Hugging Face's logo

Datasets:
ppcnet-dataset
/
PPCNet

Tasks:
Image-to-3D
Other
Modalities:
3D
Image
Text
Formats:
text
Size:
1K - 10K
Tags:
medical-imaging
spine
point-cloud
lumbar
DRR
biplanar-radiographs
Libraries:
Datasets
License:
Dataset card Data Studio Files
xet
 Community
1
ppcnet-dataset commited on
Commit
1070bc2
·
verified ·
1 Parent(s): 861cf1d

Update README.md

Browse files
Files changed (1) hide show
  1. README.md +1 -9
README.md CHANGED
@@ -19,7 +19,7 @@ size_categories:
19
  <div align="center">
20
 
21
  # PPCNet Dataset
22
- ### Biplanar DRRs + Projection Matrices + Ground-Truth Point Clouds
23
 
24
  [![License](https://img.shields.io/badge/License-MIT-yellow.svg)](LICENSE)
25
  [![Patients](https://img.shields.io/badge/Patients-1%2C037-blue)]()
@@ -59,8 +59,6 @@ Each of the 1,037 patient folders contains:
59
 
60
  Fixed split in `dataset_split.json` (seed=42):
61
 
62
- <div align="center">
63
-
64
  | Split | Patients | Usage |
65
  |:-----:|:--------:|:-----:|
66
  | Train | 829 | Model training |
@@ -68,16 +66,12 @@ Fixed split in `dataset_split.json` (seed=42):
68
  | Test | 105 | Final evaluation |
69
  | **Total** | **1,037** | |
70
 
71
- </div>
72
-
73
  ---
74
 
75
  ## 🔬 DRR Generation
76
 
77
  DRRs are generated using **Plastimatch's ray-casting algorithm** with the following parameters:
78
 
79
- <div align="center">
80
-
81
  | Parameter | Value |
82
  |:---------:|:-----:|
83
  | Source-to-Axis Distance (SAD) | 1,000 mm |
@@ -88,8 +82,6 @@ DRRs are generated using **Plastimatch's ray-casting algorithm** with the follow
88
  | Bone Enhancement | 2.5× for HU > 300 |
89
  | Post-processing | CLAHE normalisation |
90
 
91
- </div>
92
-
93
  Each DRR comes with a calibrated **3×4 perspective projection matrix** encoding the exact source–detector geometry.
94
 
95
  ---
 
19
  <div align="center">
20
 
21
  # PPCNet Dataset
22
+ ## Biplanar DRRs + Projection Matrices + Ground-Truth Point Clouds
23
 
24
  [![License](https://img.shields.io/badge/License-MIT-yellow.svg)](LICENSE)
25
  [![Patients](https://img.shields.io/badge/Patients-1%2C037-blue)]()
 
59
 
60
  Fixed split in `dataset_split.json` (seed=42):
61
 
 
 
62
  | Split | Patients | Usage |
63
  |:-----:|:--------:|:-----:|
64
  | Train | 829 | Model training |
 
66
  | Test | 105 | Final evaluation |
67
  | **Total** | **1,037** | |
68
 
 
 
69
  ---
70
 
71
  ## 🔬 DRR Generation
72
 
73
  DRRs are generated using **Plastimatch's ray-casting algorithm** with the following parameters:
74
 
 
 
75
  | Parameter | Value |
76
  |:---------:|:-----:|
77
  | Source-to-Axis Distance (SAD) | 1,000 mm |
 
82
  | Bone Enhancement | 2.5× for HU > 300 |
83
  | Post-processing | CLAHE normalisation |
84
 
 
 
85
  Each DRR comes with a calibrated **3×4 perspective projection matrix** encoding the exact source–detector geometry.
86
 
87
  ---