Vanilla nnU-Net compatible and README updates

GLOW-FDG/fold_0/checkpoint_final.pth

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:947622aeb87829d9adc5c993b5151fd419f6dd995849987eddb80c2eacb81f3a
-size 819730354
+oid sha256:8f688e4a5ef8614b8bfb185a29ecf4072185ad16dd98fc2cd28486e8c637213a
+size 409647917

GLOW-FDG/fold_1/checkpoint_final.pth

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:aab8a3033565dd31babfe708bef565ea494a837f7bcac09cb3b09c5c5127853b
-size 819730418
+oid sha256:1c3d51495bc8d192064ec13d7ea9270dd2e7847141f2cbb921c126ec3d21e6ac
+size 409647917

GLOW-FDG/fold_2/checkpoint_final.pth

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:81e685eb0212811088d87e0892ac0144fa92562fd17131beca1864a41ca1e322
-size 819730354
+oid sha256:224c7f00e38af5098c1d54ad271c744a397d7cd41e096609885e2b06c50fa9fa
+size 409647917

GLOW-FDG/fold_3/checkpoint_final.pth

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:6a194b0a45fc0d9c9c1e136939ed8e1b930123a7b741c2f6ef55d1b8ba04fcb5
-size 819730354
+oid sha256:6f5a986370c834afe26a5dfc561c22035f427f0eda60cfa65f88136676276ec3
+size 409647917

GLOW-FDG/fold_4/checkpoint_final.pth

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:8be831064af91b597fc72e9071e828541d891954c98958e344e5baf17696aa06
-size 819730418
+oid sha256:eca663ec68cf08b0a9b72cc328e9c1536173990265def3bb46477d478f4d5565
+size 409647917

README.md

@@ -84,7 +84,7 @@ pip install nnunetv2
 ```python
 from huggingface_hub import snapshot_download
 
-model_dir = snapshot_download(repo_id="<org>/GLOW-FDG")
+model_dir = snapshot_download(repo_id="mrokuss/GLOW-FDG")
 # model_dir/GLOW-FDG/ now contains dataset.json, plans.json and fold_0..fold_4
 ```
 
@@ -94,12 +94,14 @@ Each case must contain two co-registered channels following the nnU-Net naming c
 
 ```
 input_folder/
-  CASE001_0000.nii.gz   # CT (HU)
-  CASE001_0001.nii.gz   # PET (SUV body-weight normalized)
+  CASE001_0000.nii.gz   # CT  (Hounsfield units)
+  CASE001_0001.nii.gz   # PET (body-weight SUV)
 ```
 
 PET intensities **must** be converted to body-weight SUV before inference.
 
+> **Why does `dataset.json` list both channels as `"CT"`?** nnU-Net uses the `"CT"` channel name as a signal to apply `CTNormalization`, which assumes inputs are in absolute physical units (no per-case rescaling). Both Hounsfield units (CT) and body-weight SUV (PET) are absolute units, so the same normalization scheme is applied to both — this is intentional and **not** a mistake. Channel `_0000` is still CT and channel `_0001` is still PET; the label is only about which normalization scheme nnU-Net selects.
+
 ### 4. Run inference
 
 ```python
@@ -132,6 +134,18 @@ predictor.predict_from_files(
 
 The output is a binary NIfTI mask per case where `1` denotes predicted FDG-avid cancer lesions.
 
+### Alternative: nnU-Net CLI
+
+If you prefer the standard nnU-Net command-line interface, point `nnUNetv2_predict_from_modelfolder` directly at the downloaded `GLOW-FDG/` folder — no `nnUNet_results` directory layout required:
+
+```bash
+nnUNetv2_predict_from_modelfolder \
+  -i  input_folder \
+  -o  output_folder \
+  -m  /path/to/GLOW-FDG \
+  -f  0 1 2 3 4
+```
+
 ## Training Data
 
 GLOW-FDG was trained on 1,563 FDG-PET/CT scans pooled from public and institutional sources, including AutoPET, HECKTOR, DEEP-PSMA, ACRIN-HNSCC, HN-PET-CT, NSCLC-RadGen, TCIA-STS, SAKK, and the SINERGIA melanoma cohort. All cases were manually reviewed to verify PET–mask correspondence, lesion completeness, and PET visibility. Organ labels for the auxiliary head were generated with [TotalSegmentator](https://github.com/wasserth/TotalSegmentator).