MIST/index.html at main · BioIntelligence-Lab/MIST · GitHub

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
<!DOCTYPE html>
<html lang="en">
<head>
  <meta charset="UTF-8">
  <meta name="viewport" content="width=device-width, initial-scale=1.0">
  <title>MIST: Medical Image Streaming Toolkit</title>
  <style>
    body {
      font-family: 'Inter', sans-serif;
      color: #222;
      margin: 0;
      background: #fff;
      line-height: 1.6;
    }
    header {
      text-align: center;
      padding: 100px 20px 60px;
      background: #f8f9fb;
    }
    h1 {
      font-size: 2.8em;
      font-weight: 700;
      margin-bottom: 10px;
    }
    h2 {
      text-align: center;
      font-weight: 600;
      margin-top: 80px;
    }
    p.subtitle {
      font-size: 1.3em;
      color: #555;
      margin-bottom: 40px;
    }
    .btn-container {
      margin-top: 25px;
    }
    .btn {
      display: inline-block;
      background: #0055cc;
      color: white;
      text-decoration: none;
      padding: 10px 20px;
      border-radius: 6px;
      margin: 6px;
      font-weight: 500;
      transition: background 0.3s;
    }
    .btn:hover { background: #003f99; }
    section {
      width: 85%;
      max-width: 950px;
      margin: auto;
      padding: 40px 0;
    }
    table {
      width: 100%;
      border-collapse: collapse;
      margin-top: 20px;
    }
    table th, table td {
      border-bottom: 1px solid #eee;
      text-align: left;
      padding: 10px;
    }
    footer {
      text-align: center;
      background: #f1f3f6;
      padding: 40px 0;
      font-size: 0.9em;
      color: #777;
      margin-top: 60px;
    }
    .highlight {
      background: #e9f2ff;
      padding: 20px;
      border-left: 4px solid #0055cc;
      margin: 25px 0;
      border-radius: 6px;
    }
    .center { text-align: center; }
    img.diagram {
      display: block;
      max-width: 100%;
      margin: 20px auto;
      border-radius: 8px;
      box-shadow: 0 2px 8px rgba(0,0,0,0.1);
    }
  </style>
</head>
<body>
  <header>
    <h1>MIST: Medical Image Streaming Toolkit</h1>
    <p class="subtitle">A unified framework for intelligent, progressive, and resource-efficient medical image streaming.</p>
    <div class="btn-container">
      <a href="https://doi.org/10.1007/s10278-024-01173-z" class="btn">Journal Paper (JIIM 2024)</a>
      <a href="https://openreview.net/forum?id=IIuULGCHLY&noteId=IIuULGCHLY" class="btn">Conference Paper (MIDL 2025)</a>
      <a href="https://patentscope.wipo.int/search/en/detail.jsf?docId=WO2024233969" class="btn">Patent</a>
      <a href="https://github.com/BioIntelligence-Lab/MIST" class="btn">GitHub</a>
    </div>
  </header>

  <section id="architecture">
    <h2>Architecture</h2>
    <p class="center">A schematic representation of MIST's hierarchical and adaptive streaming pipeline.</p>
    <img src="https://raw.githubusercontent.com/BioIntelligence-Lab/MIST/main/assets/overview.png" alt="MIST Architecture Overview" class="diagram">
  </section>

  <section id="overview">
<p><strong>MIST</strong> and <strong>ISLE</strong> represent two complementary frameworks addressing the challenges of large-scale medical imaging datasets and AI-driven inference systems.</p>
<h2>Overview</h2>
<h3>MIST: Medical Imaging Streaming Toolkit</h3>
<ul class="highlights">
<li><strong>Challenge:</strong> Large-scale imaging datasets require significant storage and bandwidth, limiting accessibility for AI research and clinical deployment.</li>
<li><strong>MIST Solution:</strong> A format-agnostic database enabling streaming of medical images at multiple resolutions and formats from a single high-resolution copy.</li>
<li><strong>Evaluation:</strong> Tested across eight diverse datasets (CT, MRI, X-ray) covering multiple modalities and file formats.</li>
<li><strong>Results:</strong> Reduced storage and bandwidth requirements without impacting image quality or downstream deep learning performance.</li>
<li><strong>Impact:</strong> Creates a data-efficient, format-agnostic platform that reduces barriers to AI research in medical imaging.</li>
</ul>


