Spaces:
Running
Running
File size: 20,789 Bytes
f0e3cf2 dbbe492 f0e3cf2 6866c15 f0e3cf2 dc32ed8 57867e4 f0e3cf2 4ff7fa7 f0e3cf2 6866c15 f0e3cf2 6866c15 f0e3cf2 6866c15 f0e3cf2 6866c15 f0e3cf2 6866c15 f0e3cf2 6866c15 f0e3cf2 6866c15 f0e3cf2 dc32ed8 f0e3cf2 6866c15 f0e3cf2 6866c15 f0e3cf2 dc32ed8 f0e3cf2 47379c6 fa8dbdc 47379c6 f0e3cf2 5b53bf6 f0e3cf2 5b53bf6 f0e3cf2 02d9ab1 8018a28 02d9ab1 8018a28 02d9ab1 5b53bf6 f0e3cf2 5b53bf6 f0e3cf2 5b53bf6 f0e3cf2 5b53bf6 f0e3cf2 77bd4c7 d73dc96 cd4d7e1 d73dc96 f0e3cf2 5b53bf6 d687f6a 5b53bf6 77bd4c7 5b53bf6 f0e3cf2 5b53bf6 2bef06d c8e2547 6278da9 c8e2547 6278da9 2bef06d f0e3cf2 a00f301 f0e3cf2 a00f301 90cfd9a f0e3cf2 a00f301 90cfd9a f0e3cf2 a00f301 90cfd9a f0e3cf2 a00f301 90cfd9a a00f301 90cfd9a a00f301 90cfd9a a00f301 90cfd9a a00f301 90cfd9a a00f301 f0e3cf2 a0c58df f0e3cf2 | 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 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 354 355 356 357 358 359 360 361 362 363 364 365 366 367 368 369 370 371 372 373 374 375 376 377 378 379 380 381 382 383 384 385 386 387 388 389 390 391 392 393 394 395 396 397 398 399 400 401 402 403 404 405 406 407 408 409 410 411 412 413 414 415 416 417 418 419 420 421 422 423 424 425 426 427 428 429 430 431 432 433 434 435 436 437 438 439 440 441 442 443 444 445 446 447 448 449 450 451 | <!DOCTYPE html>
<html>
<head>
<meta charset="utf-8">
<meta name="description"
content="PARTAGES">
<meta name="keywords" content="Generative AI, Health, Medical report">
<meta name="viewport" content="width=device-width, initial-scale=1">
<title>PARTAGES</title>
<!-- CSS -->
<link rel="stylesheet" href="./static/css/bulma.min.css">
<link rel="stylesheet" href="./static/css/bulma-carousel.min.css">
<link rel="stylesheet" href="./static/css/bulma-slider.min.css">
<link rel="stylesheet" href="./static/css/fontawesome.all.min.css">
<!-- Academicons (pour icône arXiv etc.) -->
<link rel="stylesheet" href="https://cdn.jsdelivr.net/gh/jpswalsh/academicons@1/css/academicons.min.css">
<!-- Fonts -->
<link rel="stylesheet" href="https://fonts.googleapis.com/css?family=Google+Sans|Noto+Sans|Castoro">
<!-- Ton style perso -->
<link rel="stylesheet" href="./static/css/index.css">
<script src="https://ajax.googleapis.com/ajax/libs/jquery/3.5.1/jquery.min.js"></script>
<script defer src="./static/js/fontawesome.all.min.js"></script>
<script src="./static/js/bulma-carousel.min.js"></script>
<script src="./static/js/bulma-slider.min.js"></script>
<script src="./static/js/index.js"></script>
</head>
<body>
<section class="hero">
<div class="hero-body">
<div class="container is-max-desktop">
<div class="columns is-centered">
<div class="column has-text-centered">
<h1 class="title is-1 publication-title">PARTAGES</h1>
<div class="is-size-5 publication-authors">
<span class="author-block">
<a href="https://www.health-data-hub.fr/" target="_blank">Health Data Hub</a><sup>1</sup>,</span>
<span class="author-block">
<a href="https://www.lisn.upsaclay.fr/" target="_blank">LISN</a><sup>2</sup>,</span>
<span class="author-block">
<a href="https://www.greyc.fr/" target="_blank">GREYC</a><sup>2</sup>,
</span>
<span class="author-block">
<a href="https://www.liglab.fr/fr" target="_blank">LIG</a><sup>2</sup>,
</span>
<span class="author-block">
<a href="https://www.limics.fr/" target="_blank">LIMICS</a><sup>2</sup>,
</span>
<span class="author-block">
<a href="https://lia.univ-avignon.fr/" target="_blank">LIA</a><sup>1,2</sup>,
</span>
<span class="author-block">
