Table 2.
Description of technical features of insoles for posture and activity recognition.
Sensing Elements | Sampling Frequency (Hz) | Data Transmission Methods | Populations | Age (Years) | Algorithms | Comparison Methods | Settings | Test Durations | |
---|---|---|---|---|---|---|---|---|---|
Hegde et al. [26] | 3 FSRs 402, ACC. | 50 | Bluetooth | 15 adults | M: 26.6 (3.4) F: 23.3 (5) | MLD | ActivPAL device | Laboratory and community | 1 h (laboratory), 8 AM–9 PM (free-living) |
Fulk et al. [31] | 5 FSRs, ACC. | 400 | Bluetooth | 12 stroke survivors | 62.1 (8.2) | ANN | Video record | Laboratory | 2 min/activity |
Achkar et al. [32] | 8 FSRs, IMU*, barometer | 200 | Wire connection | 10 older adults | 65–75 | DT | ACC., gyroscope | Community | 1 h/participant |
Achkar et al. [33] | 8 FSRs, IMU*, barometer | 200 | Wire connection | 10 older adults | 69.9 (3.1) | DT | 2D ACC., gyroscope | Community | 4 h (total) |
Anlauff et al. [38] | 4 FSRs | 200 | Bluetooth | 8 adults | 25.2 (NA) | NA | Visual observation | NA | 45 min (total) |
Chen et al. [39] | 4 FlexiForce, IMUs* | 100 | Wireless mode | 7 adults | 24.1 (0.5) | LDA | Visual observation | NA | 15 min/experiment |
Fulk et al. [40] | 5 FSRs, Acc. | 25 | Wireless link | 8 stroke survivors | 60.1 (9.9) | SVM | Visual observation | Laboratory | 1 min/activity |
Zhang et al. [41] | 32 miniature pressure sensors | 32 | Wire connection | 40 adults | 27.3 (13.2) | ANN | Visual observation | Outside, laboratory | 50 m |
Zhang et al. [42] | 5 FSRs, ACC. | 25 | Bluetooth | 12 stroke survivors | 62.1 (8.2) | DT | Visual observation | Laboratory | 2 min/activity |
Edgar et al. [43] | 3 pressure sensors, ACC. | 100 | Bluetooth | 1 adult | 22 | ANN | Visual observation | Indoor and outdoor | 3 min/activity |
Hegde et al. [44] | 2 or 3 FSRs 402 | NA | Bluetooth | 3 adults | 24 (4.5) | MLD | Visual observation | Laboratory | 20 min/activity |
Lin et al. [45] | 48 pressure sensors, IMU | 100 | Bluetooth | 8 people | NA | KNN | Visual observation | Indoor | NA |
Lin et al. [46] | 48 pressure sensors, IMU | 100 | Bluetooth | 8 people | NA | NA | Visual observation | Indoor | 10 trials/participant |
Peng et al. [47] | 7 FSR402 | 25 | Wireless module | 1 adult | 24 | SVM | Visual observation | Indoor | NA |
Sazonov et al. [48] | 5 FSRs, ACC. | 400 | Bluetooth | 19 adults | 28.1 (6.9) | SVM, MLP, MLD | Video | Indoor and free-living | 52.5 h (total) |
Sazonov et al. [49] | 5 FSRs, ACC. | 25 | Wireless module | 9 adults | 23.7 (4.3) | SVM | Visual observation | Laboratory | 11 h 36 min (total) |
Shang et al. [50] | 2 pressure sensors, ACC. | NA | Wireless module | 3 adults | NA | Threshold method | Visual observation | Laboratory | NA |
Sugimoto et al. [51] | 7 pressure sensors | 20 | USB port | 2 adults | NA | LDA | Visual observation | NA | 2 min |
Tang et al. [52] | 5 FSRs, ACC. | 25 | Wireless module | 9 adults | 23.6 (4.3) | SVM with rejection | Visual observation | NA | NA |
Tang et al. [53] | 5 FSRs, ACC. | 400 | Wireless module | 9 adults | 23.3 (4.3) | SVM, MLP | Visual observation | Indoor | 11.5 h (total) |
Zhang et al. [54] | 5 FSRs, ACC. | 25 | Wireless module | 9 adults | 27.3 (4.3) | DT | Visual observation | NA | 11.36 h (total) |
Zhang et al. [55] | 4 pressure sensors | 35 | Bluetooth | 10 adults | 24–56 | NA | Visual observation | Community | NA |
Chen et al. [56] | 4 pressure sensors | 250 | Wireless module | 5 adults, 1 amputee person | 23.2 (1.3); 45 | DT, LDA | Visual observation | NA | 8 h |
Cates et al. [57] | 4 FSRs, ACC. | 20 | Bluetooth | 20 adults | 28 (5) | SVM | Visual observation | NA | 2 min/activity |
Hegde et al. [58] | 3 pressure sensors, ACC. | 50 | Bluetooth | 4 adults | 28 (0.5) | MLD | Visual observation | Laboratory | 10 min/activity |
Cuong Pham et al. [59] | ACC. | 50 | Wireless module | 10 adults | 22 (1.7) | CNN | Visual observation | NA | 10–30 min/activity |
Nguyen et al. [60] | 8 pressure sensors, ACC. | 50 | Bluetooth | 3 adults | 24–29 | DT, KNN, SVM | PPAC and FF + GPS | Indoor and outdoor | 15 and 20 m, 17-step (stairs) |
ACC.: accelerometer; SVM: Support vector machine; MLP: Multi-layer perceptron; ANN: artificial neural network; LDA: Linear discriminant analysis; KNN: k-nearest-neighbors; MLD: Multinomial Logistic Discrimination; CNN: convolution neural networks; DT: decision tree; NA: not applicable; * Physilog module including an IMU* (accelerometer, gyroscope and magnetometer) and a barometer sensor: Physilog® 10D Silver, GaitUP CH; FF+GPS: foot force sensor and GPS; PPAC: plantar-pressure based ambulatory classification; min: minute; h: hour; FSR: force sensitive resistor.