Case studies

Building a diverse CT dataset for AI-based metal artifact reduction

Radiology
Metal Artifact Reduction
AI Development
AI Validation
Multi-Institutional Data
Imaging Data
Clinical Data

Background

Metal implants can generate artifacts on CT scans, obscuring surrounding anatomy and degrading diagnostic quality1.

This challenge spans multiple implant categories, including orthopedic, dental, cardiovascular, spinal, and neurovascular devices, requiring effective metal artifact reduction techniques for reliable image interpretation.

The development of robust AI-based metal artifact reduction (MAR) algorithms requires well-characterized CT datasets that reflect real-world diversity in implant types, scanners, and acquisition protocols while meeting strict technical specifications.

Objective

A top-five global medtech company engaged Aigora to curate a tailored CT dataset supporting the development and validation of an AI-based MAR algorithm.

Delivery

Aigora delivered the dataset through a multi-stage curation process that identified eligible CT studies across its network, validated imaging and reconstruction parameters, manually classified implant types, and prepared the dataset for downstream AI algorithm development.

Dataset characteristics

  • Medical imaging

    Thin-slice CT images across multiple implant categories, including scanner manufacturer and model, reconstruction kernels, and acquisition parameters

  • Clinical data

    Patient demographics, implant category and subtype

References

  1. Sharma S, Kaushal A, Patel S, Kumar V, Prakash M, Mandeep D. Methods to address metal artifacts in post-processed CT images - A do-it-yourself guide for orthopedic surgeons. J Clin Orthop Trauma. 2021;20:101493. doi:10.1016/j.jcot.2021.101493

‍

Project metrics

  • Number of cases

    350

  • Number of implant categories

    8

  • Delivery timeline

    10 weeks

More case studies

See all

Advancing microsatellite instability testing with gastric cancer H&E-stained whole slide image data

Oncology
Biomarker Pre-Screening
Pathology Data
Molecular Test Results
Read case study

Enabling AI-driven research in retinal diseases with longitudinal ophthalmic imaging data

Ophthalmology
Biomarker Discovery
Treatment Response Assessment
Multimodal Imaging Data
Clinical Data
Read case study

Identifying novel tissue biomarkers in inflammatory and autoimmune diseases with curated whole slide image data and biospecimens

Inflammatory Bowel Disease
Scleroderma
Biomarker Discovery
Translational Research
Clinical Data
Read case study