Vendor datasheets in the VendorDatasheet format; add TotalSegmentator training-data datasheets
Browse files
manifest.json
CHANGED
|
@@ -25,6 +25,6 @@
|
|
| 25 |
"tables/pipeline/embeddings.parquet": "619cb89c967ad84a05257bb4b32d095954f964de42cbb1772bf12491f8e3d6fc",
|
| 26 |
"tables/pipeline/features.parquet": "a28c242ac9b3d8e2ef05e8eeea6a1c4af1abde6525c61855553ba4e81e4fd30b",
|
| 27 |
"tables/pipeline/score.parquet": "8fd35e01873fcd0898009ce94459e0e9b83e2f1017d8049b653c78b39d18cfe1",
|
| 28 |
-
"vendor_datasheet/aaa-baseline-v0.json": "
|
| 29 |
}
|
| 30 |
}
|
|
|
|
| 25 |
"tables/pipeline/embeddings.parquet": "619cb89c967ad84a05257bb4b32d095954f964de42cbb1772bf12491f8e3d6fc",
|
| 26 |
"tables/pipeline/features.parquet": "a28c242ac9b3d8e2ef05e8eeea6a1c4af1abde6525c61855553ba4e81e4fd30b",
|
| 27 |
"tables/pipeline/score.parquet": "8fd35e01873fcd0898009ce94459e0e9b83e2f1017d8049b653c78b39d18cfe1",
|
| 28 |
+
"vendor_datasheet/aaa-baseline-v0.json": "f9a5441b4dc4c29c0e48ac0c577c1c99d27c73655bde39071f477d0dd3a84327"
|
| 29 |
}
|
| 30 |
}
|
vendor_datasheet/aaa-baseline-v0.json
CHANGED
|
@@ -1,76 +1,217 @@
|
|
| 1 |
{
|
| 2 |
-
"
|
|
|
|
|
|
|
| 3 |
"model": {
|
| 4 |
"name": "AAA baseline v0",
|
| 5 |
-
"
|
| 6 |
-
"
|
| 7 |
},
|
| 8 |
-
"
|
| 9 |
-
|
| 10 |
-
"
|
| 11 |
-
"
|
| 12 |
-
"
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 13 |
"n": 333,
|
| 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 |
"n": 330,
|
| 46 |
-
"
|
| 47 |
-
|
| 48 |
-
|
| 49 |
-
|
| 50 |
-
|
| 51 |
-
|
| 52 |
],
|
| 53 |
-
"
|
| 54 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 55 |
},
|
| 56 |
-
|
| 57 |
-
|
|
|
|
| 58 |
"n": 333,
|
| 59 |
-
"
|
| 60 |
-
|
| 61 |
-
|
| 62 |
-
|
| 63 |
-
|
| 64 |
-
|
| 65 |
],
|
| 66 |
-
"
|
| 67 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 68 |
}
|
| 69 |
-
|
| 70 |
"not_disclosed": [
|
| 71 |
"individual cases",
|
| 72 |
"joint distributions",
|
| 73 |
"embeddings",
|
| 74 |
-
"protocol names"
|
|
|
|
|
|
|
|
|
|
|
|
|
| 75 |
]
|
| 76 |
}
|
|
|
|
| 1 |
{
|
| 2 |
+
"schema_version": "0.1.0",
|
| 3 |
+
"datasheet_id": "aaa-baseline-v0",
|
| 4 |
+
"title": "AAA baseline v0: training data summary",
|
| 5 |
"model": {
|
| 6 |
"name": "AAA baseline v0",
|
| 7 |
+
"version": "0",
|
| 8 |
+
"intended_use": "abdominal aortic aneurysm detection on CT; infrarenal max outer-wall diameter (centerline-orthogonal) >= 30 mm on the TotalSegmentator aorta mask"
|
| 9 |
},
|
| 10 |
+
"describes": "training",
|
| 11 |
+
"scope": {
|
| 12 |
+
"description": "Adult abdominal CT showing the infrarenal aorta, iliac arteries and kidneys, from the TotalSegmentator dataset (University Hospital Basel, 10 coded institutes)",
|
| 13 |
+
"modality": "CT",
|
| 14 |
+
"body_regions": [
|
| 15 |
+
"abdomen",
|
| 16 |
+
"pelvis"
|
| 17 |
+
],
|
| 18 |
+
"countries": [
|
| 19 |
+
"Switzerland"
|
| 20 |
+
],
|
| 21 |
+
"inclusion": [
|
| 22 |
+
"aorta, iliac arteries and kidneys in the field of view",
|
| 23 |
+
"measurable infrarenal aortic diameter"
|
| 24 |
+
]
|
| 25 |
+
},
|
| 26 |
+
"n_scans": 333,
|
| 27 |
+
"n_patients": 333,
|
| 28 |
+
"sources": [
|
| 29 |
+
{
|
| 30 |
+
"source_id": "slicepoint-aaa-ct",
|
| 31 |
+
"kind": "dataset_release",
|
| 32 |
+
"citation": "SlicePoint AAA CT, vendor_train cohort (Hugging Face, slicepoint/slicepoint-aaa-ct)",
|
| 33 |
+
"url": "https://huggingface.co/datasets/slicepoint/slicepoint-aaa-ct"
|
| 34 |
+
}
|
| 35 |
+
],
|
| 36 |
+
"variables": [
|
| 37 |
+
{
|
| 38 |
+
"variable": "age_years",
|
| 39 |
+
"label": "Age",
|
| 40 |
"n": 333,
|
| 41 |
+
"sources": [
|
| 42 |
+
{
|
| 43 |
+
"source_id": "slicepoint-aaa-ct",
|
| 44 |
+
"value_source": "computed",
|
| 45 |
+
"locator": "cohorts/vendor_train"
|
| 46 |
+
}
|
| 47 |
],
|
| 48 |
+
"summary": {
|
| 49 |
+
"kind": "continuous",
|
| 50 |
+
"unit": "years",
|
| 51 |
+
"mean": 63.45,
|
| 52 |
+
"sd": 15.14,
|
| 53 |
+
"median": 65.0,
|
| 54 |
+
"q1": 54.0,
|
| 55 |
+
"q3": 74.0,
|
| 56 |
+
"min": 19.0,
|
| 57 |
+
"max": 98.0
|
| 58 |
+
}
|
| 59 |
},
|
| 60 |
+
{
|
| 61 |
+
"variable": "sex",
|
| 62 |
+
"label": "Sex",
|
