update

OAIZIB-CM.py

@@ -1,114 +1,168 @@
 import os
 import pandas as pd
 import nibabel as nib
-from datasets import GeneratorBasedBuilder, SplitGenerator, Split, DatasetInfo, Features, Array3D, Value
+import datasets
+from datasets import (
+    GeneratorBasedBuilder,
+    SplitGenerator,
+    Split,
+    DatasetInfo,
+    Features,
+    Array3D,
+    Value,
+)
+
+logger = datasets.logging.get_logger(__name__)
+
+_CITATION = """\
+@article{YAO2024103035,
+  title = {CartiMorph: A framework for automated knee articular cartilage morphometrics},
+  journal = {Medical Image Analysis},
+  author = {Yongcheng Yao and Junru Zhong and Liping Zhang and Sheheryar Khan and Weitian Chen},
+  volume = {91},
+  pages = {103035},
+  year = {2024},
+  issn = {1361-8415},
+  doi = {https://doi.org/10.1016/j.media.2023.103035}
+}
+
+@InProceedings{10.1007/978-3-031-82007-6_16,
+  author = {Yao, Yongcheng and Chen, Weitian},
+  editor = {Wu, Shandong and Shabestari, Behrouz and Xing, Lei},
+  title = {Quantifying Knee Cartilage Shape and Lesion: From Image to Metrics},
+  booktitle = {Applications of Medical Artificial Intelligence},
+  year = {2025},
+  publisher = {Springer Nature Switzerland},
+  address = {Cham},
+  pages = {162--172}
+}
+"""
+
+_DESCRIPTION = """\
+This is the official release of the OAIZIB-CM dataset.
+
+OAIZIB-CM is based on the OAIZIB dataset.
+OAIZIB paper: Automated Segmentation of Knee Bone and Cartilage combining Statistical Shape Knowledge 
+and Convolutional Neural Networks: Data from the Osteoarthritis Initiative.
+In OAIZIB-CM, tibial cartilage is split into medial and lateral tibial cartilages.
+OAIZIB-CM includes CLAIR-Knee-103R, consisting of:
+  - a template image learned from 103 MR images of subjects without radiographic OA,
+  - a corresponding 5-ROI segmentation mask for cartilages and bones, and
+  - a corresponding 20-ROI atlas for articular cartilages.
+
+This dataset is released under the CC BY-NC 4.0 license.
+"""
+
+_HOMEPAGE_URL = "https://github.com/YongchengYAO/CartiMorph"
+
 
 class OAIZIBCMDataset(GeneratorBasedBuilder):
-    VERSION = "1.0.0"
+    VERSION = datasets.Version("1.0.0")
 
     def _info(self):
         return DatasetInfo(
-            description=(
-                "This is the official release of the OAIZIB-CM dataset.\n\n"
-                "OAIZIB-CM is based on the OAIZIB dataset.\n"
-                "OAIZIB paper: Automated Segmentation of Knee Bone and Cartilage combining Statistical Shape Knowledge "
-                "and Convolutional Neural Networks: Data from the Osteoarthritis Initiative.\n"
-                "In OAIZIB-CM, tibial cartilage is split into medial and lateral tibial cartilages.\n"
-                "OAIZIB-CM includes CLAIR-Knee-103R, consisting of:\n"
-                "  - a template image learned from 103 MR images of subjects without radiographic OA,\n"
-                "  - a corresponding 5-ROI segmentation mask for cartilages and bones, and\n"
-                "  - a corresponding 20-ROI atlas for articular cartilages.\n\n"
-                "Papers:\n"
-                "The dataset originates from these projects:\n"
-                " - CartiMorph: https://github.com/YongchengYAO/CartiMorph\n"
-                " - CartiMorph Toolbox: https://github.com/YongchengYAO/CartiMorph-Toolbox\n"
-                " - https://github.com/YongchengYAO/CMT-AMAI24paper\n\n"
-                "Please cite the following if you use the dataset:\n\n"
-                "@article{YAO2024103035,\n"
-                "  title = {CartiMorph: A framework for automated knee articular cartilage morphometrics},\n"
-                "  journal = {Medical Image Analysis},\n"
-                "  author = {Yongcheng Yao and Junru Zhong and Liping Zhang and Sheheryar Khan and Weitian Chen},\n"
-                "  volume = {91},\n"
-                "  pages = {103035},\n"
-                "  year = {2024},\n"
-                "  issn = {1361-8415},\n"
-                "  doi = {https://doi.org/10.1016/j.media.2023.103035}\n"
-                "}\n\n"
-                "@InProceedings{10.1007/978-3-031-82007-6_16,\n"
-                "  author = {Yao, Yongcheng and Chen, Weitian},\n"
-                "  editor = {Wu, Shandong and Shabestari, Behrouz and Xing, Lei},\n"
-                "  title = {Quantifying Knee Cartilage Shape and Lesion: From Image to Metrics},\n"
-                "  booktitle = {Applications of Medical Artificial Intelligence},\n"
-                "  year = {2025},\n"
-                "  publisher = {Springer Nature Switzerland},\n"
-                "  address = {Cham},\n"
-                "  pages = {162--172}\n"
-                "}\n\n"
-                "License:\n"
-                "This dataset is released under the CC BY-NC 4.0 license. It is compulsory to cite the above papers if you use the dataset.\n"
+            description=_DESCRIPTION,
+            features=Features(
+                {
+                    # Let datasets library infer the shape dimensions
+                    "image": Array3D(dtype="float32"),
+                    "mask": Array3D(dtype="float32"),
+                    "image_path": Value("string"),
+                    "mask_path": Value("string"),
+                }
             ),
-            features=Features({
-                "image": Array3D(dtype="float32", shape=(None, None, None)),
-                "mask": Array3D(dtype="float32", shape=(None, None, None)),
-                "image_path": Value("string"),
-                "mask_path": Value("string"),
-            }),
+            homepage=_HOMEPAGE_URL,
+            citation=_CITATION,
         )
 
