تشغيل أي مهارة في Manus بنقرة واحدة

habit-preprocess

النجوم٦

التفرعات٢

آخر تحديث١٥ مايو ٢٠٢٦ في ٠٨:٢٠

Preprocess medical images (CT/MRI) for HABIT — DICOM-to-NIfTI conversion, resampling, multi-modal registration, N4 bias correction, z-score / histogram standardization, CLAHE. Use when the user mentions "图像预处理", "重采样", "配准", "DICOM 转 NIfTI", "N4", "标准化", "preprocess", "register", "bias field". Runs the `habit preprocess` CLI.

التثبيت

التثبيت باستخدام Codex أو Claude انسخ هذا Prompt والصقه في Codex أو Claude أو مساعد آخر ليراجع صفحة Skill ويثبّتها لك.

تشغيل في Manus

المصدر

lichao312214129

lichao312214129/HABIT

فتح مستودع GitHub عرض مستودعات المنشئ

تنزيل

تشغيل في Manus

المهن ذات الصلةSOC

استنادا إلى تصنيف SOC المهني

مطوّرو البرمجياتمهن الحاسوب والرياضيات·SOC 15-1252

مستكشف الملفات

2 ملفات

SKILL.md

readonly

name	habit-preprocess
description	Preprocess medical images (CT/MRI) for HABIT — DICOM-to-NIfTI conversion, resampling, multi-modal registration, N4 bias correction, z-score / histogram standardization, CLAHE. Use when the user mentions "图像预处理", "重采样", "配准", "DICOM 转 NIfTI", "N4", "标准化", "preprocess", "register", "bias field". Runs the `habit preprocess` CLI.

HABIT Image Preprocessing

Drives the habit preprocess CLI command. The pipeline is fully YAML-driven — each step is optional and applied in the order listed in the config.

CLI

habit preprocess --config <your_config.yaml>

Required Information

Before generating a config, confirm:

Field	Notes
`data_dir`	path to subject folders (raw or DICOM)
`out_dir`	output root
Modality folder names	e.g. `T1, T2, DWI, ADC`
MRI or CT?	drives whether to use `n4_correction`
Multi-modal?	drives whether to use `registration`
Registration `fixed_image`	usually T2
(DICOM only) `dcm2niix_path`	full path to `dcm2niix.exe`

Available methods (in execution order)

Method	Purpose	Use for
`dcm2nii`	DICOM → NIfTI	raw DICOM input
`n4_correction`	bias field correction	MRI only
`resample`	unify voxel spacing	always
`registration`	align modalities	multi-modal
`zscore_normalization`	standardize intensity	MRI before ML (NOT for DCE phases or CT HU)
`histogram_standardization`	Nyúl method	multi-center MRI
`adaptive_histogram_equalization`	CLAHE	optional contrast boost

Decision tree

DICOM input? → use dcm2nii first. See config/preprocessing/config_image_preprocessing_dcm2nii.yaml.
MRI? → enable n4_correction then resample then zscore_normalization.
CT? → just resample. NEVER N4 (CT has no bias field). See config/preprocessing/config_image_preprocessing.yaml.
Multi-modal? → add registration with one image as fixed_image (usually T2).
DCE-MRI? → use config/preprocessing/config_image_preprocessing.yaml. Do NOT z-score phases independently.
Multi-center cohort? → add histogram_standardization after N4.

Reference templates

Config index: skills/CONFIG_SOURCES.md.

File	Use for
`config/preprocessing/config_image_preprocessing.yaml`	starting scaffold (any)
`config/preprocessing/config_image_preprocessing_dcm2nii.yaml`	DICOM → NIfTI
`config/preprocessing/config_image_preprocessing.yaml`	T1+T2(+DWI/ADC) standard MRI
`config/preprocessing/config_image_preprocessing.yaml`	DCE-MRI dynamic
`config/preprocessing/config_image_preprocessing.yaml`	CT (resample only)

Full annotated reference: config/preprocessing/config_image_preprocessing.yaml.

Save intermediate outputs (debugging)

save_options:
  save_intermediate: true
  intermediate_steps: []   # empty = save every step

Each step then writes to <out_dir>/<stage>_NN/... (e.g. n4_correction_01/). The final aggregated output is always at <out_dir>/processed_images/.

