| A | B | C | D | E | F | G | H | I | J | K | L | M | N | O | P | Q | R | S | T | U | V | W | X | Y | Z | |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
1 | Article | Population | Pain Setting | Pain Scale (Ground Truth) | ML Classifiers | Output | Results | Accuracy (DOP) | Accuracy (PI) | Average Accuracy | Average Accuracy | |||||||||||||||
2 | Artificial intelligence to evaluate postoperative pain based on facial expression recognition - Fontaine - 2022 | Adult patients from a single university hospital | Postoperative pain | Numerical Rating Scale (NRS) | Convolutional Neural Network (CNN) | Pain Intensity estimation | Pain Intensity - Accuracy = 53% Detection of Pain - Accuracy = 89% | 89.00% | 53.00% | 85.05% | 73.90% | |||||||||||||||
3 | Deep Pain: Exploiting Long Short-Term Memory Networks for Facial Expression Classification | UNBC-McMaster database | Self-identified shoulder pain | Prkachin and Solomon Pain Intensity (PSPI) | Convolutional Neural Network - Long short-term memory (CNN - LSTM) | Pain Detection and Pain Intensity | Pain Detection - Accuracy = 83% - AUC = 93.3% Pain Intensity Estimation - MSE = 0.74 | 83.00% | 26.00% | DOP = Detection of Pain | PI = Pain Intensity | |||||||||||||||
4 | Automatic Decoding of Facial Movements Reveals Deceptive Pain Expressions | Healthy Subjects | Cold pressor-induced pain | Pain stimuli-dependent assessments | Support Vector Machine (SVM) | Detection of genuine vs. faked pain | AUC = 0.91% Accuracy = 85% | 85.00% | N/A | |||||||||||||||||
5 | Automatic vs. Human Recognition of Pain Intensity from Facial Expression on the X-ITE Pain Database | X-ITE Pain Database | Heat-induced and electrical-induced pain | NRS categorized into 4 pain intensities (no pain, low, medium, and severe) | 2-CNN's with sample weighting (complementing each other) | Pain Intensity estimation with two groupings of pain levels (none/low/severe & none/moderate/severe) | Accuracy = 51.7% | N/A | 51.70% | |||||||||||||||||
6 | Ensemble neural network approach detecting pain intensity from facial expressions - ScienceDirect | UNBC-McMaster database | self-identified shoulder pain | Prkachin and Solomon Pain Intensity (PSPI) | Convolutional Neural Network - Recurrent neural network (CNN - RNN) | Pain Intensity estimation | UNBC-McMaster - Accuracy = 86% - AUC = 90.5% | N/A | 86.00% | |||||||||||||||||
7 | Ensemble neural network approach detecting pain intensity from facial expressions - ScienceDirect | MIntPAIN database | electrical-induced pain | Stimuli-based pain levels (0–4) | Convolutional Neural Network - Recurrent neural network (CNN - RNN) | Pain Intensity estimation | MIntPAIN - Accuracy = 92.26% - AUC = 93.67% | N/A | 92.26% | |||||||||||||||||
8 | A novel approach for pain intensity detection based on facial feature deformations - ScienceDirect | UNBC-McMaster database | self-identified shoulder pain | Prkachin and Solomon Pain Intensity (PSPI) | Double Machine Learning (DML) combined with Support Vector Machine (SVM) | Pain Intensity estimation (PSPI) | Accuracy = 96% | N/A | 96.00% | |||||||||||||||||
9 | Automatically Detecting Pain in Video Through Facial Action Units | IEEE Journals & Magazine | UNBC-McMaster database | self-identified shoulder pain | Prkachin and Solomon Pain Intensity (PSPI) | Support Vector Machine (SVM) | Pain detection | Accuracy = 80.9% AUC = 84.7% | 80.90% | N/A | |||||||||||||||||
