-
Notifications
You must be signed in to change notification settings - Fork 0
Expand file tree
/
Copy pathindex.html
More file actions
328 lines (296 loc) · 17.2 KB
/
index.html
File metadata and controls
328 lines (296 loc) · 17.2 KB
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
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
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
<!DOCTYPE html>
<html lang="en">
<head>
<meta charset="UTF-8">
<meta name="viewport" content="width=device-width, initial-scale=1.0">
<title>Deep Learning-enabled Gait Assessment in Parkinson's Disease</title>
<meta name="description" content="A smartphone video-based framework for precise assessment of Parkinson's disease severity using deep learning">
<link rel="preconnect" href="https://fonts.googleapis.com">
<link rel="preconnect" href="https://fonts.gstatic.com" crossorigin>
<link href="https://fonts.googleapis.com/css2?family=Inter:wght@300;400;500;600&family=Merriweather:ital,wght@0,300;0,400;0,700;1,400&display=swap" rel="stylesheet">
<link rel="stylesheet" href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.0/css/all.min.css">
<link rel="stylesheet" href="style.css">
</head>
<body>
<header>
<div class="container">
<h1>Deep Learning-enabled Accurate Assessment of Gait Impairments in Parkinson's Disease</h1>
<div class="authors-container">
<div class="authors-list">
<span class="author-name">Jianda Han<sup>1,2,†</sup></span>,
<span class="author-name">Zhihua Tian<sup>1,†</sup></span>,
<span class="author-name">Jialing Wu<sup>3,†</sup></span>,
<span class="author-name">Kai Zhang<sup>1</sup></span>,
<span class="author-name">Shaohua Li<sup>1</sup></span>,
<span class="author-name">Fahd Baig<sup>4</sup></span>,
<span class="author-name">Peipei Liu<sup>3</sup></span>,
<span class="author-name">Ravi Vaidyanathan<sup>5,6</sup></span>,
<span class="author-name">Francesca Morgante<sup>4</sup></span>,
<span class="author-name">Weiguang Huo<sup>1,2,*</sup></span>
</div>
<div class="affiliations-list">
<span class="affiliation"><sup>1</sup> College of Artificial Intelligence, Nankai University, Tianjin, China</span>
<span class="affiliation"><sup>2</sup> Institute of Intelligence Technology and Robotic Systems, Shenzhen Research Institute, Nankai University, Shenzhen, China</span>
<span class="affiliation"><sup>3</sup> Department of Neurology, Tianjin Huanhu Hospital, Tianjin, China</span>
<span class="affiliation"><sup>4</sup> Neurosciences and Cell Biology Institute, Neuromodulation and Motor Control Section, City St George's University of London, London, UK</span>
<span class="affiliation"><sup>5</sup> Department of Mechanical Engineering, Imperial College London, London, UK</span>
<span class="affiliation"><sup>6</sup> Dementia Research Institute Care Research & Technology Centre, Imperial College London, London, UK</span>
</div>
<div class="author-notes">
<p><sup>†</sup> These authors contributed equally to this work</p>
</div>
</div>
<div class="header-buttons">
<a href="https://codeocean.com/capsule/7700825" class="btn btn-primary" target="_blank" rel="noopener">
<i class="fa-solid fa-database"></i> Code & Data
</a>
<a href="https://gaitanalysis.simplaj.fun" class="btn btn-outline" target="_blank" rel="noopener">
<i class="fa-solid fa-play"></i> Online Demo
</a>
</div>
</div>
</header>
<section id="abstract">
<div class="container">
<h2 class="section-title">Abstract</h2>
<div class="paper-abstract">
<p><strong>Background:</strong> Gait impairments are among the most prevalent and disabling symptoms in Parkinson's Disease (PD), featuring complex and highly heterogeneous manifestations.</p>
<p><strong>Methods:</strong> We propose a deep learning-based framework to assess gait impairments using smartphone-recorded videos. This framework demonstrated high proficiency in predicting PD severity.</p>
<p><strong>Results:</strong> The model achieved a micro-average area under the receiver operating characteristic curve (AUC) of <strong>0.87</strong> and an F1 score of <strong>0.806</strong>, comparable to the average performance of three clinical specialists. Additionally, it effectively discerned the comprehensive efficacy of medications on gait impairments with a precision of <strong>73.68%</strong>.</p>