| 63 |
+
"n": 333,
|
| 64 |
+
"sources": [
|
| 65 |
+
{
|
| 66 |
+
"source_id": "slicepoint-aaa-ct",
|
| 67 |
+
"value_source": "computed",
|
| 68 |
+
"locator": "cohorts/vendor_train"
|
| 69 |
+
}
|
| 70 |
+
],
|
| 71 |
+
"summary": {
|
| 72 |
+
"kind": "categorical",
|
| 73 |
+
"counts": {
|
| 74 |
+
"F": 142.0,
|
| 75 |
+
"M": 191.0
|
| 76 |
+
}
|
| 77 |
+
}
|
| 78 |
},
|
| 79 |
+
{
|
| 80 |
+
"variable": "scanner_vendor",
|
| 81 |
+
"label": "Scanner manufacturer",
|
| 82 |
+
"n": 333,
|
| 83 |
+
"sources": [
|
| 84 |
+
{
|
| 85 |
+
"source_id": "slicepoint-aaa-ct",
|
| 86 |
+
"value_source": "computed",
|
| 87 |
+
"locator": "cohorts/vendor_train"
|
| 88 |
+
}
|
| 89 |
+
],
|
| 90 |
+
"summary": {
|
| 91 |
+
"kind": "categorical",
|
| 92 |
+
"counts": {
|
| 93 |
+
"ge": 15.0,
|
| 94 |
+
"philips": 29.0,
|
| 95 |
+
"siemens": 286.0,
|
| 96 |
+
"unknown": 3.0
|
| 97 |
+
}
|
| 98 |
+
}
|
| 99 |
},
|
| 100 |
+
{
|
| 101 |
+
"variable": "contrast_phase",
|
| 102 |
+
"label": "Contrast phase",
|
| 103 |
+
"n": 333,
|
| 104 |
+
"sources": [
|
| 105 |
+
{
|
| 106 |
+
"source_id": "slicepoint-aaa-ct",
|
| 107 |
+
"value_source": "computed",
|
| 108 |
+
"locator": "cohorts/vendor_train"
|
| 109 |
+
}
|
| 110 |
+
],
|
| 111 |
+
"summary": {
|
| 112 |
+
"kind": "categorical",
|
| 113 |
+
"counts": {
|
| 114 |
+
"arterial": 72.0,
|
| 115 |
+
"non-contrast": 62.0,
|
| 116 |
+
"venous": 199.0
|
| 117 |
+
}
|
| 118 |
+
}
|
| 119 |
},
|
| 120 |
+
{
|
| 121 |
+
"variable": "slice_thickness_class",
|
| 122 |
+
"label": "Slice thickness class",
|
| 123 |
+
"n": 333,
|
| 124 |
+
"sources": [
|
| 125 |
+
{
|
| 126 |
+
"source_id": "slicepoint-aaa-ct",
|
| 127 |
+
"value_source": "computed",
|
| 128 |
+
"locator": "cohorts/vendor_train"
|
| 129 |
+
}
|
| 130 |
+
],
|
| 131 |
+
"summary": {
|
| 132 |
+
"kind": "categorical",
|
| 133 |
+
"counts": {
|
| 134 |
+
"standard": 103.0,
|
| 135 |
+
"thick": 48.0,
|
| 136 |
+
"thin": 182.0
|
| 137 |
+
}
|
| 138 |
+
}
|
| 139 |
},
|
| 140 |
+
{
|
| 141 |
+
"variable": "kvp",
|
| 142 |
+
"label": "Tube voltage",
|
| 143 |
"n": 330,
|
| 144 |
+
"sources": [
|
| 145 |
+
{
|
| 146 |
+
"source_id": "slicepoint-aaa-ct",
|
| 147 |
+
"value_source": "computed",
|
| 148 |
+
"locator": "cohorts/vendor_train"
|
| 149 |
+
}
|
| 150 |
],
|
| 151 |
+
"summary": {
|
| 152 |
+
"kind": "continuous",
|
| 153 |
+
"unit": "kV",
|
| 154 |
+
"mean": 114.0,
|
| 155 |
+
"sd": 13.65,
|
| 156 |
+
"median": 120.0,
|
| 157 |
+
"q1": 100.0,
|
| 158 |
+
"q3": 120.0,
|
| 159 |
+
"min": 80.0,
|
| 160 |
+
"max": 150.0
|
| 161 |
+
},
|
| 162 |
+
"n_missing": 3
|
| 163 |
},
|
| 164 |
+
{
|
| 165 |
+
"variable": "aaa",
|
| 166 |
+
"label": "AAA (diameter >= 30 mm)",
|
| 167 |
"n": 333,
|
| 168 |
+
"sources": [
|
| 169 |
+
{
|
| 170 |
+
"source_id": "slicepoint-aaa-ct",
|
| 171 |
+
"value_source": "computed",
|
| 172 |
+
"locator": "cohorts/vendor_train"
|
| 173 |
+
}
|
| 174 |
],
|
| 175 |
+
"summary": {
|
| 176 |
+
"kind": "categorical",
|
| 177 |
+
"counts": {
|
| 178 |
+
"false": 297.0,
|
| 179 |
+
"true": 36.0
|
| 180 |
+
}
|
| 181 |
+
}
|
| 182 |
+
},
|
| 183 |
+
{
|
| 184 |
+
"variable": "max_aortic_diameter_mm",
|
| 185 |
+
"label": "Max infrarenal aortic diameter",
|
| 186 |
+
"n": 333,
|
| 187 |
+
"sources": [
|
| 188 |
+
{
|
| 189 |
+
"source_id": "slicepoint-aaa-ct",
|
| 190 |
+
"value_source": "computed",
|
| 191 |
+
"locator": "cohorts/vendor_train"
|
| 192 |
+
}
|
| 193 |
+
],
|
| 194 |
+
"summary": {
|
| 195 |
+
"kind": "continuous",
|
| 196 |
+
"unit": "mm",
|
| 197 |
+
"mean": 23.42,
|
| 198 |
+
"sd": 8.87,
|
| 199 |
+
"median": 21.02,
|
| 200 |
+
"q1": 19.21,
|
| 201 |
+
"q3": 23.9,
|
| 202 |
+
"min": 12.5,
|
| 203 |
+
"max": 72.76
|
| 204 |
+
}
|
| 205 |
}
|
| 206 |
+
],
|
| 207 |
"not_disclosed": [
|
| 208 |
"individual cases",
|
| 209 |
"joint distributions",
|
| 210 |
"embeddings",
|
| 211 |
+
"protocol names",
|
| 212 |
+
"race and ethnicity (not recorded in the source)"
|
| 213 |
+
],
|
| 214 |
+
"notes": [
|
| 215 |
+
"Synthetic example: a summary of the vendor_train cohort written as a vendor might disclose it."
