# lfz/DSB2017

The solution of team 'grt123' in DSB2017

Repository: https://github.com/lfz/DSB2017
Canonical: https://ross.abutalabs.com/products/dsb2017
Language: Python
License: MIT
License Family: permissive
Last push: 2023-06-28T10:35:43+00:00

## Health v2 (maintenance only)
Score: 32/100 (v2, computed 2026-09-03T02:20:16.233290+00:00)
- activity 0, release rhythm 35, longevity 100
- inputs: {"age_days": 3408, "days_push": 1162, "days_rel": null, "gap_med": null, "n_releases_24m": 0}
- flags: no_releases
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 1241, forks 412 (observed 2026-08-28T04:04:06.260449+00:00)

## What it is
The winning solution of team 'grt123' for the 2017 Data Science Bowl (DSB2017), a deep learning pipeline for detecting lung cancer from CT scans. It uses a 3D Faster R-CNN nodule detector built on a U-net backbone plus a cancer classifier, implemented in PyTorch.

## Use cases
- detect lung cancer from CT scans
- train a 3D nodule detector on LUNA16 data
- reproduce the DSB2017 winning solution
- run lung nodule detection on stage 2 test data
- classify lung nodules as cancerous using deep learning
- learn 3D Faster R-CNN for medical imaging

## When to choose
- you need a proven lung cancer screening pipeline on CT scans
- you want to reproduce or build on the DSB2017 winning approach
- you have LUNA16 and stage1 data and multiple GPUs for training

## When to avoid
- you need a maintained, production-ready medical imaging tool
- you work outside Python 2.7 / CUDA 8 legacy environments
- you lack significant GPU resources, as training takes days on 8 GPUs

## Facets
- artifact type: application
- maturity: maintenance
- function: deep-learning, machine-learning, image-processing, computer-vision
- domain: healthcare, deep-learning, machine-learning, computer-vision
- platform: python
- tags: lung-cancer-detection, medical-imaging, kaggle-competition, 3d-convolutional-networks, faster-rcnn, pytorch, ct-scans, luna16, linux, gpu

## Member repositories
- lfz/DSB2017 (main) score 32

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:04:06.260449+00:00.
- Health v2: computed from the inputs above; adoption is never an input.
- Inferred fields (summary, facets, guidance): AI-extracted, prompt v1, taxonomy v1, on 2026-08-30T05:08:29.414958+00:00, confidence not recorded.
  - readme: https://github.com/lfz/DSB2017 (fetched 2026-08-28T04:04:06.260449+00:00, sha 550353928fdc)
- Data as of 2026-08-30T08:39:29.467469+00:00.