Validate output (MANDATORY after run)

python skills/habit-preprocess/scripts/validate_preprocess_output.py <out_dir> --modalities T1 T2 DWI ADC

This script checks every subject has all required modalities, no constant or all-zero volumes, masks present and non-empty.

Output structure

out_dir/processed_images/
├── images/<subject>/<modality>/<modality>.nii.gz
└── masks/<subject>/<modality>/<modality>.nii.gz

This is the expected input for habit get-habitat and habit extract.

Common pitfalls

Windows paths must use forward slashes in YAML (./software/dcm2niix.exe).
fixed_image must NOT appear in moving_images — it stays fixed.
only_inmask: true requires a mask_key: value.
Mask files are NOT modified by preprocessing (only resampled if registration runs).
N4 + CT is wrong — N4 is MRI-specific.
Apply resample BEFORE registration so all images share the same grid.
registration backends: ants (default, ANTsPy), simpleitk (SimpleITK only), or elastix (elastix / transformix executables; optional elastix_path in YAML). See docs/source/user_guide/image_preprocessing_zh.rst.

Logs

Preprocessing log is at <out_dir>/preprocess.log. Always check it on errors.

Next step

After preprocessing succeeds, the typical next skill is habit-habitat-analysis.

المزيد من هذا المستودع

نفس المستودع

habit-quickstart

lichao312214129/HABIT

Entry-point router for the HABIT (Habitat Analysis Biomedical Imaging Toolkit) package. Use this FIRST whenever a user mentions HABIT, habitat analysis, tumor sub-region clustering, intra-tumor heterogeneity, "生境分析", "亚区聚类", "habitat segmentation", or wants a multi-step radiomics workflow. Verifies environment, validates data layout, then hands off to the right specialist skill.

2026-07-216

habit-troubleshoot

lichao312214129/HABIT

Diagnose and fix HABIT runtime errors — covers preprocess, habitat clustering, feature extraction, and ML/comparison failures. Use when the user pastes a Python traceback, when a `habit ...` CLI command fails, when an output file is missing or corrupt, or when results look biologically wrong (degenerate clusters, all-NaN features, AUC=0.5). Triggers on phrases like "报错", "出错", "不工作", "为什么", "error", "failed", "traceback", "crash", "AUC 太低", "没有生成结果".

2026-07-216

habit-feature-extraction

lichao312214129/HABIT

Extract quantitative features from HABIT habitat maps — traditional radiomics, whole-habitat radiomics, per-habitat radiomics, MSI (Most Significant Intensity), and ITH (Intra-Tumor Heterogeneity) scores. Use when the user has habitat .nrrd maps and wants per-subject feature CSVs ready for ML. Triggers on "提取生境特征", "MSI 特征", "ITH 异质性", "habitat features", "extract radiomics from habitats". Runs `habit extract`.

2026-07-066

habit-habitat-analysis

lichao312214129/HABIT

Generate tumor habitat (sub-region) maps from medical images using voxel-level clustering. Use when the user wants to identify intra-tumor heterogeneity zones, perform supervoxel clustering, generate habitat .nrrd maps, run kinetic DCE habitat, or use voxel-radiomics texture clustering. Triggers on "生境分析", "亚区聚类", "habitat", "supervoxel", "kinetic", "DCE 分期生境", "intra-tumor heterogeneity". Runs `habit get-habitat`.

2026-07-066

habit-machine-learning

lichao312214129/HABIT

Train, predict, or k-fold cross-validate ML classifiers on habitat / radiomics feature CSVs using HABIT. Supports LogisticRegression, RandomForest, XGBoost, SVM, MLP, AutoGluon, and 12+ feature selection methods (LASSO, mRMR, RFECV, ICC, correlation, ANOVA, ...). Use when the user wants to build a prediction model from feature CSVs. Triggers on "训练模型", "建模", "K折交叉验证", "k-fold", "LASSO", "feature selection", "predict mode". Runs `habit model` or `habit cv`.

2026-07-066

habit-model-comparison

lichao312214129/HABIT

Compare multiple trained classification models with publication-quality plots — ROC, DCA, calibration, precision-recall, DeLong's AUC test. Use when the user has prediction CSVs from 2+ models and wants side-by-side comparison. Triggers on "模型比较", "ROC 对比", "DeLong 检验", "决策曲线", "校准曲线", "model comparison", "compare AUC", "DCA". Runs `habit compare`.

