luna16

LUNA16-Lung-Nodule-Analysis-2016-Challenge

Developing a well-documented repository for the Lung Nodule Detection task on the Luna16 dataset. This work is inspired by the ideas of the first-placed team at DSB2017, "grt123".

3D-UCaps: 3D Capsules Unet for Volumetric Image Segmentation (MICCAI 2021)

A Clone version from Original SegCaps source code with enhancements on MS COCO dataset.

This study presents the development and validation of AI models for both nodule detection and cancer classification tasks. This benchmarking across multiple datasets establishes the DLCSD as a reliable resource for lung cancer AI research.

Lung nodule detection- LUNA 16

Lung Nodules Segmentation from CT scans using CNN.

Boost lung Cancer Detection using Generative model and Semi-Supervised Learning

Lung cancer detection using deep learning models

Binary tumor-vs-healthy classification of pulmonary nodules from LUNA16 CT scans using DenseNet121 transfer learning, with a cached patch-extraction pipeline and exploratory DICOM/PET-CT notebooks.

Python script that extract images from LUNA16 dataset to a human readable format

Implementation of lung nodule detection and false positive reduction using CT images for LUNA16-LUng-Nodule-Analysis-2016-Challenge.

In progress (data acquisition); CT lung nodule classifier for the LUNA16 false-positive reduction task (888 scans, 1,186 nodules). Plan: SimpleITK preprocessing, PyRadiomics + gradient-boosting baseline, then a PyTorch CNN; scan-level 10-fold cross-validation; official FROC/CPM evaluation; FastAPI + Docker deployment.

( Specialized for the LUNA16/DBT datasets )

Full Processing on Luna16 Challange (All data resampled to 1mm x 0.7mm x 0.7mm)

Official codebase for my Year 3 dissertation on 3D pulmonary nodule segmentation using LUNA16 dataset. Compares a supervised 3D U-Net baseline vs. semi-supervised Mean Teacher architecture (peak validation Dice: 0.6617, with a DSC of 81.77%). Includes model definitions, metadata, and key NIfTI qualitative visual result files.