     def _split_generators(self, dl_manager):
-        # Define the base URL where the CSV files are hosted on the HF Hub.
-        # This URL is constructed to point to the files from the desired revision.
+        # Download metadata files (CSV) that describe the dataset
         base_url = "https://huggingface.co/datasets/YongchengYAO/OAIZIB-CM/resolve/load_dataset-support"
         train_csv_url = f"{base_url}/train.csv"
         test_csv_url = f"{base_url}/test.csv"
-        
-        # Use dl_manager to download the CSV files
-        train_csv_path = dl_manager.download(train_csv_url)
-        test_csv_path = dl_manager.download(test_csv_url)
-        
-        # Debug: print the downloaded paths (optional)
-        print("Downloaded Train CSV Path:", train_csv_path)
-        print("Downloaded Test CSV Path:", test_csv_path)
-        
-        train_df = pd.read_csv(train_csv_path)
-        test_df = pd.read_csv(test_csv_path)
-        
+
+        # Download CSV files containing dataset metadata
+        csv_paths = dl_manager.download({"train": train_csv_url, "test": test_csv_url})
+
+        logger.info(f"Downloaded CSV paths: {csv_paths}")
+
+        # Check if local data directory exists
+        data_dir = os.path.join(os.getcwd(), "data")
+
+        train_df = pd.read_csv(csv_paths["train"])
+        test_df = pd.read_csv(csv_paths["test"])
+
         return [
-            SplitGenerator(name=Split.TRAIN, gen_kwargs={"df": train_df, "split": "train"}),
-            SplitGenerator(name=Split.TEST, gen_kwargs={"df": test_df, "split": "test"}),
+            SplitGenerator(
+                name=Split.TRAIN,
+                gen_kwargs={"df": train_df, "split": "train", "data_dir": data_dir},
+            ),
+            SplitGenerator(
+                name=Split.TEST,
+                gen_kwargs={"df": test_df, "split": "test", "data_dir": data_dir},
+            ),
         ]
 
-    def _generate_examples(self, df, split):
-        # Determine base directories for images and masks based on the split
+    def _generate_examples(self, df, split, data_dir):
+        # Set up directory paths based on the split
         if split == "train":
-            base_img_dir = os.path.join("data", "imagesTr")
-            base_mask_dir = os.path.join("data", "labelsTr")
+            img_dir = os.path.join(data_dir, "imagesTr")
+            mask_dir = os.path.join(data_dir, "labelsTr")
         elif split == "test":
-            base_img_dir = os.path.join("data", "imagesTs")
-            base_mask_dir = os.path.join("data", "labelsTs")
+            img_dir = os.path.join(data_dir, "imagesTs")
+            mask_dir = os.path.join(data_dir, "labelsTs")
         else:
-            base_img_dir = ""
-            base_mask_dir = ""
+            raise ValueError(f"Unknown split: {split}")
+
+        # Log directories for debugging
+        logger.info(f"Looking for {split} images in: {img_dir}")
+        logger.info(f"Looking for {split} masks in: {mask_dir}")
+
+        # Verify directories exist
+        os.makedirs(img_dir, exist_ok=True)
+        os.makedirs(mask_dir, exist_ok=True)
+
+        # Track yield count for debugging
+        count = 0
+        skipped = 0
 
         for idx, row in df.iterrows():
             img_file = row["image"]
             mask_file = row["mask"]
-            img_path = os.path.join(base_img_dir, img_file)
-            mask_path = os.path.join(base_mask_dir, mask_file)
-            try:
-                image = nib.load(img_path).get_fdata().astype("float32")
-            except Exception as e:
-                print(f"Error loading image from {img_path}: {e}")
+            img_path = os.path.join(img_dir, img_file)
+            mask_path = os.path.join(mask_dir, mask_file)
+
+            # Check if files exist
+            if not os.path.exists(img_path):
+                logger.warning(f"Image not found: {img_path}")
+                skipped += 1
+                continue
+
+            if not os.path.exists(mask_path):
+                logger.warning(f"Mask not found: {mask_path}")
+                skipped += 1
                 continue
+
             try:
-                mask = nib.load(mask_path).get_fdata().astype("float32")
+                # Load and convert the image and mask data
+                img_nib = nib.load(img_path)
+                image = img_nib.get_fdata().astype("float32")
+
+                mask_nib = nib.load(mask_path)
+                mask = mask_nib.get_fdata().astype("float32")
+
+                yield idx, {
+                    "image": image,
+                    "mask": mask,
+                    "image_path": img_path,
+                    "mask_path": mask_path,
+                }
+                count += 1
+
             except Exception as e:
-                print(f"Error loading mask from {mask_path}: {e}")
-                continue
+                logger.error(f"Error processing {img_path} and {mask_path}: {e}")
+                skipped += 1
 
-            yield idx, {
-                "image": image,
-                "mask": mask,
-                "image_path": img_path,
-                "mask_path": mask_path,
-            }
+        logger.info(
+            f"Successfully yielded {count} examples for {split} split, skipped {skipped}"
+        )