Differential normal and abnormal kidneys, Imaging.
| Article DOI | https://doi.org/10.1016/j.crad.2023.04.015 |
| Objective | To differentiate normal from abnormal/scarred kidneys using SPECT in pediatric patients on DMSA scan |
| AI Approach | CNN |
| Data Source(s) | Single institutional series (301 patients) |
| Model Input | 3D SPECT, 2D Maximum Intensity Projections, 2.5D Maximum Intensity Projections |
| Model Outcome | Abnormal kidneys |
| Model Metrics | AUC: 0.91 (3D), 0.94 (2.5D), 0.87-0.92 (2D) |
| Model Usability | NA |
AI = Artificial intelligence, CNN = Convolutional Neural Networks
AIPEDURO 
Leave a comment