NoduleMNIST3D

Dataset Information

The NoduleMNIST3D is based on the LIDC-IDRI, a large public lung nodule dataset, containing images from thoracic CT scans. The dataset is designed for both lung nodule segmentation and 5-level malignancy classification task. To perform binary classification, we categorize cases with malignancy level 1/2 into negative class and 4/5 into positive class, ignoring the cases with malignancy level 3. We split the source dataset with a ratio of 7:1:2 into training, validation and test set, and center-crop the spatially normalized images (with a spacing of 1mm×1mm×1mm) into 28×28×28.

Task: binary-class

Labels:

0: benign, 1: malignant

Samples:

• Train: 1158
• Validation: 165
• Test: 310

Experiment Parameter

Learning Rate: 1e-5

Training Epoch: 100

Convergence Epoch: 40