|
| 216 |
]
|
| 217 |
}
|
vendor_datasheet/totalsegmentator-v2-train-abdominal.json
ADDED
|
@@ -0,0 +1,366 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"schema_version": "0.1.0",
|
| 3 |
+
"datasheet_id": "totalsegmentator-v2-train-abdominal",
|
| 4 |
+
"title": "TotalSegmentator, training data (abdominal AAA subset)",
|
| 5 |
+
"datasheet_version": "TotalSegmentator dataset v2.0.1",
|
| 6 |
+
"model": {
|
| 7 |
+
"name": "TotalSegmentator",
|
| 8 |
+
"version": "2.x, task total",
|
| 9 |
+
"manufacturer": "University Hospital Basel",
|
| 10 |
+
"intended_use": "Segmentation of 104+ anatomic structures in CT (research use)",
|
| 11 |
+
"regulatory_ids": []
|
| 12 |
+
},
|
| 13 |
+
"describes": "training",
|
| 14 |
+
"scope": {
|
| 15 |
+
"description": "The training-split scans in SlicePoint AAA CT: adult abdominal CT showing the infrarenal aorta, iliac arteries and kidneys, University Hospital Basel",
|
| 16 |
+
"modality": "CT",
|
| 17 |
+
"body_regions": [
|
| 18 |
+
"abdomen",
|
| 19 |
+
"pelvis"
|
| 20 |
+
],
|
| 21 |
+
"countries": [
|
| 22 |
+
"Switzerland"
|
| 23 |
+
],
|
| 24 |
+
"inclusion": [
|
| 25 |
+
"aorta, iliac arteries and kidneys in the field of view",
|
| 26 |
+
"measurable infrarenal aortic diameter"
|
| 27 |
+
],
|
| 28 |
+
"exclusion": []
|
| 29 |
+
},
|
| 30 |
+
"n_scans": 481,
|
| 31 |
+
"sources": [
|
| 32 |
+
{
|
| 33 |
+
"source_id": "ts-v201-meta",
|
| 34 |
+
"kind": "dataset_release",
|
| 35 |
+
"citation": "Wasserthal J. et al. TotalSegmentator dataset v2.0.1, meta.csv (Zenodo)",
|
| 36 |
+
"url": "https://doi.org/10.5281/zenodo.10047292"
|
| 37 |
+
},
|
| 38 |
+
{
|
| 39 |
+
"source_id": "ts-v300-meta",
|
| 40 |
+
"kind": "dataset_release",
|
| 41 |
+
"citation": "TotalSegmentator dataset v3.0.0, meta.csv (Zenodo): acquisition slice thickness",
|
| 42 |
+
"url": "https://doi.org/10.5281/zenodo.22688904"
|
| 43 |
+
},
|
| 44 |
+
{
|
| 45 |
+
"source_id": "ts-paper",
|
| 46 |
+
"kind": "publication",
|
| 47 |
+
"citation": "Wasserthal J. et al. TotalSegmentator: Robust Segmentation of 104 Anatomic Structures in CT Images. Radiology: Artificial Intelligence 2023;5(5)",
|
| 48 |
+
"doi": "10.1148/ryai.230024"
|
| 49 |
+
}
|
| 50 |
+
],
|
| 51 |
+
"variables": [
|
| 52 |
+
{
|
| 53 |
+
"variable": "age_years",
|
| 54 |
+
"label": "Age",
|
| 55 |
+
"summary": {
|
| 56 |
+
"kind": "continuous",
|
| 57 |
+
"unit": "years",
|
| 58 |
+
"mean": 63.703,
|
| 59 |
+
"sd": 15.11,
|
| 60 |
+
"median": 66.0,
|
| 61 |
+
"q1": 54.0,
|
| 62 |
+
"q3": 74.0,
|
| 63 |
+
"min": 19.0,
|
| 64 |
+
"max": 94.0,
|
| 65 |
+
"histogram": {
|
| 66 |
+
"edges": [
|
| 67 |
+
10.0,
|
| 68 |
+
20.0,
|
| 69 |
+
30.0,
|
| 70 |
+
40.0,
|
| 71 |
+
50.0,
|
| 72 |
+
60.0,
|
| 73 |
+
70.0,
|
| 74 |
+
80.0,
|
| 75 |
+
90.0,
|
| 76 |
+
100.0,
|
| 77 |
+
110.0
|
| 78 |
+
],
|
| 79 |
+
"counts": [
|
| 80 |
+
1.0,
|
| 81 |
+
11.0,
|
| 82 |
+
22.0,
|
| 83 |
+
48.0,
|
| 84 |
+
89.0,
|
| 85 |
+
119.0,
|
| 86 |
+
124.0,
|
| 87 |
+
58.0,
|
| 88 |
+
9.0,
|
| 89 |
+
0.0
|
| 90 |
+
]
|
| 91 |
+
}
|
| 92 |
+
},
|
| 93 |
+
"n": 481,
|
| 94 |
+
"sources": [
|
| 95 |
+
{
|
| 96 |
+
"source_id": "ts-v201-meta",
|
| 97 |
+
"value_source": "computed",
|
| 98 |
+
"locator": "meta.csv: age"
|
| 99 |
+
}
|
| 100 |
+
]
|
| 101 |
+
},
|
| 102 |
+
{
|
| 103 |
+
"variable": "sex",
|
| 104 |
+
"label": "Sex",
|
| 105 |
+
"summary": {
|
| 106 |
+
"kind": "categorical",
|
| 107 |
+
"counts": {
|
| 108 |
+
"M": 277.0,
|