<a href="https://www.ls2n.fr/" target="_blank">LS2N</a><sup>2</sup>
</span>
<span class="author-block">
<a href="https://www.lis-lab.fr/" target="_blank">LIS</a><sup>2</sup>
</span>
<span class="author-block">
<a href="https://www.uness.fr/" target="_blank">UNESS</a><sup>2</sup>
</span>
<span class="author-block">
<a href="https://www.bordeaux-population-health.center/fr/" target="_blank">Bordeaux Population Health</a><sup>2</sup>
</span>
<span class="author-block">
<a href="https://www.cnrs.fr/fr" target="_blank">CNRS</a><sup>2</sup>
</span>
<span class="author-block">
<a href="https://www.inria.fr/fr" target="_blank">INRIA</a><sup>2</sup>
</span>
<span class="author-block">
<a href="https://www.aphp.fr/" target="_blank">AP-HP</a><sup>2</sup>
</span>
<span class="author-block">
<a href="https://www.centreleonberard.fr/" target="_blank">Centre Léon Bérard</a><sup>2</sup>
</span>
<span class="author-block">
<a href="https://www.chu-rouen.fr/" target="_blank">CHU de Rouen</a><sup>2</sup>
</span>
<span class="author-block">
<a href="https://teamhcl.chu-lyon.fr/groupement-de-cooperation-sanitaire-houraa-gcs-houraa" target="_blank">GCS HOURAA</a><sup>2</sup>
</span>
<span class="author-block">
<a href="https://www.gustaveroussy.fr/" target="_blank">Institut Gustave Roussy</a><sup>2</sup>
</span>
<span class="author-block">
<a href="https://curie.fr/" target="_blank">Institut Curie</a><sup>2</sup>
</span>
<span class="author-block">
<a href="https://www.hpsj.fr/" target="_blank">Hôpitaux Saint-Joseph &
Marie-Lannelongue</a><sup>2</sup>
</span>
<span class="author-block">
<a href="https://www.hopital-foch.com/" target="_blank">Hôpital Foch</a><sup>2</sup>
</span>
<span class="author-block">
<a href="https://www.elsan.care/fr" target="_blank">ELSAN</a><sup>2</sup>
</span>
<span class="author-block">
<a href="google.com/search?q=CHU+de+Toulouse&rlz=1C1GCEA_enFR1064FR1064&oq=CHU+de+Toulouse&gs_lcrp=EgZjaHJvbWUyCQgAEEUYORiABDIHCAEQABiABDIHCAIQABiABDIHCAMQABiABDIHCAQQABiABDIHCAUQABiABDIHCAYQABiABDIHCAcQABiABDIHCAgQABiABDIHCAkQABiABNIBBzgzMWowajSoAgCwAgA&sourceid=chrome&ie=UTF-8" target="_blank">CHU de Toulouse</a><sup>2</sup>
</span>
<span class="author-block">
<a href="https://www.chu-bordeaux.fr/" target="_blank">CHU de Bordeaux</a><sup>2</sup>
</span>
<span class="author-block">
<a href="https://www.chu-amiens.fr/" target="_blank">CHU Amiens-Picardi</a><sup>2</sup>
</span>
<span class="author-block">
<a href="https://www.chu-brest.fr/" target="_blank">CHU Brest</a><sup>2</sup>
</span>
<span class="author-block">
<a href="https://www.chu-lille.fr/" target="_blank">CHU de Lille</a><sup>2</sup>
</span>
<span class="author-block">
<a href="https://www.chu-nancy.fr/" target="_blank">CHRU de Nancy</a><sup>2</sup>
</span>
<span class="author-block">
<a href="https://www.chu-reunion.fr/" target="_blank">CHU de La Réunion</a><sup>2</sup>
</span>
<span class="author-block">
<a href="https://www.ramsaysante.fr/" target="_blank">Ramsay Santé</a><sup>2</sup>
</span>
<span class="author-block">
<a href="https://recital.ai/" target="_blank">reciTAL</a><sup>2</sup>
</span>
<span class="author-block">
<a href="https://mistral.ai/fr" target="_blank">Mistral AI</a><sup>2</sup>
</span>
</div>
<!-- Data -->
<div class="column has-text-centered">
<div class="publication-links">
<!-- Code Link. -->
<span class="link-block">
<a href="https://huggingface.co/HealthDataHub/models" target="_blank"
class="external-link button is-normal is-rounded is-dark">
<span class="icon">
<i class="fab fa-github"></i>
</span>
