@@ -11,13 +11,14 @@ <h1>Welcome to the ISIC Challenge</h1>
1111 Beginning in 2016, ISIC has sponsored annual challenges for the computer science
1212 community in association with leading computer vision conferences. Over the years, the
1313 challenges have grown in scale, complexity, and participation, using high-quality
14- human-validated training and test sets of thousands of CC-0-licensed images and metadata.
14+ human-validated training and test sets of thousands of images and metadata.
1515 The earlier challenges were focused primarily on diagnostic accuracy for distinguishing
16- melanoma from other benign and malignant skin lesions. By 2018, the diagnostic performance
16+ dermoscopic photos of melanoma from other skin lesions. By 2018, the diagnostic performance
1717 of the leading algorithms was consistently outperforming clinicians in "reader studies".
18- Additional challenges in 2019 and 2020 were designed to address the out-of-distribution
19- problem and assess the impact of clinical context respectively. The 2020 challenge had
20- 3,300 participants from around the world. In addition to the annual Grand Challenges,
18+ Challenges in 2019 and 2020 were designed to address out-of-distribution classes
19+ and assess the impact of clinical context, respectively.
20+ The challenge in 2024 shifted focus from dermoscopy to 3D total-body-photography and involved 3,400 participants.
21+ In addition to the annual Grand Challenges,
2122 ISIC hosts "live challenges" that allow researchers and students to benchmark the
2223 performance of their algorithms using ISIC images on an ongoing basis.
2324 </ p >
@@ -53,8 +54,8 @@ <h3 id="about-the-isic-archive">About the ISIC Archive</h3>
5354 the largest publicly available collection of quality controlled dermoscopic
5455 images of skin lesions.
5556 </ p >
56- < p > Presently, the ISIC Archive contains over 13 ,000
57- dermoscopic images, which were collected from leading clinical centers
57+ < p > Presently, the ISIC Archive contains over 500 ,000 public
58+ images, which were collected from leading clinical centers
5859 internationally and acquired from a variety of devices within each center. Broad
5960 and international participation in image contribution is designed to insure a
6061 representative clinically relevant sample.
@@ -67,38 +68,6 @@ <h3 id="about-the-isic-archive">About the ISIC Archive</h3>
6768 (i.e., global and focal morphologic elements in the image known to discriminate
6869 between types of skin lesions).
6970 </ p >
70- < p >
71- The software infrastructure of the ISIC
72- Archive is built using the open-source
73- < a href ="https://girder.readthedocs.org/ "
74- >
75- Girder platform
76- < svg
77- xmlns ="http://www.w3.org/2000/svg " aria-hidden ="true " x ="0px " y ="0px " viewBox ="0
78- 0 100 100 " width ="15 " height ="15 " class ="icon outbound ">
79- < path
80- fill ="currentColor "
81- d ="M18.8,85.1h56l0,0c2.2,0,4-1.8,4-4v-32h-8v28h-48v-48h28v-8h-32l0,0c-2.2,0-4,1.8-4,4v56C14.8,83.3,16.6,85.1,18.8,85.1z "> </ path >
82- < polygon fill ="currentColor " points ="45.7,48.7 51.3,54.3 77.2,28.5 77.2,37.2
83- 85.2,37.2 85.2,14.9 62.8,14.9 62.8,22.9 71.5,22.9 "> </ polygon >
84- </ svg >
85- </ a >
86- , and the
87- source code for the Archive itself is
88- < a
89- href ="https://github.com/ImageMarkup/isic-archive ">
90- freely available on GitHub
91- < svg xmlns ="http://www.w3.org/2000/svg "
92- aria-hidden ="true " x ="0px " y ="0px " viewBox ="0 0 100 100 " width ="15 " height ="15 "
93- class ="icon outbound ">
94- < path fill ="currentColor "
95- d ="M18.8,85.1h56l0,0c2.2,0,4-1.8,4-4v-32h-8v28h-48v-48h28v-8h-32l0,0c-2.2,0-4,1.8-4,4v56C14.8,83.3,16.6,85.1,18.8,85.1z "> </ path >
96- < polygon fill ="currentColor " points ="45.7,48.7 51.3,54.3 77.2,28.5 77.2,37.2
97- 85.2,37.2 85.2,14.9 62.8,14.9 62.8,22.9 71.5,22.9 "> </ polygon >
98- </ svg >
99- </ a >
100- .
101- </ p >
10271
10372 < h3 > About Melanoma</ h3 >
10473 < p > Skin cancer is a major public health problem, with over
0 commit comments