| 109 |
+
"F": 204.0
|
| 110 |
+
}
|
| 111 |
+
},
|
| 112 |
+
"n": 481,
|
| 113 |
+
"sources": [
|
| 114 |
+
{
|
| 115 |
+
"source_id": "ts-v201-meta",
|
| 116 |
+
"value_source": "computed",
|
| 117 |
+
"locator": "meta.csv: gender"
|
| 118 |
+
}
|
| 119 |
+
]
|
| 120 |
+
},
|
| 121 |
+
{
|
| 122 |
+
"variable": "scanner_vendor",
|
| 123 |
+
"label": "Scanner manufacturer",
|
| 124 |
+
"summary": {
|
| 125 |
+
"kind": "categorical",
|
| 126 |
+
"counts": {
|
| 127 |
+
"siemens": 421.0,
|
| 128 |
+
"philips": 42.0,
|
| 129 |
+
"ge": 18.0
|
| 130 |
+
}
|
| 131 |
+
},
|
| 132 |
+
"n": 481,
|
| 133 |
+
"sources": [
|
| 134 |
+
{
|
| 135 |
+
"source_id": "ts-v201-meta",
|
| 136 |
+
"value_source": "computed",
|
| 137 |
+
"locator": "meta.csv: manufacturer"
|
| 138 |
+
}
|
| 139 |
+
]
|
| 140 |
+
},
|
| 141 |
+
{
|
| 142 |
+
"variable": "scanner_model",
|
| 143 |
+
"label": "Scanner model",
|
| 144 |
+
"summary": {
|
| 145 |
+
"kind": "categorical",
|
| 146 |
+
"counts": {
|
| 147 |
+
"somatom definition flash": 127.0,
|
| 148 |
+
"somatom definition as+": 71.0,
|
| 149 |
+
"somatom force": 58.0,
|
| 150 |
+
"emotion 16": 57.0,
|
| 151 |
+
"somatom definition edge": 53.0,
|
| 152 |
+
"ingenuity core 128": 33.0,
|
| 153 |
+
"sensation 16": 24.0,
|
| 154 |
+
"sensation 64": 17.0,
|
| 155 |
+
"lightspeed vct": 17.0,
|
| 156 |
+
"somatom go.top": 12.0,
|
| 157 |
+
"iqon - spectral ct": 5.0,
|
| 158 |
+
"brilliance 64": 4.0,
|
| 159 |
+
"somatom definition as": 2.0,
|
| 160 |
+
"discovery ste": 1.0
|
| 161 |
+
}
|
| 162 |
+
},
|
| 163 |
+
"n": 481,
|
| 164 |
+
"sources": [
|
| 165 |
+
{
|
| 166 |
+
"source_id": "ts-v201-meta",
|
| 167 |
+
"value_source": "computed",
|
| 168 |
+
"locator": "meta.csv: scanner_model"
|
| 169 |
+
}
|
| 170 |
+
]
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"variable": "kvp",
|
| 174 |
+
"label": "Tube voltage",
|
| 175 |
+
"summary": {
|
| 176 |
+
"kind": "continuous",
|
| 177 |
+
"unit": "kV",
|
| 178 |
+
"mean": 113.368,
|
| 179 |
+
"sd": 12.97,
|
| 180 |
+
"median": 120.0,
|
| 181 |
+
"q1": 100.0,
|
| 182 |
+
"q3": 120.0,
|
| 183 |
+
"min": 80.0,
|
| 184 |
+
"max": 150.0,
|
| 185 |
+
"histogram": {
|
| 186 |
+
"edges": [
|
| 187 |
+
65.0,
|
| 188 |
+
75.0,
|
| 189 |
+
85.0,
|
| 190 |
+
95.0,
|
| 191 |
+
105.0,
|
| 192 |
+
115.0,
|
| 193 |
+
125.0,
|
| 194 |
+
135.0,
|
| 195 |
+
145.0,
|
| 196 |
+
155.0,
|
| 197 |
+
165.0
|
| 198 |
+
],
|
| 199 |
+
"counts": [
|
| 200 |
+
0.0,
|
| 201 |
+
12.0,
|
| 202 |
+
6.0,
|
| 203 |
+
147.0,
|
| 204 |
+
37.0,
|
| 205 |
+
228.0,
|
| 206 |
+
32.0,
|
| 207 |
+
11.0,
|
| 208 |
+
8.0,
|
| 209 |
+
0.0
|
| 210 |
+
]
|
| 211 |
+
}
|
| 212 |
+
},
|
| 213 |
+
"n": 481,
|
| 214 |
+
"sources": [
|
| 215 |
+
{
|
| 216 |
+
"source_id": "ts-v201-meta",
|
| 217 |
+
"value_source": "computed",
|
| 218 |
+
"locator": "meta.csv: kvp"
|
| 219 |
+
}
|
| 220 |
+
]
|
| 221 |
+
},
|
| 222 |
+
{
|
| 223 |
+
"variable": "slice_thickness_mm",
|
| 224 |
+
"label": "Slice thickness (acquisition)",
|
| 225 |
+
"summary": {
|
| 226 |
+
"kind": "continuous",
|
| 227 |
+
"unit": "mm",
|
| 228 |
+
"mean": 2.193,
|
| 229 |
+
"sd": 1.214,
|
| 230 |
+
"median": 1.5,
|
| 231 |
+
"q1": 1.5,
|
| 232 |
+
"q3": 3.0,
|
| 233 |
+
"min": 0.6,
|
| 234 |
+
"max": 5.0,
|
| 235 |
+
"histogram": {
|
| 236 |
+
"edges": [
|
| 237 |
+
0.0,
|
| 238 |
+
0.7,
|
| 239 |
+
0.95,
|
| 240 |
+
1.1,
|
| 241 |
+
1.4,
|
| 242 |
+
1.75,
|
| 243 |
+
2.25,
|
| 244 |
+
2.75,
|
| 245 |
+
3.4,
|