<span>Code</span>
</a>
</span>
<!-- Dataset Link. -->
<span class="link-block">
<a href="https://huggingface.co/HealthDataHub/datasets" target="_blank"
class="external-link button is-normal is-rounded is-dark">
<span class="icon">
<i class="far fa-images"></i>
</span>
<span>Data</span>
</a>
</span>
</div>
</div>
</div>
</div>
</div>
</div>
</section>
<section class="section">
<div class="container is-max-desktop">
<!-- Logo -->
<div class="has-text-centered" style="margin-bottom: 2rem;">
<img src="./static/images/PARTAGES BASELINE_RVB.png" alt="Project logo" style="max-height: 120px;">
</div>
<!--/ Logo -->
<!-- Abstract. -->
<div class="columns is-centered has-text-centered">
<div class="column is-four-fifths">
<h2 class="title is-3">Abstract</h2>
<div class="content has-text-justified">
<p>
<strong>PARTAGES (Advanced Development of Digital Commons for Generative Artificial Intelligence in Healthcare)</strong> is a project coordinated by the Health Data Hub. A winner of the “Digital Commons for Generative AI” call for projects under the France 2030 plan, it aims to accelerate and democratize the use of large language models (LLMs) for healthcare professionals.
</p>
<p>
<strong>Its goal: </strong> to create a national momentum fostering the emergence of open generative AI solutions in healthcare, as well as their adoption within the healthcare ecosystem—whether academic, research-based, or industrial.
</p>
</div>
</div>
</div>
<!--/ Abstract. -->
<!-- Text. -->
<div class="columns is-centered">
<div class="column is-four-fifths">
<h2 class="title is-3">A major national project</h2>
<div class="content has-text-justified">
<p>
With a budget of <strong>€9.4 million</strong>, PARTAGES is supported by a <strong>unique consortium of 32 partners</strong> mobilized on a national scale:
</p>
<ul>
<li><strong>10 research teams</strong> (CNRS, Inria, universities),</li>
<li><strong>20 public and private healthcare institutions</strong> (AP-HP, Institut Curie, Centre Léon Bérard, Ramsay Santé, ELSAN, 12 university hospitals...),</li>
<li><strong>DeepTech companies specializing in AI</strong>, including Mistral and ReciTAL.</li>
</ul>
</div>
</div>
</div>
<!--/ Text. -->
<!-- Image -->
<div class="has-text-centered" style="margin-bottom: 2rem;">
<img src="./static/images/A major national project.png" alt="Project logo" style="width: 300px;">
</div>
<!--/ Image -->
<!-- Text. -->
<div class="columns is-centered">
<div class="column is-four-fifths">
<h2 class="title is-3">A four-step approach</h2>
<div class="content has-text-justified">
<p>
<strong>1. Developing Medical LLMs</strong>
</p>
<p>
Building an open-source corpus of French medical text data to train, evaluate, and distribute multiple open-source medical language models.
</p>
<p>
PARCOMED - PARTAGES Corpus of Open MEdical Documents available <a href="https://huggingface.co/datasets/HealthDataHub/PARCOMED" target="_blank">here</a>
</p>
<p>
PARCOMED research only - PARTAGES Corpus of Open MEdical Documents (for research purpose only) available <a href="https://huggingface.co/datasets/HealthDataHub/PARCOMED" target="_blank">here</a>
</p>
<p>
<strong>2. Creating an open database of fictional medical reports</strong>
</p>
<p>
Creation and release as open data of a unique corpus of over 6,000 fictional medical reports, covering 20 specialties, including annotated reports for use cases. This effort involved more than 100 residents and young physicians and will be used in particular to train specialized models.
