|
|
--- |
|
|
license: apache-2.0 |
|
|
task_categories: |
|
|
- zero-shot-classification |
|
|
size_categories: |
|
|
- 10M<n<100M |
|
|
--- |
|
|
Leukemia Detection and Classification Model |
|
|
Model Description |
|
|
This deep learning model is designed for the detection and classification of leukemia from medical images. The model can identify cancerous cells and classify different types or stages of leukemia, providing automated assistance for medical diagnosis. |
|
|
Model Details |
|
|
|
|
|
Model Type: [Specify: CNN/ResNet/VGG/YOLO/EfficientNet/etc.] |
|
|
Task: Multi-class classification and detection |
|
|
Input: Medical microscopy images of blood cells |
|
|
Output: Classification labels and/or bounding boxes for detected leukemia cells |
|
|
Framework: [PyTorch/TensorFlow/Keras] |
|
|
License: Apache 2.0 |
|
|
|
|
|
Intended Use |
|
|
Primary Use Case |
|
|
This model is intended to assist medical professionals in: |
|
|
|
|
|
Early detection of leukemia from blood cell images |
|
|
Classification of leukemia subtypes (ALL, AML, CML, CLL) |
|
|
Screening and diagnostic support in clinical settings |
|
|
Research and educational purposes |
|
|
|
|
|
Direct Use |
|
|
The model can be used directly for inference on microscopic blood cell images to detect and classify leukemia. |
|
|
Downstream Use |
|
|
|
|
|
Integration into clinical diagnostic systems |
|
|
Medical image analysis pipelines |
|
|
Research tools for hematology studies |
|
|
Educational platforms for medical training |
|
|
|
|
|
Out-of-Scope Use |
|
|
⚠️ Important Limitations: |
|
|
|
|
|
This model is NOT a replacement for professional medical diagnosis |
|
|
Should NOT be used as the sole basis for treatment decisions |
|
|
Requires validation by qualified healthcare professionals |
|
|
Not intended for use without proper medical oversight |
|
|
|
|
|
Training Details |
|
|
Training Data |
|
|
Dataset: [Specify your dataset name/source] |
|
|
|
|
|
Number of images: [X samples] |
|
|
Classes: [ Normal, ALL, AML, CML, CLL] |
|
|
Image resolution: [224x224, 640x640] |
|
|
Data split: [70% train, 15% validation, 15% test] |
|
|
|
|
|
Preprocessing: |
|
|
|
|
|
Image resizing to [dimensions] |
|
|
Normalization |
|
|
Data augmentation (rotation, flipping, brightness adjustment) |
|
|
[Other preprocessing steps] |
|
|
|
|
|
Training Procedure |
|
|
Hyperparameters: |
|
|
|
|
|
Optimizer: [ Adam, SGD] |
|
|
Learning rate: [0.001] |
|
|
Batch size: [ 32] |
|
|
Epochs: [ 100] |
|
|
Loss function: [ CrossEntropyLoss, Focal Loss] |
|
|
[Additional parameters] |
|
|
|
|
|
Training Environment: |
|
|
|
|
|
Hardware: [ NVIDIA GPU, Google Colab] |
|
|
Training time: [approximate duration] |
|
|
|