| 246 |
+
4.5,
|
| 247 |
+
6.0
|
| 248 |
+
],
|
| 249 |
+
"counts": [
|
| 250 |
+
2.0,
|
| 251 |
+
6.0,
|
| 252 |
+
25.0,
|
| 253 |
+
16.0,
|
| 254 |
+
226.0,
|
| 255 |
+
76.0,
|
| 256 |
+
0.0,
|
| 257 |
+
68.0,
|
| 258 |
+
1.0,
|
| 259 |
+
61.0
|
| 260 |
+
]
|
| 261 |
+
}
|
| 262 |
+
},
|
| 263 |
+
"n": 481,
|
| 264 |
+
"sources": [
|
| 265 |
+
{
|
| 266 |
+
"source_id": "ts-v300-meta",
|
| 267 |
+
"value_source": "computed",
|
| 268 |
+
"locator": "meta.csv: slice_thickness",
|
| 269 |
+
"note": "v3.0.0 metadata joined on image_id"
|
| 270 |
+
}
|
| 271 |
+
]
|
| 272 |
+
},
|
| 273 |
+
{
|
| 274 |
+
"variable": "study_type",
|
| 275 |
+
"label": "Study type",
|
| 276 |
+
"summary": {
|
| 277 |
+
"kind": "categorical",
|
| 278 |
+
"counts": {
|
| 279 |
+
"ct thorax-abdomen-pelvis": 95.0,
|
| 280 |
+
"ct neck-thorax-abdomen-pelvis": 77.0,
|
| 281 |
+
"ct abdomen-pelvis": 60.0,
|
| 282 |
+
"ct abdomen": 46.0,
|
| 283 |
+
"ct polytrauma": 36.0,
|
| 284 |
+
"ct thorax-abdomen": 28.0,
|
| 285 |
+
"ct angiography neck-thx-abd-pelvis-leg": 21.0,
|
| 286 |
+
"ct thorax-chest": 16.0,
|
| 287 |
+
"ct aortic valve": 15.0,
|
| 288 |
+
"ct angiography thorax-abdomen-pelvis": 14.0,
|
| 289 |
+
"ct spine": 12.0,
|
| 290 |
+
"ct angiography abdomen-pelvis": 12.0,
|
| 291 |
+
"ct pelvis": 11.0,
|
| 292 |
+
"ct angiography pelvis-leg": 10.0,
|
| 293 |
+
"ct angiography abdomen-pelvis-leg": 9.0,
|
| 294 |
+
"other": 19.0
|
| 295 |
+
}
|
| 296 |
+
},
|
| 297 |
+
"n": 481,
|
| 298 |
+
"sources": [
|
| 299 |
+
{
|
| 300 |
+
"source_id": "ts-v201-meta",
|
| 301 |
+
"value_source": "computed",
|
| 302 |
+
"locator": "meta.csv: study_type"
|
| 303 |
+
}
|
| 304 |
+
]
|
| 305 |
+
},
|
| 306 |
+
{
|
| 307 |
+
"variable": "site",
|
| 308 |
+
"label": "Institute (coded)",
|
| 309 |
+
"summary": {
|
| 310 |
+
"kind": "categorical",
|
| 311 |
+
"counts": {
|
| 312 |
+
"I": 313.0,
|
| 313 |
+
"C": 57.0,
|
| 314 |
+
"E": 33.0,
|
| 315 |
+
"A": 30.0,
|
| 316 |
+
"B": 29.0,
|
| 317 |
+
"G": 10.0,
|
| 318 |
+
"J": 5.0,
|
| 319 |
+
"F": 4.0
|
| 320 |
+
}
|
| 321 |
+
},
|
| 322 |
+
"n": 481,
|
| 323 |
+
"sources": [
|
| 324 |
+
{
|
| 325 |
+
"source_id": "ts-v201-meta",
|
| 326 |
+
"value_source": "computed",
|
| 327 |
+
"locator": "meta.csv: institute"
|
| 328 |
+
}
|
| 329 |
+
]
|
| 330 |
+
},
|
| 331 |
+
{
|
| 332 |
+
"variable": "pathology",
|
| 333 |
+
"label": "Pathology",
|
| 334 |
+
"summary": {
|
| 335 |
+
"kind": "categorical",
|
| 336 |
+
"counts": {
|
| 337 |
+
"no_pathology": 139.0,
|
| 338 |
+
"tumor": 102.0,
|
| 339 |
+
"unclear": 84.0,
|
| 340 |
+
"vascular": 44.0,
|
| 341 |
+
"other": 37.0,
|
| 342 |
+
"inflammation": 37.0,
|
| 343 |
+
"trauma": 33.0,
|
| 344 |
+
"bleeding": 5.0
|
| 345 |
+
}
|
| 346 |
+
},
|
| 347 |
+
"n": 481,
|
| 348 |
+
"sources": [
|
| 349 |
+
{
|
| 350 |
+
"source_id": "ts-v201-meta",
|
| 351 |
+
"value_source": "computed",
|
| 352 |
+
"locator": "meta.csv: pathology"
|
| 353 |
+
}
|
| 354 |
+
]
|
| 355 |
+
}
|
| 356 |
+
],
|
| 357 |
+
"not_disclosed": [
|
| 358 |
+
"contrast phase",
|
| 359 |
+
"race and ethnicity (not recorded in the source)",
|
| 360 |
+
"joint distributions"
|
| 361 |
+
],
|
| 362 |
+
"notes": [
|
| 363 |
+
"Computed from the dataset's row-level metadata (training split only), not read off the paper's Figure 3, which describes the v1 dataset (1,204 scans).",
|
| 364 |
+
"Institute codes are the dataset's own. The paper's Figure 3 site letters are a different coding."