</div>
<div class="metrics">
<div class="metric">
<div class="metric-value">0.87</div>
<div class="metric-label">Micro-average AUC</div>
</div>
<div class="metric">
<div class="metric-value">0.806</div>
<div class="metric-label">F1 Score</div>
</div>
<div class="metric">
<div class="metric-value">73.68%</div>
<div class="metric-label">Medication Effect Precision</div>
</div>
<div class="metric">
<div class="metric-value">Expert-level</div>
<div class="metric-label">Performance</div>
</div>
</div>
</div>
</section>
<section id="introduction" class="bg-light">
<div class="container">
<h2 class="section-title">Introduction</h2>
<p class="lead-text">Parkinson's disease (PD) is the second most prevalent progressive neurodegenerative disorder, affecting more than 10 million people worldwide. Gait impairments are among the most common and disabling symptoms in PD, characterized by intricate underlying mechanisms and substantial individual variation in clinical manifestations.</p>
<div class="card-grid">
<div class="card">
<div class="card-icon">
<i class="fa-solid fa-triangle-exclamation"></i>
</div>
<div class="card-content">
<h3>Problem</h3>
<p>Current clinical rating scales like UPDRS have low sensitivity, inherent subjectivity, and dependency on clinical specialists, limiting their utility in routine assessment of gait impairments.</p>
</div>
</div>
<div class="card">
<div class="card-icon">
<i class="fa-solid fa-mobile-screen-button"></i>
</div>
<div class="card-content">
<h3>Solution</h3>
<p>Our framework uses smartphone-recorded videos and deep learning to provide objective, precise assessment of gait impairments, enabling home-based monitoring.</p>
</div>
</div>
<div class="card">
<div class="card-icon">
<i class="fa-solid fa-fingerprint"></i>
</div>
<div class="card-content">
<h3>Innovation</h3>
<p>We extract both traditional clinically used motion markers and discover novel digital biomarkers sensitive to disease progression and medication response.</p>
</div>
</div>
</div>
</div>
</section>
<section id="method">
<div class="container">
<h2 class="section-title">Methodology</h2>
<p class="lead-text">We developed a novel Siamese contrastive network architecture that efficiently extracts precise spatiotemporal motion characteristics of the entire body joints from gait videos recorded with a single smartphone.</p>
<div class="card-grid">
<div class="card">
<div class="card-icon">
<i class="fa-solid fa-network-wired"></i>
</div>
<div class="card-content">
<h3>Siamese Contrastive Network</h3>
<p>Our architecture fuses gait videos recorded from both left and right lateral perspectives during shuttle walks, ensuring accurate identification of lateral motion characteristics.</p>
</div>
</div>
<div class="card">
<div class="card-icon">
<i class="fa-solid fa-child-reaching"></i>
</div>
<div class="card-content">
<h3>Pose Estimation</h3>
<p>We extract body skeletons from smartphone videos using DWPose, with modifications to correct misdetections of leg keypoints for accurate feature extraction.</p>
</div>
</div>
<div class="card">
<div class="card-icon">
<i class="fa-solid fa-magnifying-glass-chart"></i>
</div>
<div class="card-content">
<h3>Interpretable Framework</h3>
<p>Our model analyzes personalized joint impacts on gait impairment severity over walking time, enabling extraction of both traditional and novel digital biomarkers.</p>
</div>
</div>
</div>
</div>
</section>
<section id="results" class="bg-light">
<div class="container">
<h2 class="section-title">Key Results</h2>
<div class="card-grid">
<div class="card">
<div class="card-icon">
<i class="fa-solid fa-chart-column"></i>
</div>
<div class="card-content">
<h3>Severity Assessment</h3>
<p>The model demonstrated high accuracy in predicting gait impairment severity, with performance comparable to clinical experts (F1 score: 0.806).</p>
</div>
</div>
<div class="card">
<div class="card-icon">
<i class="fa-solid fa-pills"></i>
</div>
<div class="card-content">
<h3>Medication Effect Discrimination</h3>
<p>Our framework accurately identified medication-induced changes in gait impairments with 73.68% precision, including subtle responses undetectable by UPDRS.</p>
</div>
</div>
<div class="card">
<div class="card-icon">
