Lung Cancer Detection is a deep-learning project that classifies lung CT scan images into three categories based on cancer risk.
The model leverages YOLOv8, a powerful neural network architecture designed for image classification and object detection.
The dataset contains 1,054 medical lung scan images, curated and hosted on Roboflow Universe, and a trained model is already available for experimentation and inference.
| Class Type | Description |
|---|---|
| Normal | Lungs showing no visible abnormalities or cancer-related nodules. |
| Benign | Non-cancerous nodules or minor abnormalities detected. |
| Malignant | Cancerous nodules or signs strongly indicating lung cancer. |
| Parameter | Value |
|---|---|
| Dataset Version | v1 |
| Total Images | 1,054 |
| Model Version | YOLOv8 Classification Model (v1) |
| Dataset Host | Roboflow Universe |
The dataset and current model release are stable. Future dataset updates and incremental versions are planned.
🔹 Live Demo (Gradio App via Hugging Face Spaces):
👉 https://huggingface.co/spaces/bedead/Lung-cancer-classification
🔹 Dataset (Roboflow Universe):
👉 https://universe.roboflow.com/satyam-mishra-gfl0c/lung-cancer-detection-afeuf
To maintain dataset quality and medical reliability:
- Use only anonymized and ethically sourced medical images.
- Label using the following categories only:
Normal,Benign,Malignant. - Ensure image clarity (CT/X-ray format recommended).
- Add metadata only if anonymized and ethically permissible.
- Expert cross-verification is recommended due to medical sensitivity.
📧 For contribution or dataset collaboration, contact:
Satyam Mishra — [email protected]
Patent Application Number:
202341039647
This project and methodology are protected under Indian patent application laws.
- YOLOv8 (Ultralytics)
- Python
- Roboflow
- Gradio (Demo UI)
- Hugging Face Spaces
If you found this useful, don’t forget to ⭐ star the repository!