|
| 365 |
+
]
|
| 366 |
+
}
|
vendor_datasheet/totalsegmentator-v2-train-all.json
ADDED
|
@@ -0,0 +1,371 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"schema_version": "0.1.0",
|
| 3 |
+
"datasheet_id": "totalsegmentator-v2-train-all",
|
| 4 |
+
"title": "TotalSegmentator, training data (all scans)",
|
| 5 |
+
"datasheet_version": "TotalSegmentator dataset v2.0.1",
|
| 6 |
+
"model": {
|
| 7 |
+
"name": "TotalSegmentator",
|
| 8 |
+
"version": "2.x, task total",
|
| 9 |
+
"manufacturer": "University Hospital Basel",
|
| 10 |
+
"intended_use": "Segmentation of 104+ anatomic structures in CT (research use)",
|
| 11 |
+
"regulatory_ids": []
|
| 12 |
+
},
|
| 13 |
+
"describes": "training",
|
| 14 |
+
"scope": {
|
| 15 |
+
"description": "Routine clinical CT of every body region (head to feet, including whole-body, polytrauma and extremity studies) from University Hospital Basel, years 2012, 2016 and 2020",
|
| 16 |
+
"modality": "CT",
|
| 17 |
+
"body_regions": [
|
| 18 |
+
"head",
|
| 19 |
+
"neck",
|
| 20 |
+
"thorax",
|
| 21 |
+
"abdomen",
|
| 22 |
+
"pelvis",
|
| 23 |
+
"extremities"
|
| 24 |
+
],
|
| 25 |
+
"countries": [
|
| 26 |
+
"Switzerland"
|
| 27 |
+
],
|
| 28 |
+
"inclusion": [],
|
| 29 |
+
"exclusion": []
|
| 30 |
+
},
|
| 31 |
+
"n_scans": 1082,
|
| 32 |
+
"sources": [
|
| 33 |
+
{
|
| 34 |
+
"source_id": "ts-v201-meta",
|
| 35 |
+
"kind": "dataset_release",
|
| 36 |
+
"citation": "Wasserthal J. et al. TotalSegmentator dataset v2.0.1, meta.csv (Zenodo)",
|
| 37 |
+
"url": "https://doi.org/10.5281/zenodo.10047292"
|
| 38 |
+
},
|
| 39 |
+
{
|
| 40 |
+
"source_id": "ts-v300-meta",
|
| 41 |
+
"kind": "dataset_release",
|
| 42 |
+
"citation": "TotalSegmentator dataset v3.0.0, meta.csv (Zenodo): acquisition slice thickness",
|
| 43 |
+
"url": "https://doi.org/10.5281/zenodo.22688904"
|
| 44 |
+
},
|
| 45 |
+
{
|
| 46 |
+
"source_id": "ts-paper",
|
| 47 |
+
"kind": "publication",
|
| 48 |
+
"citation": "Wasserthal J. et al. TotalSegmentator: Robust Segmentation of 104 Anatomic Structures in CT Images. Radiology: Artificial Intelligence 2023;5(5)",
|
| 49 |
+
"doi": "10.1148/ryai.230024"
|
| 50 |
+
}
|
| 51 |
+
],
|
| 52 |
+
"variables": [
|
| 53 |
+
{
|
| 54 |
+
"variable": "age_years",
|
| 55 |
+
"label": "Age",
|
| 56 |
+
"summary": {
|
| 57 |
+
"kind": "continuous",
|
| 58 |
+
"unit": "years",
|
| 59 |
+
"mean": 63.414,
|
| 60 |
+
"sd": 14.847,
|
| 61 |
+
"median": 65.0,
|
| 62 |
+
"q1": 55.0,
|
| 63 |
+
"q3": 74.0,
|
| 64 |
+
"min": 15.0,
|
| 65 |
+
"max": 94.0,
|
| 66 |
+
"histogram": {
|
| 67 |
+
"edges": [
|
| 68 |
+
10.0,
|
| 69 |
+
20.0,
|
| 70 |
+
30.0,
|
| 71 |
+
40.0,
|
| 72 |
+
50.0,
|
| 73 |
+
60.0,
|
| 74 |
+
70.0,
|
| 75 |
+
80.0,
|
| 76 |
+
90.0,
|
| 77 |
+
100.0,
|
| 78 |
+
110.0
|
| 79 |
+
],
|
| 80 |
+
"counts": [
|
| 81 |
+
5.0,
|
| 82 |
+
29.0,
|
| 83 |
+
36.0,
|
| 84 |
+
110.0,
|
| 85 |
+
210.0,
|
| 86 |
+
286.0,
|
| 87 |
+
266.0,
|
| 88 |
+
124.0,
|
| 89 |
+
16.0,
|
| 90 |
+
0.0
|
| 91 |
+
]
|
| 92 |
+
}
|
| 93 |
+
},
|
| 94 |
+
"n": 1082,
|
| 95 |
+
"sources": [
|
| 96 |
+
{
|
| 97 |
+
"source_id": "ts-v201-meta",
|
| 98 |
+
"value_source": "computed",
|
| 99 |
+
"locator": "meta.csv: age"
|
| 100 |
+
}
|
| 101 |
+
]
|
| 102 |
+
},
|
| 103 |
+
{
|
| 104 |
+
"variable": "sex",
|
| 105 |
+
"label": "Sex",
|
| 106 |
+
"summary": {
|
| 107 |
+
"kind": "categorical",
|
| 108 |
+
"counts": {
|
| 109 |
+
"M": 636.0,
|
| 110 |
+
"F": 445.0
|
| 111 |
+
}
|
| 112 |
+
},
|
| 113 |
+
