Title:Universal Lymph Node Detection in T2 MRI using Neural Networks

Abstract:Purpose: Identification of abdominal Lymph Nodes (LN) that are suspicious for metastasis in T2 Magnetic Resonance Imaging (MRI) scans is critical for staging of lymphoproliferative diseases. Prior work on LN detection has been limited to specific anatomical regions of the body (pelvis, rectum) in single MR slices. Therefore, the development of a universal approach to detect LN in full T2 MRI volumes is highly desirable.
Methods: In this study, a Computer Aided Detection (CADきゃど) pipeline to universally identify abdominal LN in volumetric T2 MRI using neural networks is proposed. First, we trained various neural network models for detecting LN: Faster RCNN with and without Hard Negative Example Mining (HNEM), FCOS, FoveaBox, VFNet, and Detection Transformer (DETR). Next, we show that the state-of-the-art (SOTA) VFNet model with Adaptive Training Sample Selection (ATSS) outperforms Faster RCNN with HNEM. Finally, we ensembled models that surpassed a 45% mAP threshold. We found that the VFNet model and one-stage model ensemble can be interchangeably used in the CADきゃど pipeline.
Results: Experiments on 122 test T2 MRI volumes revealed that VFNet achieved a 51.1% mAP and 78.7% recall at 4 false positives (FP) per volume, while the one-stage model ensemble achieved a mAP of 52.3% and sensitivity of 78.7% at 4FP.
Conclusion: Our contribution is a CADきゃど pipeline that detects LN in T2 MRI volumes, resulting in a sensitivity improvement of $\sim$14 points over the current SOTA method for LN detection (sensitivity of 78.7% at 4 FP vs. 64.6% at 5 FP per volume).