Transactions of the Association for Computational Linguistics, 2020
Transformer-based models have pushed state of the art in many areas of NLP, but our understanding... more Transformer-based models have pushed state of the art in many areas of NLP, but our understanding of what is behind their success is still limited. This paper is the first survey of over 150 studies of the popular BERT model. We review the current state of knowledge about how BERT works, what kind of information it learns and how it is represented, common modifications to its training objectives and architecture, the overparameterization issue, and approaches to compression. We then outline directions for future research.
Introduction: Breast Cancer (BC) is the most common type of cancer among women worldwide and a le... more Introduction: Breast Cancer (BC) is the most common type of cancer among women worldwide and a leading cause of cancer related deaths. Based on insight into immune system/cancer interactions, cancer immunotherapy strategies have been developed. Of these strategies, immune checkpoints regulators are among the most important. Programmed Death Ligand 1 (PD-L1) is one of the 2 ligands of Programmed Death-1 (PD-1); an important immune checkpoint regulator. Clinical trials using PD-L1 antibodies have shown efficacy in various cancers. Some of these antibodies have received Food and Drug Administration (FDA) approval for clinical use. Aim: Detection of Immunohistochemical (IHC) expression of PD-L1 by Tumour Cells (TC) and stromal Tumour Infiltrating Lymphocytes (TILs) in female BC, as well as, correlation of this expression with the pathologic parameters of the tumours. Materials and Methods: One hundred BC tissue sections were collected from mastectomy specimens. IHC expression of PD-L1 was investigated in TC and TILs. Correlation of PD-L1 expression with the tumours pathologic parameters was performed using Statistical Package of Social Science (SPSS) Software program, version 25.0. Results: TC PD-L1 expression was detected in 61% of cases and showed statistically significant correlation with higher tumour grade, higher prognostic stage, Estrogen receptor negativity, Human Epidermal Growth Factor Receptor Type 2 (HER2) enriched molecular subtype and LVI (p-value <0.05). TILs PD-L1 expression was detected in 55% of cases and showed statistically significant correlation with higher tumour grade, higher prognostic stage, Estrogen and Progesterone receptor negativity, HER2 overexpression, higher Ki-67 index, triple negative molecular subtype and higher stromal TILs (p-value <0.05). Conclusion: The present results supported PD-L1 expression in BC by both TC and TILs, with higher expression in subset of tumours that are high grade, TILs rich and lacking hormone receptors, highlighting them as candidates for anti-PD-1/PD-L1 therapy.
Transactions of the Association for Computational Linguistics, 2020
Transformer-based models have pushed state of the art in many areas of NLP, but our understanding... more Transformer-based models have pushed state of the art in many areas of NLP, but our understanding of what is behind their success is still limited. This paper is the first survey of over 150 studies of the popular BERT model. We review the current state of knowledge about how BERT works, what kind of information it learns and how it is represented, common modifications to its training objectives and architecture, the overparameterization issue, and approaches to compression. We then outline directions for future research.
Introduction: Breast Cancer (BC) is the most common type of cancer among women worldwide and a le... more Introduction: Breast Cancer (BC) is the most common type of cancer among women worldwide and a leading cause of cancer related deaths. Based on insight into immune system/cancer interactions, cancer immunotherapy strategies have been developed. Of these strategies, immune checkpoints regulators are among the most important. Programmed Death Ligand 1 (PD-L1) is one of the 2 ligands of Programmed Death-1 (PD-1); an important immune checkpoint regulator. Clinical trials using PD-L1 antibodies have shown efficacy in various cancers. Some of these antibodies have received Food and Drug Administration (FDA) approval for clinical use. Aim: Detection of Immunohistochemical (IHC) expression of PD-L1 by Tumour Cells (TC) and stromal Tumour Infiltrating Lymphocytes (TILs) in female BC, as well as, correlation of this expression with the pathologic parameters of the tumours. Materials and Methods: One hundred BC tissue sections were collected from mastectomy specimens. IHC expression of PD-L1 was investigated in TC and TILs. Correlation of PD-L1 expression with the tumours pathologic parameters was performed using Statistical Package of Social Science (SPSS) Software program, version 25.0. Results: TC PD-L1 expression was detected in 61% of cases and showed statistically significant correlation with higher tumour grade, higher prognostic stage, Estrogen receptor negativity, Human Epidermal Growth Factor Receptor Type 2 (HER2) enriched molecular subtype and LVI (p-value <0.05). TILs PD-L1 expression was detected in 55% of cases and showed statistically significant correlation with higher tumour grade, higher prognostic stage, Estrogen and Progesterone receptor negativity, HER2 overexpression, higher Ki-67 index, triple negative molecular subtype and higher stromal TILs (p-value <0.05). Conclusion: The present results supported PD-L1 expression in BC by both TC and TILs, with higher expression in subset of tumours that are high grade, TILs rich and lacking hormone receptors, highlighting them as candidates for anti-PD-1/PD-L1 therapy.
Uploads
Papers by Ольга Ковалева