Note: The job is a remote job and is open to candidates in USA. ProVoyance builds AI-driven medical imaging software that converts radiological scans into detailed 3D anatomical models for procedure planning and navigation. The role involves developing deep learning algorithms for image-guided interventions, working closely with AI scientists, software engineers, and clinicians to take algorithms from research through validation and into FDA-cleared products.
Responsibilities
- Design, train, and evaluate deep learning models for segmentation, landmark detection, registration, and 3D anatomical modeling
- Develop AI supporting image-guided interventions including surgical planning, navigation, and tumor ablation
- Build image processing pipelines for DICOM/NIfTI volumetric data across X-ray, CT, and MRI
- Analyze large clinical imaging datasets and improve data quality
- Design reproducible experiments and evaluate performance with appropriate statistical methods
- Write production-quality, well-tested code and collaborate with engineering on integration
- Support V&V activities for regulated medical device development
Skills
- Bachelor's in CS, Biomedical/Electrical/Computer Engineering, Applied Math, Physics, or related
- 2+ years professional or graduate research experience in medical image analysis, computer vision, or deep learning (graduate research and internships count)
- Proficiency in Python, PyTorch, NumPy, Git
- Experience with at least one medical imaging library: MONAI, ITK/SimpleITK, VTK, or OpenCV
- Familiarity with DICOM/NIfTI and 3D volumetric images (CT, MRI, X-ray)
- Working knowledge of CNNs, U-Net, Vision Transformers, segmentation, and registration
- Comfortable in Linux environments
- Master's in a related field
- 3D deep learning; nnU-Net, MONAI, TorchIO, or TotalSegmentator
- Model optimization: CUDA, TensorRT, ONNX
- Docker and cloud (GCP/Vertex AI a plus given our current migration; AWS also relevant)
- Reproducible ML pipelines and experiment tracking (MLflow, W&B)
- Annotation tooling (3D Slicer, ITK-SNAP)
- Regulated medical device development; familiarity with IEC 62304, ISO 13485, or GMLP
- Open-source contributions or publications
Company Overview