Translating prognostic quantification of c-MYC and BCL2 from tissue microarrays to whole slide images in diffuse large B-cell lymphoma using deep learning
Abstract Background c-MYC and BCL2 positivity are important prognostic factors for diffuse large B-cell lymphoma. However, manual quantification is subject to significant intra- and inter-observer variability. We developed an automated method for quantification in whole-slide images of tissue sectio...
| Main Authors: | , , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
BMC
2024-01-01
|
| Series: | Diagnostic Pathology |
| Subjects: | |
| Online Access: | https://doi.org/10.1186/s13000-023-01425-6 |
| _version_ | 1797350138530758656 |
|---|---|
| author | Thomas E. Tavolara M. Khalid Khan Niazi Andrew L. Feldman David L. Jaye Christopher Flowers Lee A.D. Cooper Metin N. Gurcan |
| author_facet | Thomas E. Tavolara M. Khalid Khan Niazi Andrew L. Feldman David L. Jaye Christopher Flowers Lee A.D. Cooper Metin N. Gurcan |
| author_sort | Thomas E. Tavolara |
| collection | DOAJ |
| description | Abstract Background c-MYC and BCL2 positivity are important prognostic factors for diffuse large B-cell lymphoma. However, manual quantification is subject to significant intra- and inter-observer variability. We developed an automated method for quantification in whole-slide images of tissue sections where manual quantification requires evaluating large areas of tissue with possibly heterogeneous staining. We train this method using annotations of tumor positivity in smaller tissue microarray cores where expression and staining are more homogeneous and then translate this model to whole-slide images. Methods Our method applies a technique called attention-based multiple instance learning to regress the proportion of c-MYC-positive and BCL2-positive tumor cells from pathologist-scored tissue microarray cores. This technique does not require annotation of individual cell nuclei and is trained instead on core-level annotations of percent tumor positivity. We translate this model to scoring of whole-slide images by tessellating the slide into smaller core-sized tissue regions and calculating an aggregate score. Our method was trained on a public tissue microarray dataset from Stanford and applied to whole-slide images from a geographically diverse multi-center cohort produced by the Lymphoma Epidemiology of Outcomes study. Results In tissue microarrays, the automated method had Pearson correlations of 0.843 and 0.919 with pathologist scores for c-MYC and BCL2, respectively. When utilizing standard clinical thresholds, the sensitivity/specificity of our method was 0.743 / 0.963 for c-MYC and 0.938 / 0.951 for BCL2. For double-expressors, sensitivity and specificity were 0.720 and 0.974. When translated to the external WSI dataset scored by two pathologists, Pearson correlation was 0.753 & 0.883 for c-MYC and 0.749 & 0.765 for BCL2, and sensitivity/specificity was 0.857/0.991 & 0.706/0.930 for c-MYC, 0.856/0.719 & 0.855/0.690 for BCL2, and 0.890/1.00 & 0.598/0.952 for double-expressors. Survival analysis demonstrates that for progression-free survival, model-predicted TMA scores significantly stratify double-expressors and non double-expressors (p = 0.0345), whereas pathologist scores do not (p = 0.128). Conclusions We conclude that proportion of positive stains can be regressed using attention-based multiple instance learning, that these models generalize well to whole slide images, and that our models can provide non-inferior stratification of progression-free survival outcomes. |
| first_indexed | 2024-03-08T12:41:23Z |
| format | Article |
| id | doaj.art-7c68eb8f1a58429a94ccbcc88f6fa070 |
| institution | Directory Open Access Journal |
| issn | 1746-1596 |
| language | English |
| last_indexed | 2024-03-08T12:41:23Z |
| publishDate | 2024-01-01 |
| publisher | BMC |
| record_format | Article |
| series | Diagnostic Pathology |
| spelling | doaj.art-7c68eb8f1a58429a94ccbcc88f6fa0702024-01-21T12:08:19ZengBMCDiagnostic Pathology1746-15962024-01-0119111310.1186/s13000-023-01425-6Translating prognostic quantification of c-MYC and BCL2 from tissue microarrays to whole slide images in diffuse large B-cell lymphoma using deep learningThomas E. Tavolara0M. Khalid Khan Niazi1Andrew L. Feldman2David L. Jaye3Christopher Flowers4Lee A.D. Cooper5Metin N. Gurcan6Center for Artificial Intelligence Research, Wake Forest University School of MedicineCenter for Artificial Intelligence Research, Wake Forest University School of MedicineDepartment of Laboratory Medicine and Pathology, Mayo ClinicDepartment of Pathology and Laboratory Medicine, Emory University School of MedicineDepartment of Lymphoma/Myeloma, The University of Texas MD Anderson Cancer CenterDepartment of Pathology, Northwestern University Feinberg School of MedicineCenter for Artificial Intelligence Research, Wake Forest University School of MedicineAbstract Background c-MYC and BCL2 positivity are important prognostic factors for diffuse large B-cell lymphoma. However, manual quantification is subject to significant intra- and inter-observer variability. We