Skip to main content

Table 6 GROUP II: classification results and confusion matrices for TEST1

From: EMG-based pattern recognition approach in post stroke robot-aided rehabilitation: a feasibility study

Patient id and classification rate

Actual classes

Predicted classes

  

North

East

South

West

P1

North

28.6

25

0

58.3

43.3%

East

14.3

75

28.6

16.7

South

42.9

0

71.4

0

West

14.3

0

0

25

P2

North

100

0

0

0

70%

East

0

77.8

0

0

South

0

22.2

85.7

60

West

0

0

14.3

40

P3

North

42.9

26.6

0

0

30%

East

0

28.6

66.7

28.6

South

28.6

42.9

11.1

28.6

West

28.6

0

22.2

42.9

P4

North

12.5

57.1

0

0

26.7%

East

87.5

28.6

0

0

South

0

14.3

37.5

71.4

West

0

0

62.5

28.6

P5

North

14.3

0

0

20

40%

East

28.6

57.1

16.7

20

South

14.3

42.9

83.3

40

West

42.9

0

0

20

P6

North

83.3

0

0

0

70%

East

16.7

45.5

12.5

0

South

0

54.5

75

0

West

0

0

12.5

100

P7

North

100

0

0

0

66.7%

East

0

57.1

60

0

South

0

42.9

40

0

 

West

0

0

0

100

  1. Elements on the left-right diagonal indicate the percentage of correct classification, while elements on the off-diagonals denote percentage of misclassification. SVM was trained (70% of trials) and tested (30% of trials) individually on each patient with the muscle dataset selected for healthy subjects; training and testing data were randomly selected and mutually exclusive; test was repeated 5 times for each patient and accuracy rates were averaged across iterations.