AI CBCT Lesion Segmentation Plugin
Краткое
Freelancer Client is hiring: AI CBCT Lesion Segmentation Plugin.
Location: Remote
I have a collection of DICOM CBCT volumes exported from Sirona GALILEOS / SIDEXIS and I need a 3D Slicer extension that streamlines research-grade periapical lesion analysis. The core priority is highly accurate detection and segmentation; everything else follows from that.
The workflow I picture starts with frictionless DICOM import, passes the volumes through a pre-trained model for initial lesion detection, lets the user refine or approve the mask, and then automatically computes total lesion volume in mm³. Once a baseline study is finished, the same tool should accept a follow-up scan, register it to the baseline, and output comparable measurements so longitudinal change can be plotted immediately.
3D Slicer extension (Python scripted module or C++ loadable, whichever integrates best)
Integration of a pre-trained deep learning network for periapical lesion suggestion
Interactive editing tools leveraging Slicer’s Segment Editor for quick correction
Skills: C Programming, Software Architecture, C++ Programming, Computer Vision
Budget: $30–$250 USD
Source: Freelancer Client via Remote / Online. Apply on the source website.
Оригинал
I have a collection of DICOM CBCT volumes exported from Sirona GALILEOS / SIDEXIS and I need a 3D Slicer extension that streamlines research-grade periapical lesion analysis. The core priority is highly accurate detection and segmentation; everything else follows from that.
The workflow I picture starts with frictionless DICOM import, passes the volumes through a pre-trained model for initial lesion detection, lets the user refine or approve the mask, and then automatically computes total lesion volume in mm³. Once a baseline study is finished, the same tool should accept a follow-up scan, register it to the baseline, and output comparable measurements so longitudinal change can be plotted immediately.
Key deliverables
• 3D Slicer extension (Python scripted module or C++ loadable, whichever integrates best)
• Integration of a pre-trained deep learning network for periapical lesion suggestion
• Interactive editing tools leveraging Slicer’s Segment Editor for quick correction
• Automatic volume calculation and CSV export of all metrics
• Saving of clean segmentation masks (NIfTI or labelmap) for later reuse
• Side-by-side or overlaid comparison mode to track volume change between timepoints
Acceptance criteria
1. On a provided test dataset the tool produces a lesion mask within clinically acceptable boundaries that needs ≤20 % manual correction.
2. Volume output matches manual ground-truth measurements within ±5 %.
3. Baseline vs follow-up report lists both absolute volume and percentage change.
4. Extension installs through Slicer’s Extension Manager on Windows and macOS without extra compilation steps.
Time frame is ASAP, so reusable open-source libraries (MONAI, PyTorch, ITK-Snap bridges, etc.) are welcome if they shorten development while keeping accuracy high.
I will supply annotated sample CBCT volumes, ground-truth segmentations, and any additional documentation you need once the project starts.
Локация & Details
Перейти к источнику →About this listing
This remote opportunity was imported from Freelancer and is shown here for discovery. To apply, follow the link to the original posting.