developed an automated method for quantification in whole-slide images of tissue sections where manual quantification requires evaluating large areas of tissue with possibly heterogeneous staining. We train this method using annotations of tumor positivity in smaller tissue microarray cores where expression and staining are more homogeneous and then translate this model to whole-slide images. Methods Our method applies a technique called attention-based multiple instance learning to regress the proportion of c-MYC-positive and BCL2-positive tumor cells from pathologist-scored tissue microarray cores. This technique does not require annotation of individual cell nuclei and is trained instead on core-level annotations of percent tumor positivity. We translate this model to scoring of whole-slide images by tessellating the slide into smaller core-sized tissue regions and calculating an aggregate score. Our method was trained on a public tissue microarray dataset from Stanford and applied to whole-slide images from a geographically diverse multi-center cohort produced by the Lymphoma Epidemiology of Outcomes study. Results In tissue microarrays, the automated method had Pearson correlations of 0.843 and 0.919 with pathologist scores for c-MYC and BCL2, respectively. When utilizing standard clinical thresholds, the sensitivity/specificity of our method was 0.743 / 0.963 for c-MYC and 0.938 / 0.951 for BCL2. For double-expressors, sensitivity and specificity were 0.720 and 0.974. When translated to the external WSI dataset scored by two pathologists, Pearson correlation was 0.753 & 0.883 for c-MYC and 0.749 & 0.765 for BCL2, and sensitivity/specificity was 0.857/0.991 & 0.706/0.930 for c-MYC, 0.856/0.719 & 0.855/0.690 for BCL2, and 0.890/1.00 & 0.598/0.952 for double-expressors. Survival analysis demonstrates that for progression-free survival, model-predicted TMA scores significantly stratify double-expressors and non double-expressors (p = 0.0345), whereas pathologist scores do not (p = 0.128). Conclusions We conclude that proportion of positive stains can be regressed using attention-based multiple instance learning, that these models generalize well to whole slide images, and that our models can provide non-inferior stratification of progression-free survival outcomes.https://doi.org/10.1186/s13000-023-01425-6Deep learningDiffuse large B-cell Lymphomac-MYCBCL2ImmunohistochemistryMultiple instance learning |
| spellingShingle | Thomas E. Tavolara M. Khalid Khan Niazi Andrew L. Feldman David L. Jaye Christopher Flowers Lee A.D. Cooper Metin N. Gurcan Translating prognostic quantification of c-MYC and BCL2 from tissue microarrays to whole slide images in diffuse large B-cell lymphoma using deep learning Diagnostic Pathology Deep learning Diffuse large B-cell Lymphoma c-MYC BCL2 Immunohistochemistry Multiple instance learning |
| title | Translating prognostic quantification of c-MYC and BCL2 from tissue microarrays to whole slide images in diffuse large B-cell lymphoma using deep learning |
| title_full | Translating prognostic quantification of c-MYC and BCL2 from tissue microarrays to whole slide images in diffuse large B-cell lymphoma using deep learning |
| title_fullStr | Translating prognostic quantification of c-MYC and BCL2 from tissue microarrays to whole slide images in diffuse large B-cell lymphoma using deep learning |
| title_full_unstemmed | Translating prognostic quantification of c-MYC and BCL2 from tissue microarrays to whole slide images in diffuse large B-cell lymphoma using deep learning |
| title_short | Translating prognostic quantification of c-MYC and BCL2 from tissue microarrays to whole slide images in diffuse large B-cell lymphoma using deep learning |
| title_sort | translating prognostic quantification of c myc and bcl2 from tissue microarrays to whole slide images in diffuse large b cell lymphoma using deep learning |
| topic | Deep learning Diffuse large B-cell Lymphoma c-MYC BCL2 Immunohistochemistry Multiple instance learning |
| url | https://doi.org/10.1186/s13000-023-01425-6 |
| work_keys_str_mv | AT thomasetavolara translatingprognosticquantificationofcmycandbcl2fromtissuemicroarraystowholeslideimagesindiffuselargebcelllymphomausingdeeplearning AT mkhalidkhanniazi translatingprognosticquantificationofcmycandbcl2fromtissuemicroarraystowholeslideimagesindiffuselargebcelllymphomausingdeeplearning AT andrewlfeldman translatingprognosticquantificationofcmycandbcl2fromtissuemicroarraystowholeslideimagesindiffuselargebcelllymphomausingdeeplearning AT davidljaye translatingprognosticquantificationofcmycandbcl2fromtissuemicroarraystowholeslideimagesindiffuselargebcelllymphomausingdeeplearning AT christopherflowers translatingprognosticquantificationofcmycandbcl2fromtissuemicroarraystowholeslideimagesindiffuselargebcelllymphomausingdeeplearning AT leeadcooper translatingprognosticquantificationofcmycandbcl2fromtissuemicroarraystowholeslideimagesindiffuselargebcelllymphomausingdeeplearning AT metinngurcan translatingprognosticquantificationofcmycandbcl2fromtissuemicroarraystowholeslideimagesindiffuselargebcelllymphomausingdeeplearning |