You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
`BrainLes preprocessing` is a comprehensive tool for preprocessing tasks in biomedical imaging, with a focus on (but not limited to) multi-modal brain MRI. It can be used to build modular preprocessing pipelines:
11
+
`BrainLes preprocessing` is a comprehensive, modular toolkit for preprocessing multi-modal brain MRI and other biomedical imaging data. It provides flexible preprocessing pipelines that can be customized to your specific needs.
12
12
13
-
This includes **normalization**, **co-registration**, **atlas registration**, **skullstripping / brain extraction**, **N4 Bias correction** and **defacing**.
14
-
We provide means to transform images and segmentations in both directions between native and atlas space.
13
+
## Features
15
14
16
-
BrainLes is written modular and `backend-agnostic` meaning it allows to skip or swap registration, brain extraction, N4 bias correction and defacing tools.
15
+
### Core Preprocessing Steps
16
+
-**Normalization**: Intensity normalization using various methods (e.g., percentile-based)
17
+
-**Co-registration**: Align multiple modalities to a reference modality
18
+
-**Atlas Registration**: Register images to standard atlas spaces (MNI152, SRI24, etc.)
19
+
-**Brain Extraction**: Skull stripping using state-of-the-art methods (HD-BET, SynthStrip)
-**Defacing**: Anonymize images by removing facial features
22
+
23
+
### Preprocessing Modes
24
+
-**Atlas-Centric**: Process images in atlas space with optional atlas-based intensity correction
25
+
-**Native Space**: Process images in patient/native space without atlas registration
26
+
27
+
### Key Benefits
28
+
-**Modular & Backend-Agnostic**: Easily swap or skip preprocessing steps and choose from multiple backends
29
+
-**Flexible Output Options**: Generate any combination of outputs (brain extracted, with skull, defaced, raw, normalized)
30
+
-**Bidirectional Transforms**: Transform images and segmentations between native and atlas space
31
+
-**Extensible**: Add custom normalizers, registrators, brain extractors, or defacing methods
17
32
18
33
<!-- TODO include image here -->
19
34
@@ -36,64 +51,205 @@ We recommend using Python `3.10 / 3.11 / 3.12`.
36
51
37
52
38
53
## Usage
39
-
A minimal example to register (to the standard atlas using ANTs) and skull strip (using HDBet) a t1c image (center modality) with 1 moving modality (flair) could look like this:
54
+
55
+
### Atlas-Centric Preprocessing
56
+
Use `AtlasCentricPreprocessor` to register images to an atlas, perform atlas correction, and skull strip. This is useful when you want all images in a common atlas space.
57
+
58
+
**Key features:**
59
+
- Co-registration of moving modalities to center modality
60
+
- Registration to atlas space
61
+
- Optional atlas correction (intensity adjustment based on atlas)
62
+
- N4 bias correction
63
+
- Brain extraction
64
+
- Defacing for anonymization
65
+
66
+
**Example with all output options:**
67
+
68
+
This example demonstrates all possible output paths. You can specify any combination of:
Use `NativeSpacePreprocessor` to perform coregistration, N4 bias correction, brain extraction, and defacing while keeping images in native space (no atlas registration).
160
+
161
+
**Key features:**
162
+
- Co-registration of moving modalities to center modality
163
+
- N4 bias correction
164
+
- Brain extraction
165
+
- Defacing for anonymization
166
+
-**No atlas registration** - stays in native/patient space
167
+
168
+
**Example with all output options:**
169
+
170
+
This example demonstrates all possible output paths. You can specify any combination of:
0 commit comments