i4h-catheter-navigation-setup

作成者: nvidia

カテーテルナビゲーションワークフローのホスト/GPU要件とPYTHONPATHを確認してください。catheter_navigationのセットアップ、インストール、またはブートストラップを求められた場合に使用します。また…

npx skills add https://github.com/nvidia/skills --skill i4h-catheter-navigation-setup

i4h Catheter Navigation - Setup

Purpose

Verify host and GPU requirements, confirm the ./i4h CLI sees the workflow, and run CPU smoke tests. Use when asked to set up catheter navigation or when hitting missing imports, GPU, or slangpy errors.

Base Code

These steps drive the i4h-workflows base code (the workflows/catheter_navigation/ tree). To reuse an existing checkout, set I4H_WORKFLOWS to its path (no clone happens). Otherwise this resolves the current repo, or clones to ~/i4h-workflows - pick that default without prompting. Run every command below from the resolved root:

ROOT="${I4H_WORKFLOWS:-$(git rev-parse --show-toplevel 2>/dev/null)}"
if [ ! -d "$ROOT/workflows/catheter_navigation" ]; then
  ROOT="${I4H_WORKFLOWS:-$HOME/i4h-workflows}"
  [ -d "$ROOT/workflows/catheter_navigation" ] || git clone https://github.com/isaac-for-healthcare/i4h-workflows "$ROOT"
fi
export I4H_WORKFLOWS="$ROOT"; cd "$ROOT"

Basics

  • Catheter navigation registers via workflows/catheter_navigation/metadata.json and runs through ./i4h run catheter_navigation <mode>.
  • Runtime is package-first: render_drr uses installed fluorosim (python -m fluorosim.examples.render_drr), while interactive_viewport is launched from the local workflow script path in metadata.json.
  • Docker image: workflows/catheter_navigation/docker/Dockerfile (drop --local on ./i4h run to use it).
  • GPU modes need slangpy, Warp, and CUDA; CPU smoke tests do not.

Preflight

command -v python3
command -v git
nvidia-smi
df -h .

Required: Linux x86_64 (Ubuntu 22.04/24.04 tested), NVIDIA GPU (CC >= 7.0), driver compatible with CUDA 12.8, >= 16 GB RAM, >= 20 GB disk.

Run the steps below in order. Each step is a separate bash call; variables persist in the local agent's tmux session.

Step 1 - resolve repo and export paths

REPO_ROOT="${I4H_WORKFLOWS:-$(git rev-parse --show-toplevel 2>/dev/null)}"; [ -d "$REPO_ROOT/workflows/catheter_navigation" ] || REPO_ROOT="$HOME/i4h-workflows"
WF_ROOT="${REPO_ROOT}/workflows/catheter_navigation"
SIM_ROOT="${WF_ROOT}/scripts/simulation"
export PYTHONPATH="${SIM_ROOT}:${PYTHONPATH:-}"
RUN_DIR="${WF_ROOT}/runs/setup_$(date +%Y%m%d_%H%M%S)"
mkdir -p "${RUN_DIR}/logs"
ln -sfn "${RUN_DIR}" "${WF_ROOT}/runs/.latest"

Step 2 - verify CLI registration

"${REPO_ROOT}/i4h" modes catheter_navigation 2>&1 | tee "${RUN_DIR}/logs/modes.log"

Step 3 - CPU smoke tests (no GPU)

python3 -m unittest workflows/catheter_navigation/tests/test_fluorosim_smoke.py \
  2>&1 | tee "${RUN_DIR}/logs/smoke.log"

Expected: Ran 7 tests ... OK. Parser error lines in stderr from negative test cases are expected.

Step 4 - optional GPU sanity (synthetic DRR)

Skip if no GPU or slangpy not installed.

"${REPO_ROOT}/i4h" run catheter_navigation render_drr --local \
  --run-args="--output ${RUN_DIR}/drr.png" \
  2>&1 | tee "${RUN_DIR}/logs/render_drr.log"

Verify

test -f "${RUN_DIR}/logs/smoke.log"
grep -q "OK" "${RUN_DIR}/logs/smoke.log"
python3 -c "import fluorosim; print('fluorosim', fluorosim.__file__)"

Prerequisites

  • Repo checkout with workflows/catheter_navigation/ present.
  • Python 3 with numpy; full GPU stack (slangpy, torch CUDA, warp) for render/viewport modes.

Limitations

  • No dedicated setup.sh yet - this skill verifies and documents host requirements; use Docker when host deps are incomplete.
  • Step 4 requires a GPU; Step 3 alone is sufficient for CI-style verification.

Troubleshooting

  • Error: fluorosim import fails - Cause: PYTHONPATH not set. Fix: re-run Step 1; confirm SIM_ROOT exists.
  • Error: ./i4h not found - Cause: not at repo root. Fix: cd "$REPO_ROOT" where ./i4h lives.
  • Error: render_drr fails with slang/GPU - Cause: missing CUDA or slangpy. Fix: use Docker (./i4h run catheter_navigation render_drr without --local) or install deps per README.

Final Response

Report setup status, smoke-test result, optional DRR output path, and recommend the next skill ([[i4h-catheter-navigation-digital-twin]] for patient data, [[i4h-catheter-navigation-viewport]] for interactive demo).

nvidiaのその他のスキル

compileiq-debug
nvidia
何かがおかしいときに使用:Search()がハングする、すべての評価がINVALID_SCOREを返す、スコアが改善しない、すべての設定が同じ数値を返す、ptxasエラー…
create-github-pr
nvidia
gh CLIを使用してGitHubのプルリクエストを作成します。ユーザーが新しいPRを作成したい、コードをレビューに提出したい、またはプルリクエストを開きたい場合に使用します。トリガーキーワード -…
nemoclaw-maintainer-cross-issue-sweep
nvidia
他のオープンなIssueをスキャンし、特定のPRが修正する可能性があるものや、誤って壊す可能性があるものを見つけます。隣接修正の機会や矛盾リスクをfile:line…と共に出力します。
fhir-basics
nvidia
エージェントにFHIR R4 APIの動作方法、利用可能なリソース、検索パラメータを使ったクエリ方法、およびすべてのレスポンス形式を正しく解析する方法を教えます…
compileiq-validate-result
nvidia
検索が完了した後、かつスピードアップの申請やACFの発送の前に使用します。dump_results CSVを読み込み、トップK候補(単一目的)を抽出します…
changelog-audit
nvidia
リリース前にWarp CHANGELOG.mdを監査:失われたエントリを復元、ユーザー影響で並べ替え、エントリの文言を洗練、行折り返し、および(リリースブランチモードで)比較をバンプ…
maintain-dynamic-plugins
nvidia
NeMo Relayの動的プラグインローダー、マニフェスト、RustネイティブSDK、gRPCワーカープロトコル、PythonワーカーSDK、ドキュメント、テスト、およびリリースワークフローのカバレッジを維持する
dgx-diagnose
nvidia
一般的なDGX Station GB300の問題(CUDAクラッシュ、誤ったGPUターゲット、vLLM/SGLangコンテナのバグ、MIG状態の問題、NVLink/Fabric Managerエラーなど)を診断します。