<i class="fa-solid fa-dna"></i>
</div>
<div class="card-content">
<h3>Biomarker Discovery</h3>
<p>We identified novel digital biomarkers sensitive to disease progression and medication response, such as linear velocities and accelerations of skeletal joints.</p>
</div>
</div>
</div>
<div class="highlight-box">
<h3><i class="fa-solid fa-check-double"></i> Clinical Validation</h3>
<p>The model's performance was validated against the consensus of three clinical specialists, demonstrating alignment with expert assessments and the ability to handle inter-rater variability.</p>
</div>
</div>
</section>
<section id="demo">
<div class="container">
<h2 class="section-title">Online Assessment System</h2>
<p class="lead-text">We developed an online assessment system (FAGI-PD) that enables home-based, objective, routine assessments of gait impairments in Parkinson's disease using smartphone-recorded videos.</p>
<div class="demo-link">
<a href="https://gaitanalysis.simplaj.fun" class="btn btn-primary" target="_blank" rel="noopener">
<i class="fa-solid fa-play"></i> Try FAGI-PD System
</a>
</div>
<!-- 演示视频 -->
<div class="video-container">
<iframe src="//player.bilibili.com/player.html?bvid=BV1R6kvBcEVC&page=1&high_quality=1&autoplay=0"
scrolling="no"
border="0"
frameborder="no"
framespacing="0"
allowfullscreen="true"
title="FAGI-PD System Demonstration">
</iframe>
</div>
<div class="card-grid">
<div class="card">
<div class="card-icon">
<i class="fa-solid fa-house-medical"></i>
</div>
<div class="card-content">
<h3>Accessible Monitoring</h3>
<p>Enables frequent, convenient assessment of gait impairments in home settings, reducing the need for clinical visits.</p>
</div>
</div>
<div class="card">
<div class="card-icon">
<i class="fa-solid fa-clock-rotate-left"></i>
</div>
<div class="card-content">
<h3>Immediate Feedback</h3>
<p>Provides immediate assessment results and trends over time for patients and clinicians.</p>
</div>
</div>
<div class="card">
<div class="card-icon">
<i class="fa-solid fa-chart-line"></i>
</div>
<div class="card-content">
<h3>Longitudinal Tracking</h3>
<p>Tracks disease progression and treatment response over time, enabling personalized therapy adjustments.</p>
</div>
</div>
</div>
</div>
</section>
<section id="publication">
<div class="container">
<h2 class="section-title">Resources</h2>
<div class="card-grid resources-grid">
<div class="card resource-card">
<div class="card-content">
<h3><i class="fa-brands fa-github"></i> Code & Data</h3>
<p>Code for data preprocessing, model training, and feature extraction available at Code Ocean</p>
<a href="https://codeocean.com/capsule/7700825" class="btn btn-small btn-outline" target="_blank" rel="noopener">Access Code</a>
</div>
</div>
<div class="card resource-card">
<div class="card-content">
<h3><i class="fa-solid fa-laptop-code"></i> Online System</h3>
<p>Try our FAGI-PD system for gait impairment assessment</p>
<a href="https://gaitanalysis.simplaj.fun" class="btn btn-small btn-outline" target="_blank" rel="noopener">Try Demo</a>
</div>
</div>
</div>
</div>
</section>
<footer>
<div class="container">
<div class="footer-content">
<div class="footer-section">
<h3>Contact</h3>
<p><strong>Weiguang Huo, Ph.D.</strong></p>
<p>College of Artificial Intelligence</p>
<p>Nankai University</p>
<a href="mailto:weiguang.huo@nankai.edu.cn">weiguang.huo@nankai.edu.cn</a>
</div>
<div class="footer-section">
<h3>Navigation</h3>
<ul class="footer-links">
<li><a href="#abstract">Abstract</a></li>
<li><a href="#method">Methodology</a></li>
<li><a href="#results">Results</a></li>
<li><a href="#demo">Demo</a></li>
</ul>
</div>
<div class="footer-section">
<h3>Affiliations</h3>
<ul class="footer-links">
<li><a href="https://ai.nankai.edu.cn" target="_blank" rel="noopener">Nankai University</a></li>
<li><a href="#" target="_blank" rel="noopener">Tianjin Huanhu Hospital</a></li>
<li><a href="https://www.imperial.ac.uk" target="_blank" rel="noopener">Imperial College London</a></li>
<li><a href="https://www.sgul.ac.uk" target="_blank" rel="noopener">St George's University of London</a></li>
</ul>
</div>
</div>
<div class="copyright">
<p>© 2025 Deep Learning-enabled Gait Assessment Project. All rights reserved.</p>
</div>
</div>
</footer>
</body>
</html>