brainchop-cli/models.json at main · neuroneural/brainchop-cli · 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
{
  "robust_tissue": {
    "folder": "robust_tissue",
    "type": "sae",
    "n_classes": 3,
    "description": "🧠 robust tissue segmentation using spatial autoencoder architecture",
    "considerations": "quantile normalization"
  },
  "big_robust_tissue": {
    "folder": "big_robust_tissue",
    "type": "sae",
    "n_classes": 18,
    "description": "🧠 large robust tissue segmentation with 18 classes using spatial autoencoder",
    "considerations": "quantile normalization"
  },
  "tissue_fast": {
    "folder": "model5_gw_ae",
    "description": "⚡ our fastest tissue segmentation model that produces gray and white matter labels",
    "considerations": "min/max normalization"
  },
  "subcortical": {
    "folder": "subcortical",
    "description": "🪓 a robust model that can handle clinical scans and other difficult input volumes producing cortical tissue segmentation accompanied by 15 subcortical regions",
    "considerations": "min/max normalization"
  },
  "subcortical-mini": {
    "folder": "model18cls",
    "description": "🪓 a smaller version of subcortical that has only 21 channels instead of 30",
    "considerations": "min/max normalization"
  },
  "DKatlas": {
    "folder": "model24chan104cls",
    "description": "🔪 segments the brain into the Desikan-Killiany Atlas with 104 labels",
    "considerations": "quantile normalization"
  },
  "mindgrab": {
    "folder": "mindgrab",
    "description": "tissue extraction",
    "considerations": "quantile normalization"
  },
  "aparc50": {
    "folder": "model30chan50cls",
    "description": "🧠 cortical parcellation model with 50 anatomical regions",
    "considerations": "quantile normalization (5%-95%)"
  }
}