Investigating the Benefit of Transfer Learning from the Locus Coeruleus for Substantia Nigra Pars Compacta Segmentation
Organizations: Faculty of Electrical Engineering and Information Technology, Institute for Medical Engineering, OVGU · Department of Neurology, Faculty of Medicine, Otto von Guericke University (OVGU) · Faculty of Computer Science, OVGU · German Center for Neurodegenerative Diseases (DZNE), Magdeburg, Germany ·
Abstract
The substantia nigra pars compacta (SNpc) is a small midbrain nucleus that degenerates early in Parkinson's disease (PD). T1-weighted FLASH MRI enables its visualization and may support early detection of neurodegeneration. However, manual segmentation is difficult and time-consuming, and deep learning models often suffer from the limited availability of labeled data. Transfer learning (TL) may help by adapting pre-trained models to small datasets. We investigate whether TL from a model trained on the locus coeruleus (LC), a nucleus with a similar appearance in this sequence, improves SNpc segmentation. Several TL strategies were evaluated using a 3D U-Net. With as few as 5-14 labeled subjects, TL and training from scratch achieved similar performance, reaching a Dice similarity coefficient of up to 0.77, close to reported inter-rater agreement. No TL strategy differed significantly from training from scratch, suggesting that initialization from the LC offers no significant benefit for SNpc segmentation.
Figures & tables
| Strategy | 14 subjects | 7 subjects | 5 subjects |
|---|---|---|---|
| Scratch | 0.78 0.08 | 0.77 0.10 | 0.73 0.15 |
| Decay last | 0.77 0.10 | 0.74 0.14 | 0.69 0.19 |
| Decay up2 | 0.76 0.12 | 0.66 0.20 | 0.58 0.22 |
| Decay up1 | 0.65 0.21 | 0.66 0.18 | 0.65 0.21 |
| Decay up0 | 0.68 0.17 | 0.72 0.16 | 0.65 0.20 |
| Decay down3 | 0.71 0.15 | 0.69 0.17 | 0.56 0.31 |
Medical Image Segmentation
Medical image segmentation delineates organs, lesions, and other structures in scans to support measurement, diagnosis, treatment planning, and research.
See all 846 papers