Principal Scientist, Imaging Analytics

Johnson & Johnson Johnson & Johnson · Pharma · New Brunswick, NJ +5

Principal Scientist role focused on applying AI/ML to medical imaging (CT, MRI, PET) for oncology drug development. The role involves developing quantitative imaging measures, predictive models, and insights to inform clinical decisions. It requires hands-on research, collaboration with internal and external partners, and publication of scientific innovation. The position bridges data science, clinical development, and external vendors, focusing on translating imaging AI from research to clinical impact.

What you'd actually do

  1. Lead end-to-end AI applications on trial imaging data (CT, MRI, PET) for quantitative imaging measures and AI-derived endpoints.
  2. Collaborate internally and externally to drive scientific innovation in foundational imaging AI that are relevant to oncology drug development — including automated segmentation, radiomics, and multimodal predictive modeling through hands-on research.
  3. Translate imaging-derived evidence into actionable insights by converting complex quantitative findings into clear scientific narratives and engaging cross-functional stakeholders
  4. Provide scientific leadership to external partnerships — including imaging AI vendors, CROs, biomarker companies, academic centers, and imaging OEMs — to accelerate model development, validation, and deployment.
  5. Publish and present scientific innovation at top scientific and clinical conferences (e.g., MICCAI, AACR, RSNA, etc.)

Skills

Required

  • Ph.D. in Computer Science, Biomedical Engineering, Electrical Engineering, or a related quantitative field.
  • 3+ years of post-doctoral or industry experience developing AI/ML for medical imaging (CT, MRI, PET) in a clinical setting.
  • Deep learning (segmentation, detection, classification, registration)
  • Radiomics
  • Multimodal predictive modeling
  • Python
  • PyTorch
  • MONAI
  • SimpleITK
  • ITK
  • PyRadiomics
  • nnU-Net
  • 3D Slicer
  • OpenCV
  • Cloud ML infrastructure
  • MLOps practices
  • DICOM I/O
  • visualization
  • registration
  • harmonization
  • annotation
  • segmentation of 3D medical images
  • Peer-reviewed publication record
  • Communication of complex scientific concepts

Nice to have

  • Analyzing solid-tumor imaging, particularly lung and head & neck (H&N).
  • Developing and applying AI/ML to oncology imaging and within oncology clinical trials.
  • Familiarity with standard oncology endpoints (e.g., RECIST 1.1, iRECIST, PERCIST).
  • Building and scaling clinical or imaging AI platforms end-to-end, including data ingestion, harmonization, model inference, visualization, and continuous monitoring.
  • Sourcing, structuring, and managing external partnerships with imaging-AI vendors, CROs / imaging core labs, biomarker companies, and academic centers.

What the JD emphasized

  • Ph.D. in Computer Science, Biomedical Engineering, Electrical Engineering, or a related quantitative field.
  • 3+ years of post-doctoral or industry experience developing AI/ML for medical imaging (CT, MRI, PET) in a clinical setting.
  • Hands-on expertise across the medical imaging AI stack: deep learning (segmentation, detection, classification, registration), radiomics, and multimodal predictive modeling.
  • Proficiency in Python and PyTorch, with practical experience in medical-imaging libraries such as MONAI, SimpleITK, ITK, PyRadiomics, nnU-Net, 3D Slicer, and OpenCV.
  • Experience with cloud ML infrastructure and MLOps practices for scalable training and inference on imaging data.
  • Extensive experience with the full imaging data workflow: DICOM I/O, visualization, registration, harmonization, annotation, and segmentation of 3D medical images.
  • Strong peer-reviewed publication record and demonstrated ability to communicate complex scientific concepts to both technical and cross-functional audiences.

Other signals

  • AI/ML for medical imaging
  • quantitative imaging metrics
  • predictive modeling