2026-07-066

name	habit-preprocess
description	Preprocess medical images (CT/MRI) for HABIT — DICOM-to-NIfTI conversion, resampling, multi-modal registration, N4 bias correction, z-score / histogram standardization, CLAHE. Use when the user mentions "图像预处理", "重采样", "配准", "DICOM 转 NIfTI", "N4", "标准化", "preprocess", "register", "bias field". Runs the `habit preprocess` CLI.

HABIT Image Preprocessing

Drives the habit preprocess CLI command. The pipeline is fully YAML-driven — each step is optional and applied in the order listed in the config.

CLI

habit preprocess --config <your_config.yaml>

Required Information

Before generating a config, confirm:

Field	Notes
`data_dir`	path to subject folders (raw or DICOM)
`out_dir`	output root
Modality folder names	e.g. `T1, T2, DWI, ADC`
MRI or CT?	drives whether to use `n4_correction`
Multi-modal?	drives whether to use `registration`
Registration `fixed_image`	usually T2
(DICOM only) `dcm2niix_path`	full path to `dcm2niix.exe`

Available methods (in execution order)

Method	Purpose	Use for
`dcm2nii`	DICOM → NIfTI	raw DICOM input
`n4_correction`	bias field correction	MRI only
`resample`	unify voxel spacing	always
`registration`	align modalities	multi-modal
`zscore_normalization`	standardize intensity	MRI before ML (NOT for DCE phases or CT HU)
`histogram_standardization`	Nyúl method	multi-center MRI
`adaptive_histogram_equalization`	CLAHE	optional contrast boost

Decision tree

DICOM input? → use dcm2nii first. See config/preprocessing/config_image_preprocessing_dcm2nii.yaml.
MRI? → enable n4_correction then resample then zscore_normalization.
CT? → just resample. NEVER N4 (CT has no bias field). See config/preprocessing/config_image_preprocessing.yaml.
Multi-modal? → add registration with one image as fixed_image (usually T2).
DCE-MRI? → use config/preprocessing/config_image_preprocessing.yaml. Do NOT z-score phases independently.
Multi-center cohort? → add histogram_standardization after N4.

Reference templates

Config index: skills/CONFIG_SOURCES.md.

File	Use for
`config/preprocessing/config_image_preprocessing.yaml`	starting scaffold (any)
`config/preprocessing/config_image_preprocessing_dcm2nii.yaml`	DICOM → NIfTI
`config/preprocessing/config_image_preprocessing.yaml`	T1+T2(+DWI/ADC) standard MRI
`config/preprocessing/config_image_preprocessing.yaml`	DCE-MRI dynamic
`config/preprocessing/config_image_preprocessing.yaml`	CT (resample only)

Full annotated reference: config/preprocessing/config_image_preprocessing.yaml.

Save intermediate outputs (debugging)

save_options:
  save_intermediate: true
  intermediate_steps: []   # empty = save every step

Each step then writes to <out_dir>/<stage>_NN/... (e.g. n4_correction_01/). The final aggregated output is always at <out_dir>/processed_images/.

Validate output (MANDATORY after run)

python skills/habit-preprocess/scripts/validate_preprocess_output.py <out_dir> --modalities T1 T2 DWI ADC

This script checks every subject has all required modalities, no constant or all-zero volumes, masks present and non-empty.

Output structure

out_dir/processed_images/
├── images/<subject>/<modality>/<modality>.nii.gz
└── masks/<subject>/<modality>/<modality>.nii.gz

This is the expected input for habit get-habitat and habit extract.

Common pitfalls

Windows paths must use forward slashes in YAML (./software/dcm2niix.exe).
fixed_image must NOT appear in moving_images — it stays fixed.
only_inmask: true requires a mask_key: value.
Mask files are NOT modified by preprocessing (only resampled if registration runs).
N4 + CT is wrong — N4 is MRI-specific.
Apply resample BEFORE registration so all images share the same grid.
registration backends: ants (default, ANTsPy), simpleitk (SimpleITK only), or elastix (elastix / transformix executables; optional elastix_path in YAML). See docs/source/user_guide/image_preprocessing_zh.rst.

Logs

Preprocessing log is at <out_dir>/preprocess.log. Always check it on errors.

Next step

After preprocessing succeeds, the typical next skill is habit-habitat-analysis.