</p>
<p>
PARHAF - an open French corpus of human-authored clinical reports of fictional patients available <a href="https://huggingface.co/datasets/HealthDataHub/PARHAF" target="_blank">here</a>
</p>
<p>
<strong>3. Develop models for targeted use cases</strong>
</p>
<p>
Using these resources, PARTAGES is developing seven specialized AI models designed for high-impact use cases in research, innovation, and the healthcare system.
</p>
<p>
<strong>4. Establish a national federated evaluation platform</strong>
</p>
<p>
Development of a sovereign federated evaluation platform, enabling the evaluation of algorithms on real-world data within a secure regulatory framework. It will be deployed in 20 healthcare facilities but may be used by any facility wishing to access it.
</p>
</div>
</div>
</div>
<!--/ Text. -->
<!-- Text. -->
<div class="columns is-centered">
<div class="column is-four-fifths">
<h2 class="title is-3">Practical Use Cases for Healthcare</h2>
<div class="content has-text-justified">
<p>
PARTAGES addresses <strong>eight priority use cases</strong> focused on the analysis, structuring, and generation of medical reports:
</p>
<ul>
<p><i class="fas fa-database" style="margin-right: 8px;"></i><strong> Data augmentation</strong> through the generation of fictitious reports</p>
<p><i class="fas fa-comment-medical" style="margin-right: 8px;"></i><strong> Automatic pseudonymization</strong> of medical reports</p>
<p><i class="fas fa-laptop-code" style="margin-right: 8px;"></i><strong> Automated medical coding (DIM</strong>) based on medical reports</p>
<p><i class="fas fa-brain" style="margin-right: 8px;"></i><strong> Automatic summarization</strong> of medical reports</p>
<p><i class="fas fa-hand-holding-medical" style="margin-right: 8px;"></i><strong> Generation of clinical</strong> cases for medical training</p>
<p><i class="fas fa-dna" style="margin-right: 8px;"></i><strong> Identification of tumor biomarkers</strong> in oncology</p>
<p><i class="fas fa-solid fa-pills" style="margin-right: 8px;"></i><strong> Analysis of treatment response</strong> in oncology</p>
<p><i class="fas fa-bacteria" style="margin-right: 8px;"></i><strong> Automatic detection in infectious diseases</strong>, particularly to combat antibiotic resistance</p>
</ul>
</div>
</div>
</div>
<!--/ Text. -->
<div class="content">
<h2 class="title is-3">An organization structured into 8 work packages</h2>
<ul>
<li>
<i class="fas fa-database" style="margin-right: 8px;"></i>
<strong>WP1 – Data collection and governance:</strong> Structuring, accessing and securing health data.
</li>
<p>
The first work package is dedicated to the overall coordination of the PARTAGES project, as well as the dissemination and promotion of its results. It ensures sound project governance, consistency among the various work packages, and adherence to the timeline, scientific objectives, and regulatory requirements.
</p>
<li>
<i class="fas fa-cogs" style="margin-right: 8px;"></i>
<strong>WP2 – Data preprocessing and harmonization:</strong> Standardizing and preparing datasets for model training.
</li>
<p>
The objective of this second phase is to establish a rigorous methodology for the production and use of project data (primarily fictitious medical records) and to ensure the quality control of all raw datasets used to train the project’s models. It also supports data quality control at the healthcare facility level for model evaluation.
</p>
<li>
<i class="fas fa-brain" style="margin-right: 8px;"></i>
<strong>WP3 – Foundation model development:</strong> Designing and training large-scale models for health data.
</li>
<p>
Work Package 3 is responsible for developing and implementing a common evaluation methodology for foundation models and all use cases, as well as for analyzing the evaluation results.
</p>
<li>
<i class="fas fa-vial" style="margin-right: 8px;"></i>
<strong>WP4 – Use cases in oncology:</strong> Identification of tumor biomarkers and clinical applications.
</li>
<p>
The objective of the fourth work package is to develop all the foundational models that will be used by the use cases, including the fine-tuning of the general-purpose generative LLM for the French-language medical domain, as well as the development of BERT-style encoder models (Bidirectional Encoder Representations from Transformers).
