT group
Cowles T et al.; 2013 [28]Observation (1–2 min) followed by action (4–6 min) performed simultaneously with the therapistNo videoCPT as deemed appropriateMI, ARATThe median (95% CI) between-group difference was not statistically significant
Buccino G et al.;2012 [3]motor tasks of actions related to the children’s daily livesDifferent perspectivesVideo (no specific motor content) + execution of same actions as the EGMUULAfter treatment, the functional score gain was significantly different in the EG and CG
Franceschini M. et al.;2012 [29]Observation of 1 task per day consisting in three different 3-min meaningful motor sequences, from the easiest to the most complex action + to imitate the observed motor sequence. The actions were 20 daily activitiesfirst-personControl treatment or “sham” AO = to observe for 5 min 5 static images (no motor content) + to perform UL movements as well as feasible for 2 min according to a standard sequence, simulating those performed by the EGBBT, FAT, FMA, MAS, FIMMDifferences between the 2 groups were found from T0 to T1 and from T1 to T2. However, no difference was found on either change in BBT performance from T1 to T2. No significant difference between the study groups was found in the FIMM and FMA performance
Ertelt D. et al.; 2007 [6]6 min videos showing daily life hand/upper limb actions + 6 min of repetitive practice of the observed actions with their paretic UL. 3 hand/upper limb movements of increasing complexity each day3 different perspectivesSame as the EG but sequences of geometric symbols and letters. The practiced hand and upper limb actions were performed by instruction of the therapist in the exact order as they were practiced in the experimental conditionFAT, WMFT, SISSignificant improvement of motor functions as compared to T0, and compared with CG, maintained for at least 8 weeks after the end of the intervention. Neural activations between EG and CG after training shows significant rise in bilateral ventral premotor cortex, bilateral superior temporal gyrus and supplementary motor area
  1. AOT Action Observation Therapy, EG Experimental Group, CG Control Group, MUUL Melbourne Assessment of Unilateral Upper Limb Function Scale, AHA Assisting Hand Assessment, FMA Fugl-Meyer Assessment, WMFT Wolf Motor Function Test, MBI Modified Barthel Index, MEP Motor Evoked Potential, MA2 Melbourne Assessment 2, ADLs Activities of Daily Living, BBT Box and Block Test, MAS Modified Ashworth Scale, BI Barthel Index, n.s not specified, APG Action Practice Group, CPT Conventional Physical Therapy, MI Motricity Index, ARAT Action Research Arm Test, AO Action Observation, FAT Frenchay Arm Test, FIMM Functional Independence Measure Motor Item, UL Upper Limb, SIS Stroke Impact Scale