10 | Enhanced deep learning algorithm development to detect pain intensity from facial expression images - ScienceDirect | UNBC-McMaster database | self-identified shoulder pain | Prkachin and Solomon Pain Intensity (PSPI) | Hybrid Convolutional Neural Network (CNN) - bidirectional Long Short-term Memory (LSTM) | Pain Intensity Estimation, categorized into four levels (PSPI 0, 1, 2-3, greater than 4) | Accuracy = 85% AUC = 88.7% MSE = 0.21 MAE = 0.18 F-measure = 78.2% | 85.00% | 79.00% | |||||||||||||||||
11 | Automatic coding of facial expressions displayed during posed and genuine pain - ScienceDirect | University students | Cold pressor-induced pain | Pain stimuli-dependent assessments | Gaussian Support Vector Machine (SVM) | genuine vs. faked pain | Accuracy = 88% | 88.00% | N/A | |||||||||||||||||
12 | Automated detection of pain levels using deep feature extraction from shutter blinds-based dynamic-sized horizontal patches with facial images | Scientific Reports | UNBC-McMaster database | self-identified shoulder pain | Prkachin and Solomon Pain Intensity (PSPI) | K-Nearest Neighbor | Pain Intensity Estimation, categorized into four levels (PSPI 0, 1, 2-3, greater than 4) | Accuracy = 95.57% Average F1 = 95.67% (The F1 score is calculated as the harmonic mean of the precision and recall scores) | 95.57% | ||||||||||||||||||
13 | The modeling of human facial pain intensity based on Temporal Convolutional Networks trained with video frames in HSV color space | UNBC-McMaster database | self-identified shoulder pain | Prkachin and Solomon Pain Intensity (PSPI) | Temporal Convolutional Network (TCN) | Pain Intensity Estimation, categorized into four levels (PSPI 0, 1, 2-3, greater than 4) | Accuracy = 94.14% AUC = 91.3% MSE = 0.186 MAE = 0.234 | 94.14% | 81.40% | |||||||||||||||||
14 | The modeling of human facial pain intensity based on Temporal Convolutional Networks trained with video frames in HSV color space | MIntPAIN database | electrical-induced pain | Stimuli-based pain levels (0–4) | Temporal Convolutional Network (TCN) | Pain Intensity Estimation, categorized into five levels (0-4) | Accuracy = 89% AUC = 92% MSE = 0.22 MAE = 0.26 | 89.00% | 78.00% | |||||||||||||||||
15 | Multiview Distance Metric Learning on facial feature descriptors for automatic pain intensity detection - ScienceDirect | UNBC-McMaster database | self-identified shoulder pain | Prkachin and Solomon Pain Intensity (PSPI) | Support Vector Machine (SVM) | Pain detection Estimation of pain intensity, categorized into four levels: PSPI 0, 1, 2, and ≥3 | Pain detection Accuracy = 89.59% Pain intensity estimation Accuracy = 75% | 89.59% | 75.00% | |||||||||||||||||
16 | Metrological Characterization of a Pain Detection System Based on Transfer Entropy of Facial Landmarks | IEEE Journals & Magazine | UNBC-McMaster database | self-identified shoulder pain | Visual Analog Scale | Linear discriminant analysis | Pain detection (VAS≥0) Pain intensity (VAS) estimation | Pain detection AUC = 0.87 Pain intensity estimation MAE = 2.44 | 85.00% | 97.56% | |||||||||||||||||
17 | Self-supervised pain intensity estimation from facial videos via statistical spatiotemporal distillation - ScienceDirect | UNBC-McMaster database | self-identified shoulder pain | Prkachin and Solomon Pain Intensity (PSPI) | Convolutional Neural Networks (CNNs) | Estimation of pain intensity UNBC-McMaster: 16 pain levels | Training with BioVid and testing on UNBC-McMaster Self-supervised model: AUC = 0.692 Supervised model: AUC = 0.801 | 75.00% | ||||||||||||||||||
18 | Self-supervised pain intensity estimation from facial videos via statistical spatiotemporal distillation - ScienceDirect | BioVid database (part A) | heat-induced pain | stimuli-based pain (5 levels) | Convolutional Neural Networks (CNNs) | Estimation of pain intensity BioVid: 5 pain levels | Training with UNBC-McMaster and testing on BioVid (Self-supervised) model: AUC = 0.655 (Supervised) model: AUC = 0.755 | 75.00% | ||||||||||||||||||