"n": 1081,
|
| 114 |
+
"n_missing": 1,
|
| 115 |
+
"sources": [
|
| 116 |
+
{
|
| 117 |
+
"source_id": "ts-v201-meta",
|
| 118 |
+
"value_source": "computed",
|
| 119 |
+
"locator": "meta.csv: gender"
|
| 120 |
+
}
|
| 121 |
+
]
|
| 122 |
+
},
|
| 123 |
+
{
|
| 124 |
+
"variable": "scanner_vendor",
|
| 125 |
+
"label": "Scanner manufacturer",
|
| 126 |
+
"summary": {
|
| 127 |
+
"kind": "categorical",
|
| 128 |
+
"counts": {
|
| 129 |
+
"siemens": 971.0,
|
| 130 |
+
"philips": 68.0,
|
| 131 |
+
"ge": 43.0
|
| 132 |
+
}
|
| 133 |
+
},
|
| 134 |
+
"n": 1082,
|
| 135 |
+
"sources": [
|
| 136 |
+
{
|
| 137 |
+
"source_id": "ts-v201-meta",
|
| 138 |
+
"value_source": "computed",
|
| 139 |
+
"locator": "meta.csv: manufacturer"
|
| 140 |
+
}
|
| 141 |
+
]
|
| 142 |
+
},
|
| 143 |
+
{
|
| 144 |
+
"variable": "scanner_model",
|
| 145 |
+
"label": "Scanner model",
|
| 146 |
+
"summary": {
|
| 147 |
+
"kind": "categorical",
|
| 148 |
+
"counts": {
|
| 149 |
+
"somatom definition flash": 310.0,
|
| 150 |
+
"somatom definition as+": 181.0,
|
| 151 |
+
"somatom definition edge": 153.0,
|
| 152 |
+
"somatom force": 117.0,
|
| 153 |
+
"emotion 16": 83.0,
|
| 154 |
+
"sensation 64": 56.0,
|
| 155 |
+
"ingenuity core 128": 51.0,
|
| 156 |
+
"lightspeed vct": 39.0,
|
| 157 |
+
"sensation 16": 35.0,
|
| 158 |
+
"somatom go.top": 27.0,
|
| 159 |
+
"iqon - spectral ct": 12.0,
|
| 160 |
+
"somatom definition as": 7.0,
|
| 161 |
+
"brilliance 64": 5.0,
|
| 162 |
+
"discovery ste": 4.0,
|
| 163 |
+
"sensation 10": 1.0,
|
| 164 |
+
"biograph128": 1.0
|
| 165 |
+
}
|
| 166 |
+
},
|
| 167 |
+
"n": 1082,
|
| 168 |
+
"sources": [
|
| 169 |
+
{
|
| 170 |
+
"source_id": "ts-v201-meta",
|
| 171 |
+
"value_source": "computed",
|
| 172 |
+
"locator": "meta.csv: scanner_model"
|
| 173 |
+
}
|
| 174 |
+
]
|
| 175 |
+
},
|
| 176 |
+
{
|
| 177 |
+
"variable": "kvp",
|
| 178 |
+
"label": "Tube voltage",
|
| 179 |
+
"summary": {
|
| 180 |
+
"kind": "continuous",
|
| 181 |
+
"unit": "kV",
|
| 182 |
+
"mean": 111.627,
|
| 183 |
+
"sd": 14.173,
|
| 184 |
+
"median": 120.0,
|
| 185 |
+
"q1": 100.0,
|
| 186 |
+
"q3": 120.0,
|
| 187 |
+
"min": 70.0,
|
| 188 |
+
"max": 150.0,
|
| 189 |
+
"histogram": {
|
| 190 |
+
"edges": [
|
| 191 |
+
65.0,
|
| 192 |
+
75.0,
|
| 193 |
+
85.0,
|
| 194 |
+
95.0,
|
| 195 |
+
105.0,
|
| 196 |
+
115.0,
|
| 197 |
+
125.0,
|
| 198 |
+
135.0,
|
| 199 |
+
145.0,
|
| 200 |
+
155.0,
|
| 201 |
+
165.0
|
| 202 |
+
],
|
| 203 |
+
"counts": [
|
| 204 |
+
2.0,
|
| 205 |
+
58.0,
|
| 206 |
+
20.0,
|
| 207 |
+
356.0,
|
| 208 |
+
47.0,
|
| 209 |
+
503.0,
|
| 210 |
+
51.0,
|
| 211 |
+
31.0,
|
| 212 |
+
14.0,
|
| 213 |
+
0.0
|
| 214 |
+
]
|
| 215 |
+
}
|
| 216 |
+
},
|
| 217 |
+
"n": 1082,
|
| 218 |
+
"sources": [
|
| 219 |
+
{
|
| 220 |
+
"source_id": "ts-v201-meta",
|
| 221 |
+
"value_source": "computed",
|
| 222 |
+
"locator": "meta.csv: kvp"
|
| 223 |
+
}
|
| 224 |
+
]
|
| 225 |
+
},
|
| 226 |
+
{
|
| 227 |
+
"variable": "slice_thickness_mm",
|
| 228 |
+
"label": "Slice thickness (acquisition)",
|
| 229 |
+
"summary": {
|
| 230 |
+
"kind": "continuous",
|
| 231 |
+
"unit": "mm",
|
| 232 |
+
"mean": 1.952,
|
| 233 |
+
"sd": 1.188,
|
| 234 |
+
"median": 1.5,
|
| 235 |
+
"q1": 1.0,
|
| 236 |
+
"q3": 2.875,
|
| 237 |
+
"min": 0.6,
|
| 238 |
+
"max": 5.0,
|
| 239 |
+
"histogram": {
|
| 240 |
+
"edges": [
|
| 241 |
+
0.0,
|
| 242 |
+
0.7,
|
| 243 |
+
0.95,
|
| 244 |
+
1.1,
|
| 245 |
+
1.4,
|
| 246 |
+
1.75,
|
| 247 |
+
2.25,
|
| 248 |
+
2.75,
|
