Not seeing the trees for the forest. The impact of neighbours on graph-based configurations in histopathology
Abstract Background Deep learning (DL) has set new standards in cancer diagnosis, significantly enhancing the accuracy of automated classification of whole slide images (WSIs) derived from biopsied tissue samples. To enable DL models to process these large images, WSIs are typically divided into tho...
Saved in:
| Main Authors: | , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
BMC
2025-01-01
|
| Series: | BMC Bioinformatics |
| Subjects: | |
| Online Access: | https://doi.org/10.1186/s12859-024-06007-x |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1841544225737408512 |
|---|---|
| author | Olga Fourkioti Matt De Vries Reed Naidoo Chris Bakal |
| author_facet | Olga Fourkioti Matt De Vries Reed Naidoo Chris Bakal |
| author_sort | Olga Fourkioti |
| collection | DOAJ |
| description | Abstract Background Deep learning (DL) has set new standards in cancer diagnosis, significantly enhancing the accuracy of automated classification of whole slide images (WSIs) derived from biopsied tissue samples. To enable DL models to process these large images, WSIs are typically divided into thousands of smaller tiles, each containing 10–50 cells. Multiple Instance Learning (MIL) is a commonly used approach, where WSIs are treated as bags comprising numerous tiles (instances) and only bag-level labels are provided during training. The model learns from these broad labels to extract more detailed, instance-level insights. However, biopsied sections often exhibit high intra- and inter-phenotypic heterogeneity, presenting a significant challenge for classification. To address this, many graph-based methods have been proposed, where each WSI is represented as a graph with tiles as nodes and edges defined by specific spatial relationships. Results In this study, we investigate how different graph configurations, varying in connectivity and neighborhood structure, affect the performance of MIL models. We developed a novel pipeline, K-MIL, to evaluate the impact of contextual information on cell classification performance. By incorporating neighboring tiles into the analysis, we examined whether contextual information improves or impairs the network’s ability to identify patterns and features critical for accurate classification. Our experiments were conducted on two datasets: COLON cancer and UCSB datasets. Conclusions Our results indicate that while incorporating more spatial context information generally improves model accuracy at both the bag and tile levels, the improvement at the tile level is not linear. In some instances, increasing spatial context leads to misclassification, suggesting that more context is not always beneficial. This finding highlights the need for careful consideration when incorporating spatial context information in digital pathology classification tasks. |
| format | Article |
| id | doaj-art-0aa40e59849745e59388c2fd25256673 |
| institution | Kabale University |
| issn | 1471-2105 |
| language | English |
| publishDate | 2025-01-01 |
| publisher | BMC |
| record_format | Article |
| series | BMC Bioinformatics |
| spelling | doaj-art-0aa40e59849745e59388c2fd252566732025-01-12T12:41:53ZengBMCBMC Bioinformatics1471-21052025-01-0126111810.1186/s12859-024-06007-xNot seeing the trees for the forest. The impact of neighbours on graph-based configurations in histopathologyOlga Fourkioti0Matt De Vries1Reed Naidoo2Chris Bakal3The Institute of Cancer ResearchThe Institute of Cancer ResearchThe Institute of Cancer ResearchThe Institute of Cancer ResearchAbstract Background Deep learning (DL) has set new standards in cancer diagnosis, significantly enhancing the accuracy of automated classification of whole slide images (WSIs) derived from biopsied tissue samples. To enable DL models to process these large images, WSIs are typically divided into thousands of smaller tiles, each containing 10–50 cells. Multiple Instance Learning (MIL) is a commonly used approach, where WSIs are treated as bags comprising numerous tiles (instances) and only bag-level labels are provided during training. The model learns from these broad labels to extract more detailed, instance-level insights. However, biopsied sections often exhibit high intra- and inter-phenotypic heterogeneity, presenting a significant challenge for classification. To address this, many graph-based methods have been proposed, where each WSI is represented as a graph with tiles as nodes and edges defined by specific spatial relationships. Results In this study, we investigate how different graph configurations, varying in connectivity and neighborhood structure, affect the performance of MIL models. We developed a novel pipeline, K-MIL, to evaluate the impact of contextual information on cell classification performance. By incorporating neighboring tiles into the analysis, we examined whether contextual information improves or impairs the network’s ability to identify patterns and features critical for accurate classification. Our experiments were conducted on two datasets: COLON cancer and UCSB datasets. Conclusions Our results indicate that while incorporating more spatial context information generally improves model accuracy at both the bag and tile levels, the improvement at the tile level is not linear. In some instances, increasing spatial context leads to misclassification, suggesting that more context is not always beneficial. This finding highlights the need for careful consideration when incorporating spatial context information in digital pathology classification tasks.https://doi.org/10.1186/s12859-024-06007-xComputational pathologyGraph-neural networksVisualizationAttentionContextCell classification |
| spellingShingle | Olga Fourkioti Matt De Vries Reed Naidoo Chris Bakal Not seeing the trees for the forest. The impact of neighbours on graph-based configurations in histopathology BMC Bioinformatics Computational pathology Graph-neural networks Visualization Attention Context Cell classification |
| title | Not seeing the trees for the forest. The impact of neighbours on graph-based configurations in histopathology |
| title_full | Not seeing the trees for the forest. The impact of neighbours on graph-based configurations in histopathology |
| title_fullStr | Not seeing the trees for the forest. The impact of neighbours on graph-based configurations in histopathology |
| title_full_unstemmed | Not seeing the trees for the forest. The impact of neighbours on graph-based configurations in histopathology |
| title_short | Not seeing the trees for the forest. The impact of neighbours on graph-based configurations in histopathology |
| title_sort | not seeing the trees for the forest the impact of neighbours on graph based configurations in histopathology |
| topic | Computational pathology Graph-neural networks Visualization Attention Context Cell classification |
| url | https://doi.org/10.1186/s12859-024-06007-x |
| work_keys_str_mv | AT olgafourkioti notseeingthetreesfortheforesttheimpactofneighboursongraphbasedconfigurationsinhistopathology AT mattdevries notseeingthetreesfortheforesttheimpactofneighboursongraphbasedconfigurationsinhistopathology AT reednaidoo notseeingthetreesfortheforesttheimpactofneighboursongraphbasedconfigurationsinhistopathology AT chrisbakal notseeingthetreesfortheforesttheimpactofneighboursongraphbasedconfigurationsinhistopathology |