19 | Automated Assessment of Children’s Postoperative Pain Using Computer Vision | Pediatrics | Pediatric patients from a tertiary care center | postoperative pain | Numerical Rating Scale (NRS) | Logistic regression and linear regression models | Pain Detection (NRS ≥ 4) Pain intensity estimation (NRS) | Pain Detection Baseline pain: AUC = 0.84 Pain intensity estimation Baseline pain: r = 0.47; z = 4.4 * | 85.00% | 47.00% | |||||||||||||||||
20 | Deep Learning for Identification of Acute Illness and Facial Cues of Illness | volunteers | states of acute illness (fake) | use of lipopolysaccharide (LPS) which triggers an immune response resembling acute illness | Neural transfer convolutional Neural Network (NT-CNN) and four Convolutional Neural Networks (CNN) | Pain detection | Pain Detection AUC = 0.67 | 67.00% | N/A | |||||||||||||||||
21 | ||||||||||||||||||||||||||
22 | ||||||||||||||||||||||||||
23 | ||||||||||||||||||||||||||
24 | ||||||||||||||||||||||||||
25 | ||||||||||||||||||||||||||
26 | ||||||||||||||||||||||||||
27 | ||||||||||||||||||||||||||
28 | ||||||||||||||||||||||||||
29 | ||||||||||||||||||||||||||
30 | ||||||||||||||||||||||||||
31 | ||||||||||||||||||||||||||
32 | ||||||||||||||||||||||||||
33 | ||||||||||||||||||||||||||
34 | ||||||||||||||||||||||||||
35 | ||||||||||||||||||||||||||
36 | ||||||||||||||||||||||||||
37 | ||||||||||||||||||||||||||
38 | ||||||||||||||||||||||||||
39 | ||||||||||||||||||||||||||
40 | ||||||||||||||||||||||||||
41 | ||||||||||||||||||||||||||
42 | ||||||||||||||||||||||||||
43 | ||||||||||||||||||||||||||
44 | ||||||||||||||||||||||||||
45 | ||||||||||||||||||||||||||
46 | ||||||||||||||||||||||||||
47 | ||||||||||||||||||||||||||
48 | ||||||||||||||||||||||||||
49 | ||||||||||||||||||||||||||
50 | ||||||||||||||||||||||||||
51 | ||||||||||||||||||||||||||
52 | ||||||||||||||||||||||||||
53 | ||||||||||||||||||||||||||
54 | ||||||||||||||||||||||||||
55 | ||||||||||||||||||||||||||
56 | ||||||||||||||||||||||||||
57 | ||||||||||||||||||||||||||
58 | ||||||||||||||||||||||||||
59 | ||||||||||||||||||||||||||
60 | ||||||||||||||||||||||||||
61 | ||||||||||||||||||||||||||
62 | ||||||||||||||||||||||||||
63 | ||||||||||||||||||||||||||
64 | ||||||||||||||||||||||||||
65 | ||||||||||||||||||||||||||
66 | ||||||||||||||||||||||||||
67 | ||||||||||||||||||||||||||
68 | ||||||||||||||||||||||||||
69 | ||||||||||||||||||||||||||
70 | ||||||||||||||||||||||||||
71 | ||||||||||||||||||||||||||
72 | ||||||||||||||||||||||||||
73 | ||||||||||||||||||||||||||
74 | ||||||||||||||||||||||||||
75 | ||||||||||||||||||||||||||
76 | ||||||||||||||||||||||||||
77 | ||||||||||||||||||||||||||
78 | ||||||||||||||||||||||||||
79 | ||||||||||||||||||||||||||
80 | ||||||||||||||||||||||||||
81 | ||||||||||||||||||||||||||
82 | ||||||||||||||||||||||||||
83 | ||||||||||||||||||||||||||
84 | ||||||||||||||||||||||||||
85 | ||||||||||||||||||||||||||
86 | ||||||||||||||||||||||||||
87 | ||||||||||||||||||||||||||
88 | ||||||||||||||||||||||||||
89 | ||||||||||||||||||||||||||
90 | ||||||||||||||||||||||||||
91 | ||||||||||||||||||||||||||
92 | ||||||||||||||||||||||||||
93 | ||||||||||||||||||||||||||
94 | ||||||||||||||||||||||||||
95 | ||||||||||||||||||||||||||
96 | ||||||||||||||||||||||||||
97 | ||||||||||||||||||||||||||
98 | ||||||||||||||||||||||||||
99 | ||||||||||||||||||||||||||
100 |