| 249 |
+
3.4,
|
| 250 |
+
4.5,
|
| 251 |
+
6.0
|
| 252 |
+
],
|
| 253 |
+
"counts": [
|
| 254 |
+
43.0,
|
| 255 |
+
82.0,
|
| 256 |
+
184.0,
|
| 257 |
+
32.0,
|
| 258 |
+
293.0,
|
| 259 |
+
176.0,
|
| 260 |
+
1.0,
|
| 261 |
+
174.0,
|
| 262 |
+
1.0,
|
| 263 |
+
96.0
|
| 264 |
+
]
|
| 265 |
+
}
|
| 266 |
+
},
|
| 267 |
+
"n": 1082,
|
| 268 |
+
"sources": [
|
| 269 |
+
{
|
| 270 |
+
"source_id": "ts-v300-meta",
|
| 271 |
+
"value_source": "computed",
|
| 272 |
+
"locator": "meta.csv: slice_thickness",
|
| 273 |
+
"note": "v3.0.0 metadata joined on image_id"
|
| 274 |
+
}
|
| 275 |
+
]
|
| 276 |
+
},
|
| 277 |
+
{
|
| 278 |
+
"variable": "study_type",
|
| 279 |
+
"label": "Study type",
|
| 280 |
+
"summary": {
|
| 281 |
+
"kind": "categorical",
|
| 282 |
+
"counts": {
|
| 283 |
+
"ct thorax-abdomen-pelvis": 120.0,
|
| 284 |
+
"ct neck-thorax-abdomen-pelvis": 96.0,
|
| 285 |
+
"ct abdomen-pelvis": 74.0,
|
| 286 |
+
"ct thorax-neck": 72.0,
|
| 287 |
+
"ct neck": 59.0,
|
| 288 |
+
"ct angiography head": 58.0,
|
| 289 |
+
"ct abdomen": 52.0,
|
| 290 |
+
"ct thorax-abdomen": 50.0,
|
| 291 |
+
"ct pelvis": 48.0,
|
| 292 |
+
"ct angiography neck-thx-abd-pelvis-leg": 47.0,
|
| 293 |
+
"ct polytrauma": 43.0,
|
| 294 |
+
"ct heart-thorakale aorta": 42.0,
|
| 295 |
+
"ct intervention": 38.0,
|
| 296 |
+
"ct spine": 38.0,
|
| 297 |
+
"ct thorax-chest": 36.0,
|
| 298 |
+
"other": 209.0
|
| 299 |
+
}
|
| 300 |
+
},
|
| 301 |
+
"n": 1082,
|
| 302 |
+
"sources": [
|
| 303 |
+
{
|
| 304 |
+
"source_id": "ts-v201-meta",
|
| 305 |
+
"value_source": "computed",
|
| 306 |
+
"locator": "meta.csv: study_type"
|
| 307 |
+
}
|
| 308 |
+
]
|
| 309 |
+
},
|
| 310 |
+
{
|
| 311 |
+
"variable": "site",
|
| 312 |
+
"label": "Institute (coded)",
|
| 313 |
+
"summary": {
|
| 314 |
+
"kind": "categorical",
|
| 315 |
+
"counts": {
|
| 316 |
+
"I": 799.0,
|
| 317 |
+
"C": 83.0,
|
| 318 |
+
"A": 70.0,
|
| 319 |
+
"E": 51.0,
|
| 320 |
+
"B": 47.0,
|
| 321 |
+
"G": 21.0,
|
| 322 |
+
"J": 5.0,
|
| 323 |
+
"F": 5.0,
|
| 324 |
+
"H": 1.0
|
| 325 |
+
}
|
| 326 |
+
},
|
| 327 |
+
"n": 1082,
|
| 328 |
+
"sources": [
|
| 329 |
+
{
|
| 330 |
+
"source_id": "ts-v201-meta",
|
| 331 |
+
"value_source": "computed",
|
| 332 |
+
"locator": "meta.csv: institute"
|
| 333 |
+
}
|
| 334 |
+
]
|
| 335 |
+
},
|
| 336 |
+
{
|
| 337 |
+
"variable": "pathology",
|
| 338 |
+
"label": "Pathology",
|
| 339 |
+
"summary": {
|
| 340 |
+
"kind": "categorical",
|
| 341 |
+
"counts": {
|
| 342 |
+
"no_pathology": 372.0,
|
| 343 |
+
"tumor": 205.0,
|
| 344 |
+
"unclear": 140.0,
|
| 345 |
+
"vascular": 136.0,
|
| 346 |
+
"inflammation": 81.0,
|
| 347 |
+
"trauma": 80.0,
|
| 348 |
+
"other": 57.0,
|
| 349 |
+
"bleeding": 11.0
|
| 350 |
+
}
|
| 351 |
+
},
|
| 352 |
+
"n": 1082,
|
| 353 |
+
"sources": [
|
| 354 |
+
{
|
| 355 |
+
"source_id": "ts-v201-meta",
|
| 356 |
+
"value_source": "computed",
|
| 357 |
+
"locator": "meta.csv: pathology"
|
| 358 |
+
}
|
| 359 |
+
]
|
| 360 |
+
}
|
| 361 |
+
],
|
| 362 |
+
"not_disclosed": [
|
| 363 |
+
"contrast phase",
|
| 364 |
+
"race and ethnicity (not recorded in the source)",
|
| 365 |
+
"joint distributions"
|
| 366 |
+
],
|
| 367 |
+
"notes": [
|
| 368 |
+
"Computed from the dataset's row-level metadata (training split only), not read off the paper's Figure 3, which describes the v1 dataset (1,204 scans).",
|
| 369 |
+
"Institute codes are the dataset's own. The paper's Figure 3 site letters are a different coding."
|
| 370 |
+
]
|
| 371 |
+
}
|