<h3>ISLE: Intelligent Streaming for AI Inference</h3>
<ul class="highlights">
<li><strong>Motivation:</strong> Growing adoption of AI systems in radiology is increasing demands for bandwidth and computational resources.</li>
<li><strong>ISLE Framework:</strong> An intelligent streaming method inspired by video-on-demand platforms to deliver only the resolution needed for AI inference using progressive encoding.</li>
<li><strong>Results (Classification):</strong> Reduced transmission by ≥90% and decoding time by ≥87%</li>
<li><strong>Results (Segmentation):</strong> Reduced transmission by ≥77% and decoding time by ≥89%</li>
<li><strong>Performance:</strong> No impact on diagnostic performance (all P > 0.05).</li>
<li><strong>Impact:</strong> Improves data and computational efficiency for AI deployment in clinical environments without compromising diagnostic accuracy.</li>
</ul>
</section>

  <section id="opensource">
    <h2>Open-Source Tools</h2>
    <table>
      <tr><th>Component</th><th>Description</th><th>Repository</th></tr>
      <tr><td>MIST</td><td>Core streaming and dataset management framework</td><td><a href="https://github.com/BioIntelligence-Lab/MIST">GitHub</a></td></tr>
      <tr><td>IntelligentStreaming</td><td>AI-aware streaming for real-time inference</td><td><a href="https://github.com/BioIntelligence-Lab/IntelligentStreaming">GitHub</a></td></tr>
      <tr><td>OpenJPHpy</td><td>Python interface for HTJ2K codec</td><td><a href="https://github.com/BioIntelligence-Lab/openjphpy">GitHub</a></td></tr>
    </table>
  </section>

  <section id="patent">
    <h2>Patents</h2>
    <p><strong>Patent:</strong> <a href="https://patents.google.com/patent/WO2024233969A1/en?oq=WO2024233969A1">WO2024233969A1</a> — <em>Systems and methods for high-throughput analysis for graphical data</em></p>
    <p><strong>Filed by:</strong> University of Maryland Baltimore <br>
       <strong>Inventors:</strong> Vishwa S. Parekh, Pranav Kulkarni, Adway Kanhere, Paul H. Yi, Eliot L. Siegel</p>
  </section>

  <section id="publications">
    <h2>Publications</h2>
    <p>Kulkarni P., Kanhere A., Siegel E.L., Yi P.H., Parekh V.S. <em>ISLE: An Intelligent Streaming Framework for High-Throughput AI Inference in Medical Imaging.</em> <strong>Journal of Imaging Informatics in Medicine. 2024 Dec;37(6):3250-63. <a href="https://doi.org/10.1007/s10278-024-01173-z">DOI</a></p>
    <p>Kulkarni P., Kanhere A., Siegel E., Yi P., Parekh V.S. <em>Towards Resource-Efficient Streaming of Large-Scale Medical Image Datasets for Deep Learning.</em> <strong>Medical Imaging with Deep Learning (MIDL)</strong> (2025). <a href="https://openreview.net/forum?id=IIuULGCHLY&noteId=IIuULGCHLY">OpenReview</a></p>
  </section>

  <section id="contact">
    <h2>Contact</h2>
    <p class="center"><strong>Dr. Vishwa S. Parekh</strong><br>
    UTHealth Houston<br>
    <a href="mailto:vishwa.s.parekh@uth.tmc.edu">vishwa.s.parekh@uth.tmc.edu</a></p>
  </section>

  <footer>
    <p>© 2025 BioIntelligence Research Lab · UTHealth Houston</p>
  </footer>
</body>
</html>