Table 1 Representative literature about quantitative measures of finger movements.
Reference | Test | Task | Sample | Parameters studied | Accuracy | Clinical correlation |
|---|---|---|---|---|---|---|
Noyce et al. 20147 | BRAIN test: keyboard | ATT 30” | 58 PD 93 AMC | KSa AT IS | KS: 56% sensitivity, 80% specificity | KS—total UPDRS-III r = −0.53 |
Khan et al. 20148 | Computer vision framework and videos | FT | 13 PD 6 HC | Accuracy 95% | NR | |
Kassavetis et al. 20159 | Smartphone application and accelerometer | ATT | 14 PD | Tapping frequency Mean moving time Distance between taps | NR | Tapping frequency on the phone (r = − 0.75; P = 0.001), the mean moving time (r = 0.65; P = 0.001), and the distance between taps (r = − 0.61; P = 0.003 |
Maetzler et al. 201510 | Digitomotography | FT | 33 PD 18 HC | IPI TF DEV | NR | IPI-UPDRS-III: r2 = 0.02, FT r2 = 0.01 TF-UPDRS-III: r2 = 0.02, FT r2 = 0.03 DEV-UPDRS: r2 = 0.16, FT r2= 0.16 |
Arora et al. 201511 | Smartphone | ATT | 20 PD | Not specified: recorded voice, posture, gait, FT, reaction time test | Whole app PD vs controls mean sensitivity: 96.2% (SD 2%) mean specificity 96.9% (SD 1.9%) (finger tapping detail not given in isolation) | Mean error of whole app and UPDRS: 1.26 |
Sano et al. 201612 | PDFTsi—magnetic sensors | FT | 21 PD | Distance, velocity, acceleration, interval | NR | Mean sequare error – 0.45 |
Lee et al. 201613 | Smartphone tapper | ATT (10”) | 57 PD 87 HC | Number taps Amplitudea Inter-tap distance Dwelling time | Total distance: AUC: 0.92 (95% CI 0.88–0.96) Dwelling time: AUC: 0.88 (95% CI 0.82–0.93) | Overall test—UPDRS-III r2 = 0.25 Overall test—UPDRS- FT sub-score r2 = 0.32 |
Ruzicka et al. 201614 | Contactless 3D motion capture system | FT (10”) | 22 PD 22 HC | AvgFrq MaxOpV AmpDec | AmpDec: AUC = 0.87 MaxOpV: AUC = 0.81 | MaxOpV-UPDRS-FT sub-score r = −0.48 |
Mitsi et al. 201715 | Tablet based application (iMotor) | ATT (30”) | 19 PD 17 HC | Total taps, tap accuracy, velocity, interval, duration, reaction time | AUC 0.98 (0.93–1) 94% sensitivity, 93% specificity | UPDRS and tap accuracy: r = −0.35 |
Van den Noort et al. 201716 | Sensors—PowerGlove | FT | 4 PD | FT, hand opening closing, pronation/supination | NR | NR |
Gao et al. 201817 | PD-monitor (sensor) | FT (30”) | 107 PD 49 HC 41 ET | EA-dynamical classifiersb | PD-monitor score: AUC = 0.89 | Right side—MDS-UPDRS-FT: r = 0.82 Left side—MDS-UPDRS-FT: r = 0.78 |
Zhan et al. 201818 | Smartphone and ML | ATT | 129 PD 0 HC | Voice, FT, gait, balance, reaction time | NR | Overall test—UPDRS r = 0.88, p < 0.001. Did not stratify different parameters |
Lipsmeier et al. 201819 | Smartphone | ATT | 44 PD 35 HC | Sustained phonation, rest tremor, postural tremor, finger-tapping, balance, and gait. Passive movements | PD vs controls p < 0.005. No AUC | NR |
Prince et al. 201820 | Smartphone | ATT | 312 PD 236 HC | Speed, rhythm, accuracy and fatigue | AUC 65.7% | NR |
Butt et al. 201821 | Leap motion controller and different ML techniques | FT (10”) | 16 PD 12 HC | Velocity, angle, amplitude, and frequency | Log regression 70.37% AUC 0.831. Naive Bayes 81.4% (AUC 0.811) | R: −0.72 |
Wissel et al. 201822 | Tablet | ATT | 11 PD 11 HC | Total number of taps, tap interval, tap duration, and tap accuracy | ON vs OFF (0.60 ≤ AUC ≤ 0.82) | Tapping data and UPDRS effect moderate (−0.55 to 0.55) |
Lee et al. 201923 | Leap motion controller (hand tracker) | FT | 8 PD | Amplitude, frequency, velocity, slope and variance | NR | R = 0.86 |
Bobic et al. 201924 | Wearable sensors and 3D gyroscope | FT | 13 PD 17 MSA 14 PSP 12 HC | Velocity, amplitude, amplitude decrement, hesitations and freezes, speed | NR | Test vs neurologists accuracy 82.69% + /- 2.72 |
Shin et al. 202025 | Conventional camera DL tracking algorithm | FT LA (10”) | 29 PD 1 HC | Amplitude (mean, variabilitya) Interpeak interval (mean, variabilitya) | NR | FT – UPDRS-III: Interpeak interval var: r = 0.66 LA-UPDRS-III: Interpeak interval var: r = 0.7 |
Williams et al. 202026 | Smartphone camera DL tracking algorithm | FT (10”) | 39 PD 30 HC | Speed Amp CV Rhythm | NR | r = 0.74 (speed in MBRS) r = 0.69 (three parameters combined) |
Li et al. 202027 | 3D FT measurement-sensor units and computer | FT | 43 PD 30 HC | Motor coordination: slowness, amplitude, hesitation | NR | NR |
Zhao et al. 201928 | Videos and time series clustering | FT | 39 PD 30 HC | Decrement | NR | NR |
Alberts et al. 202129 | Smartphone | ATT | 23 PD | Number of taps, intertap interval and errors (double tapping) | NR | FT vs UPDRS: R = − 0.31, p = 0.04 Errors/freezing vs UPDRS: R = 0.44, p < 0.01; R = 0.43, p < 0.01, respectively |
