AI Intern Medical Image Segmentation (Dental CBCT/IOS/STL) | Remote | 3 Months

Luein Analytics is building a clinical AI product for dental implant planning. We are looking for an AI intern to own a focused, high-impact piece of it: turning 3D dental scans (CBCT/DICOM/IOS/STL) into segmented anatomy, and turning those segmentations into clinically meaningful measurements & data points.

What you will work on

• Evaluate and run open-source CBCT segmentation models (e.g. nnU-Net-based models such as DentalSegmentator) against a commercial baseline, and document accuracy and failure cases
• Build a local inference pipeline: DICOM in, segmentation masks and STL meshes out
• Write Python code to extract measurements from masks and meshes: bone-to-nerve distance, ridge width/height, bone density (grey-value) readouts, soft tissue thickness
• Validate your outputs against reference cases and flag where results aren't trustworthy
• Work directly with the team building the rule engine that consumes your measurements

You should have

• Strong Python (NumPy, SciPy, scikit-image)
• Hands-on experience with PyTorch and at least one segmentation project (U-Net, nnU-Net, or MONAI), whether academic, personal, or Kaggle
• Comfort with 3D data: volumes, meshes, coordinate systems, voxel spacing
• Familiarity with DICOM/NIfTI handling (pydicom, SimpleITK, nibabel) or willingness to learn fast

Good to have

• Experience with 3D Slicer, trimesh, or Open3D
• Prior exposure to medical imaging (any modality)
• Located in Bangalore (for occasional in-person syncs)

What you will get

• Real ownership of a live product component, not a sandbox task
• Direct mentorship from the team building the platform
• Stipend: 8,000/month

Pre-placement offer (full-time role) for strong performers

Details: 3 months · Remote (Bangalore preferred) ·

Start date: 1st October 2026