Tumor Cellularity Assessment of Breast Histopathological Slides via Instance Segmentation and Pathomic Features Explainability
The segmentation and classification of cell nuclei are pivotal steps in the pipelines for the analysis of bioimages. Deep learning (DL) approaches are leading the digital pathology field in the context of nuclei detection and classification. Nevertheless, the features that are exploited by DL models...
| Main Authors: | , , , , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2023-03-01
|
| Series: | Bioengineering |
| Subjects: | |
| Online Access: | https://www.mdpi.com/2306-5354/10/4/396 |
| _version_ | 1797606373696995328 |
|---|---|
| author | Nicola Altini Emilia Puro Maria Giovanna Taccogna Francescomaria Marino Simona De Summa Concetta Saponaro Eliseo Mattioli Francesco Alfredo Zito Vitoantonio Bevilacqua |
| author_facet | Nicola Altini Emilia Puro Maria Giovanna Taccogna Francescomaria Marino Simona De Summa Concetta Saponaro Eliseo Mattioli Francesco Alfredo Zito Vitoantonio Bevilacqua |
| author_sort | Nicola Altini |
| collection | DOAJ |
| description | The segmentation and classification of cell nuclei are pivotal steps in the pipelines for the analysis of bioimages. Deep learning (DL) approaches are leading the digital pathology field in the context of nuclei detection and classification. Nevertheless, the features that are exploited by DL models to make their predictions are difficult to interpret, hindering the deployment of such methods in clinical practice. On the other hand, pathomic features can be linked to an easier description of the characteristics exploited by the classifiers for making the final predictions. Thus, in this work, we developed an explainable computer-aided diagnosis (CAD) system that can be used to support pathologists in the evaluation of tumor cellularity in breast histopathological slides. In particular, we compared an end-to-end DL approach that exploits the Mask R-CNN instance segmentation architecture with a two steps pipeline, where the features are extracted while considering the morphological and textural characteristics of the cell nuclei. Classifiers that are based on support vector machines and artificial neural networks are trained on top of these features in order to discriminate between tumor and non-tumor nuclei. Afterwards, the SHAP (Shapley additive explanations) explainable artificial intelligence technique was employed to perform a feature importance analysis, which led to an understanding of the features processed by the machine learning models for making their decisions. An expert pathologist validated the employed feature set, corroborating the clinical usability of the model. Even though the models resulting from the two-stage pipeline are slightly less accurate than those of the end-to-end approach, the interpretability of their features is clearer and may help build trust for pathologists to adopt artificial intelligence-based CAD systems in their clinical workflow. To further show the validity of the proposed approach, it has been tested on an external validation dataset, which was collected from IRCCS Istituto Tumori “Giovanni Paolo II” and made publicly available to ease research concerning the quantification of tumor cellularity. |
| first_indexed | 2024-03-11T05:14:17Z |
| format | Article |
| id | doaj.art-ad99363331c948249ed69c3ae29f1b5d |
| institution | Directory Open Access Journal |
| issn | 2306-5354 |
| language | English |
| last_indexed | 2024-03-11T05:14:17Z |
| publishDate | 2023-03-01 |
| publisher | MDPI AG |
| record_format | Article |
| series | Bioengineering |
| spelling | doaj.art-ad99363331c948249ed69c3ae29f1b5d2023-11-17T18:21:31ZengMDPI AGBioengineering2306-53542023-03-0110439610.3390/bioengineering10040396Tumor Cellularity Assessment of Breast Histopathological Slides via Instance Segmentation and Pathomic Features ExplainabilityNicola Altini0Emilia Puro1Maria Giovanna Taccogna2Francescomaria Marino3Simona De Summa4Concetta Saponaro5Eliseo Mattioli6Francesco Alfredo Zito7Vitoantonio Bevilacqua8Department of Electrical and Information Engineering (DEI), Polytechnic University of Bari, Via Edoardo Orabona n. 4, 70126 Bari, ItalyDepartment of Electrical and Information Engineering (DEI), Polytechnic University of Bari, Via Edoardo Orabona n. 4, 70126 Bari, ItalyDepartment of Electrical and Information Engineering (DEI), Polytechnic University of Bari, Via Edoardo Orabona n. 4, 70126 Bari, ItalyDepartment of Electrical and Information Engineering (DEI), Polytechnic University of Bari, Via Edoardo Orabona n. 4, 70126 Bari, ItalyMolecular Diagnostics and Pharmacogenetics Unit, IRCCS Istituto Tumori “Giovanni Paolo II”, Via O. Flacco n. 65, 70124 Bari, ItalyLaboratory of Preclinical and Translational Research, Centro di Riferimento Oncologico della Basilicata (IRCCS-CROB), Via Padre Pio n. 1, 85028 Rionero in Vulture, ItalyPathology Department, IRCCS Istituto Tumori “Giovanni Paolo II”, Via O. Flacco n. 65, 70124 Bari, ItalyPathology Department, IRCCS Istituto Tumori “Giovanni Paolo II”, Via O. Flacco n. 65, 70124 