</p>
<li>
<i class="fas fa-bacteria" style="margin-right: 8px;"></i>
<strong>WP5 – Use cases in infectious diseases:</strong> Automatic detection and support for combating antibiotic resistance.
</li>
<p>
Lot No. 5 focuses on establishing the necessary technical infrastructure, including the creation, adaptation, and documentation of the federated validation platform, in which each partner healthcare facility serves as a node.
</p>
<li>
<i class="fas fa-chart-line" style="margin-right: 8px;"></i>
<strong>WP6 – Evaluation and benchmarking:</strong> Assessing model performance, robustness and reproducibility.
</li>
<p>
The goal of the sixth batch is to manage the recruitment and supervision of healthcare professionals (senior residents and junior physicians) for the creation of a corpus of 5,000 fictional patient records and for annotation tasks.
</p>
<li>
<i class="fas fa-network-wired" style="margin-right: 8px;"></i>
<strong>WP7 – Deployment and infrastructure:</strong> Integration into secure environments and operationalization.
</li>
<p>
Lot No. 7 oversees the project's legal matters, including the implementation of the contractual framework and the monitoring of work related to the local use of healthcare facility reports.
</p>
<li>
<i class="fas fa-users" style="margin-right: 8px;"></i>
<strong>WP8 – Dissemination and collaboration:</strong> Promoting open science, sharing resources and engaging stakeholders.
</li>
<p>
This final phase covers the development of models for the specific use cases identified above.
</p>
</ul>
</div>
<div class="content">
<h2 class="title is-3">PARTAGES stakeholders</h2>
<h3 class="title is-5">
<i class="fas fa-university" style="margin-right: 8px;"></i>
Academic and research institutions
</h3>
<ul>
<li>LISN</li>
<li>GREYC</li>
<li>LIG</li>
<li>LIMICS</li>
<li>LIA</li>
<li>LS2N</li>
<li>LIS</li>
<li>UNESS</li>
<li>Bordeaux Population Health</li>
<li>CNRS</li>
<li>INRIA</li>
</ul>
<h3 class="title is-5">
<i class="fas fa-hospital" style="margin-right: 8px;"></i>
Hospitals and healthcare institutions
</h3>
<ul>
<li>AP-HP</li>
<li>Centre Léon Bérard</li>
<li>CHU de Rouen</li>
<li>GCS HOURAA</li>
<li>Institut Gustave Roussy</li>
<li>Institut Curie</li>
<li>Hôpitaux Saint-Joseph & Marie-Lannelongue</li>
<li>Hôpital Foch</li>
<li>CHU de Toulouse</li>
<li>CHU de Bordeaux</li>
<li>CHU Amiens-Picardie</li>
<li>CHU Brest</li>
<li>CHU de Lille</li>
<li>CHRU de Nancy</li>
<li>CHU de La Réunion</li>
</ul>
<h3 class="title is-5">
<i class="fas fa-building" style="margin-right: 8px;"></i>
Industry and private partners
</h3>
<ul>
<li>ELSAN</li>
<li>Ramsay Santé</li>
<li>reciTAL</li>
<li>Mistral AI</li>
</ul>
<h3 class="title is-5">
<i class="fas fa-landmark" style="margin-right: 8px;"></i>
Public coordination
</h3>
<ul>
<li>Health Data Hub</li>
</ul>
</div>
<footer class="footer">
<div class="container">
<div class="content has-text-centered">
<a class="icon-link" target="_blank"
href="./static/videos/nerfies_paper.pdf">
<i class="fas fa-file-pdf"></i>
</a>
<a class="icon-link" href="https://github.com/keunhong" target="_blank" class="external-link" disabled>
<i class="fab fa-github"></i>
</a>
</div>
<div class="columns is-centered">
<div class="column is-8">
<div class="content">
<p>
This website is licensed under a <a rel="license" target="_blank"
href="http://creativecommons.org/licenses/by-sa/4.0/">Creative
Commons Attribution-ShareAlike 4.0 International License</a>.
</p>
<p>
This means you are free to borrow the <a target="_blank"
href="https://github.com/nerfies/nerfies.github.io">source code</a> of this website,
we just ask that you link back to this page in the footer.
Please remember to remove the analytics code included in the header of the website which
you do not want on your website.
</p>
</div>
</div>
</div>
</div>
</footer>
</body>
</html>
|