RetMIL: Retentive Multiple Instance Learning for Histopathological Whole Slide Image Classification
Fuente:
arXiv
Gespeichert in:
| Hauptverfasser: | , , , , , , , |
|---|---|
| Format: | Preprint |
| Veröffentlicht: |
2024
|
| Schlagworte: | |
| Online-Zugang: | |
| Tags: |
Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
|
| _version_ | 1866929279382585344 |
|---|---|
| author | Chu, Hongbo Sun, Qiehe Li, Jiawen Chen, Yuxuan Zhang, Lizhong Guan, Tian Han, Anjia He, Yonghong |
| author_facet | Chu, Hongbo Sun, Qiehe Li, Jiawen Chen, Yuxuan Zhang, Lizhong Guan, Tian Han, Anjia He, Yonghong |
| contents | Histopathological whole slide image (WSI) analysis with deep learning has become a research focus in computational pathology. The current paradigm is mainly based on multiple instance learning (MIL), in which approaches with Transformer as the backbone are well discussed. These methods convert WSI tasks into sequence tasks by representing patches as tokens in the WSI sequence. However, the feature complexity brought by high heterogeneity and the ultra-long sequences brought by gigapixel size makes Transformer-based MIL suffer from the challenges of high memory consumption, slow inference speed, and lack of performance. To this end, we propose a retentive MIL method called RetMIL, which processes WSI sequences through hierarchical feature propagation structure. At the local level, the WSI sequence is divided into multiple subsequences. Tokens of each subsequence are updated through a parallel linear retention mechanism and aggregated utilizing an attention layer. At the global level, subsequences are fused into a global sequence, then updated through a serial retention mechanism, and finally the slide-level representation is obtained through a global attention pooling. We conduct experiments on two public CAMELYON and BRACS datasets and an public-internal LUNG dataset, confirming that RetMIL not only achieves state-of-the-art performance but also significantly reduces computational overhead. Our code will be accessed shortly. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2403_10858 |
| institution | arXiv |
| publishDate | 2024 |
| record_format | arxiv |
| spellingShingle | RetMIL: Retentive Multiple Instance Learning for Histopathological Whole Slide Image Classification Chu, Hongbo Sun, Qiehe Li, Jiawen Chen, Yuxuan Zhang, Lizhong Guan, Tian Han, Anjia He, Yonghong Computer Vision and Pattern Recognition Histopathological whole slide image (WSI) analysis with deep learning has become a research focus in computational pathology. The current paradigm is mainly based on multiple instance learning (MIL), in which approaches with Transformer as the backbone are well discussed. These methods convert WSI tasks into sequence tasks by representing patches as tokens in the WSI sequence. However, the feature complexity brought by high heterogeneity and the ultra-long sequences brought by gigapixel size makes Transformer-based MIL suffer from the challenges of high memory consumption, slow inference speed, and lack of performance. To this end, we propose a retentive MIL method called RetMIL, which processes WSI sequences through hierarchical feature propagation structure. At the local level, the WSI sequence is divided into multiple subsequences. Tokens of each subsequence are updated through a parallel linear retention mechanism and aggregated utilizing an attention layer. At the global level, subsequences are fused into a global sequence, then updated through a serial retention mechanism, and finally the slide-level representation is obtained through a global attention pooling. We conduct experiments on two public CAMELYON and BRACS datasets and an public-internal LUNG dataset, confirming that RetMIL not only achieves state-of-the-art performance but also significantly reduces computational overhead. Our code will be accessed shortly. |
| title | RetMIL: Retentive Multiple Instance Learning for Histopathological Whole Slide Image Classification |
| topic | Computer Vision and Pattern Recognition |
| url | https://arxiv.org/abs/2403.10858 |