Bari, ItalyDepartment of Electrical and Information Engineering (DEI), Polytechnic University of Bari, Via Edoardo Orabona n. 4, 70126 Bari, ItalyThe segmentation and classification of cell nuclei are pivotal steps in the pipelines for the analysis of bioimages. Deep learning (DL) approaches are leading the digital pathology field in the context of nuclei detection and classification. Nevertheless, the features that are exploited by DL models to make their predictions are difficult to interpret, hindering the deployment of such methods in clinical practice. On the other hand, pathomic features can be linked to an easier description of the characteristics exploited by the classifiers for making the final predictions. Thus, in this work, we developed an explainable computer-aided diagnosis (CAD) system that can be used to support pathologists in the evaluation of tumor cellularity in breast histopathological slides. In particular, we compared an end-to-end DL approach that exploits the Mask R-CNN instance segmentation architecture with a two steps pipeline, where the features are extracted while considering the morphological and textural characteristics of the cell nuclei. Classifiers that are based on support vector machines and artificial neural networks are trained on top of these features in order to discriminate between tumor and non-tumor nuclei. Afterwards, the SHAP (Shapley additive explanations) explainable artificial intelligence technique was employed to perform a feature importance analysis, which led to an understanding of the features processed by the machine learning models for making their decisions. An expert pathologist validated the employed feature set, corroborating the clinical usability of the model. Even though the models resulting from the two-stage pipeline are slightly less accurate than those of the end-to-end approach, the interpretability of their features is clearer and may help build trust for pathologists to adopt artificial intelligence-based CAD systems in their clinical workflow. To further show the validity of the proposed approach, it has been tested on an external validation dataset, which was collected from IRCCS Istituto Tumori “Giovanni Paolo II” and made publicly available to ease research concerning the quantification of tumor cellularity.https://www.mdpi.com/2306-5354/10/4/396nuclei detectiontumor cellularityinstance segmentationexplainable artificial intelligence |
| spellingShingle | Nicola Altini Emilia Puro Maria Giovanna Taccogna Francescomaria Marino Simona De Summa Concetta Saponaro Eliseo Mattioli Francesco Alfredo Zito Vitoantonio Bevilacqua Tumor Cellularity Assessment of Breast Histopathological Slides via Instance Segmentation and Pathomic Features Explainability Bioengineering nuclei detection tumor cellularity instance segmentation explainable artificial intelligence |
| title | Tumor Cellularity Assessment of Breast Histopathological Slides via Instance Segmentation and Pathomic Features Explainability |
| title_full | Tumor Cellularity Assessment of Breast Histopathological Slides via Instance Segmentation and Pathomic Features Explainability |
| title_fullStr | Tumor Cellularity Assessment of Breast Histopathological Slides via Instance Segmentation and Pathomic Features Explainability |
| title_full_unstemmed | Tumor Cellularity Assessment of Breast Histopathological Slides via Instance Segmentation and Pathomic Features Explainability |
| title_short | Tumor Cellularity Assessment of Breast Histopathological Slides via Instance Segmentation and Pathomic Features Explainability |
| title_sort | tumor cellularity assessment of breast histopathological slides via instance segmentation and pathomic features explainability |
| topic | nuclei detection tumor cellularity instance segmentation explainable artificial intelligence |
| url | https://www.mdpi.com/2306-5354/10/4/396 |
| work_keys_str_mv | AT nicolaaltini tumorcellularityassessmentofbreasthistopathologicalslidesviainstancesegmentationandpathomicfeaturesexplainability AT emiliapuro tumorcellularityassessmentofbreasthistopathologicalslidesviainstancesegmentationandpathomicfeaturesexplainability AT mariagiovannataccogna tumorcellularityassessmentofbreasthistopathologicalslidesviainstancesegmentationandpathomicfeaturesexplainability AT francescomariamarino tumorcellularityassessmentofbreasthistopathologicalslidesviainstancesegmentationandpathomicfeaturesexplainability AT simonadesumma tumorcellularityassessmentofbreasthistopathologicalslidesviainstancesegmentationandpathomicfeaturesexplainability AT concettasaponaro tumorcellularityassessmentofbreasthistopathologicalslidesviainstancesegmentationandpathomicfeaturesexplainability AT eliseomattioli tumorcellularityassessmentofbreasthistopathologicalslidesviainstancesegmentationandpathomicfeaturesexplainability AT francescoalfredozito tumorcellularityassessmentofbreasthistopathologicalslidesviainstancesegmentationandpathomicfeaturesexplainability AT vitoantoniobevilacqua tumorcellularityassessmentofbreasthistopathologicalslidesviainstancesegmentationandpathomicfeaturesexplainability |