Key result
The LS-SOM neural network achieved a pixel precision of 0.811 for plantar pressure image segmentation, outperforming single segNet segmentation (0.783) and other traditional methods.
Why the study?
The study was conducted to alleviate the over-segmentation problem in plantar pressure imaging data for the customization of shoe lasts.
Does a deep self-organizing map neural network based on level-set improve the segmentation of plantar pressure imaging data compared to existing methods?
Does a deep self-organizing map neural network based on level-set improve the segmentation of plantar pressure imaging data compared to existing methods?
Absolute Event Rate: 0.811% vs 0.783%
A novel deep self-organizing map neural network improves the segmentation of plantar pressure images, which may aid in designing optimal shoe lasts.
May enhance plantar pressure image analysis for orthotics; leaves open clinical validation before practice change.
This study introduces a deep self-organizing map neural network based on level-set (LS-SOM) for the customization of a shoe-last defined from plantar pressure imaging data. To alleviate the over-segmentation problem of images, which refers to segmenting images into more subcomponents, a domain-based segmentation model of plantar pressure images was constructed. The domain growth algorithm was subsequently modified by optimizing its parameters. A SOM with 10, 15, 20, and 30 hidden layers was compared and validated according to domain growth characteristics by using merging and splitting algorithms. Furthermore, we incorporated a level set segmentation method into the plantar pressure image algorithm to enhance its efficiency. Compared to the literature, this proposed method has significantly improved pixel accuracy, average cross-combination ratio, frequency-weighted cross-combination ratio, and boundary F1 index comparison. Using the proposed methods, shoe lasts can be designed optimally, and wearing comfort is enhanced, particularly for people with high blood pressure.
No takes yet. Share an insight, caveat, or question.
Wang et al. (2024) studied Plantar pressure imaging (n=2,000). LS-SOM neural network vs. Other segmentation algorithms (e.g., Single segNet) was evaluated on Pixel Precision (PA). The LS-SOM neural network achieved a pixel precision of 0.811 for plantar pressure image segmentation, outperforming single segNet segmentation (0.783) and other traditional methods.
Synapse has enriched 5 closely related papers on similar clinical questions. Consider them for comparative context: