Explainable Artificial Intelligence Reveals Novel Insight into Tumor Microenvironment Conditions Linked with Better Prognosis in Patients with Breast Cancer

Chakraborty, Debaditya; Ivan, Cristina; Amero, Paola; Khan, Maliha; Rodríguez-Aguayo, Cristian; Başağaoğlu, Hakan; López-Berestein, Gabriel

doi:10.3390/cancers13143450

Cited by 33 publications

(18 citation statements)

References 67 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…It is becoming imperative to apply AI models that are inherently interpretable, together with XAI methods to tally accurate predictions with a better understanding of the underlying reasoning of the AI approaches. The need for using XAI techniques to explain individual decisions of predictive models has only recently emerged [ 37 ]. The use of XAI in oncology is still in its infancy, and is not widely appreciated.…”

Section: Discussionmentioning

confidence: 99%

Factors Predicting Surgical Effort Using Explainable Artificial Intelligence in Advanced Stage Epithelial Ovarian Cancer

Laios

Kalampokis

Johnson

et al. 2022

Cancers

View full text Add to dashboard Cite

(1) Background: Surgical cytoreduction for epithelial ovarian cancer (EOC) is a complex procedure. Encompassed within the performance skills to achieve surgical precision, intra-operative surgical decision-making remains a core feature. The use of eXplainable Artificial Intelligence (XAI) could potentially interpret the influence of human factors on the surgical effort for the cytoreductive outcome in question; (2) Methods: The retrospective cohort study evaluated 560 consecutive EOC patients who underwent cytoreductive surgery between January 2014 and December 2019 in a single public institution. The eXtreme Gradient Boosting (XGBoost) and Deep Neural Network (DNN) algorithms were employed to develop the predictive model, including patient- and operation-specific features, and novel features reflecting human factors in surgical heuristics. The precision, recall, F1 score, and area under curve (AUC) were compared between both training algorithms. The SHapley Additive exPlanations (SHAP) framework was used to provide global and local explainability for the predictive model; (3) Results: A surgical complexity score (SCS) cut-off value of five was calculated using a Receiver Operator Characteristic (ROC) curve, above which the probability of incomplete cytoreduction was more likely (area under the curve [AUC] = 0.644; 95% confidence interval [CI] = 0.598–0.69; sensitivity and specificity 34.1%, 86.5%, respectively; p = 0.000). The XGBoost outperformed the DNN assessment for the prediction of the above threshold surgical effort outcome (AUC = 0.77; 95% [CI] 0.69–0.85; p < 0.05 vs. AUC 0.739; 95% [CI] 0.655–0.823; p < 0.95). We identified “turning points” that demonstrated a clear preference towards above the given cut-off level of surgical effort; in consultant surgeons with <12 years of experience, age <53 years old, who, when attempting primary cytoreductive surgery, recorded the presence of ascites, an Intraoperative Mapping of Ovarian Cancer score >4, and a Peritoneal Carcinomatosis Index >7, in a surgical environment with the optimization of infrastructural support. (4) Conclusions: Using XAI, we explain how intra-operative decisions may consider human factors during EOC cytoreduction alongside factual knowledge, to maximize the magnitude of the selected trade-off in effort. XAI techniques are critical for a better understanding of Artificial Intelligence frameworks, and to enhance their incorporation in medical applications.

show abstract

Section: Discussionmentioning

confidence: 99%

Factors Predicting Surgical Effort Using Explainable Artificial Intelligence in Advanced Stage Epithelial Ovarian Cancer

Laios

Kalampokis

Johnson

et al. 2022

Cancers

View full text Add to dashboard Cite

show abstract

“…They identified survival-predicting genes in a sizeable pan-cancer cohort ( 9 ). The data-driven relationship between immune cell composition in the tumor microenvironment (TME) and >=5-year survival in breast cancer patients was investigated using an interpretable artificial intelligence (XAI) model to accurately predict clinical outcomes, thereby designing innovative strategies to cure cancer ( 11 ).…”

Section: Discussionmentioning

confidence: 99%

“…The tumor microenvironment (TME) plays an important role in tumorigenesis and progression. Novel techniques in artificial intelligence (AI) can help determine areas of therapeutic need, enhance clinical trial interpretation, identify novel targets, and generate accurate predictions that are impossible with traditional statistical techniques ( 11 ).AI is used to assess TME, prognosis and the benefit of adjuvant chemotherapy in patients after radical gastric cancer surgery and is a valuable addition to the current TNM staging system ( 12 ).…”

Section: Introductionmentioning

confidence: 99%

Mapping intellectual structures and research hotspots in the application of artificial intelligence in cancer: A bibliometric analysis

et al. 2022

View full text Add to dashboard Cite

BackgroundArtificial intelligence (AI) is more and more widely used in cancer, which is of great help to doctors in diagnosis and treatment. This study aims to summarize the current research hotspots in the Application of Artificial Intelligence in Cancer (AAIC) and to assess the research trends in AAIC.MethodsScientific publications for AAIC-related research from 1 January 1998 to 1 July 2022 were obtained from the Web of Science database. The metrics analyses using bibliometrics software included publication, keyword, author, journal, institution, and country. In addition, the blustering analysis on the binary matrix was performed on hot keywords.ResultsThe total number of papers in this study is 1592. The last decade of AAIC research has been divided into a slow development phase (2013-2018) and a rapid development phase (2019-2022). An international collaboration centered in the USA is dedicated to the development and application of AAIC. Li J is the most prolific writer in AAIC. Through clustering analysis and high-frequency keyword research, it has been shown that AI plays a significantly important role in the prediction, diagnosis, treatment and prognosis of cancer. Classification, diagnosis, carcinogenesis, risk, and validation are developing topics. Eight hotspot fields of AAIC were also identified.ConclusionAAIC can benefit cancer patients in diagnosing cancer, assessing the effectiveness of treatment, making a decision, predicting prognosis and saving costs. Future AAIC research may be dedicated to optimizing AI calculation tools, improving accuracy, and promoting AI.

show abstract

“…In addition, we further found that ASB16-AS1 expression was associated to different levels of immune infiltration in cancer types through a variety of immune cell content assessment methods. Immune cell content evaluation has differences, and multiple evaluation methods are mutually verified to effectively judge the relationship between ASB16-AS1 and immune cell subtypes [ 34 , 35 ]. Many immune cell subtypes are believed to play an important regulatory role in the development and treatment of cancers.…”

Section: Discussionmentioning

confidence: 99%

Expression, prognosis value, and immune infiltration of lncRNA ASB16-AS1 identified by pan-cancer analysis

Liao

Wang

et al. 2021

Bioengineered

View full text Add to dashboard Cite

Long non-coding RNA known as ASB16 antisense RNA1 (ASB16-AS1) has been proven to be an oncogene, and the relationship between ASB16-AS1 and immunity is still under studied. This study aims to explore the expression and prognostic potential of ASB16-AS1, and to visualize the relationship between ASB16-AS1 expression and immune infiltration in pan-cancer analysis. We clarified ASB16-AS1 expression patterns and its relationship with prognosis through multi-platform and multi-database sources. We also verified the function of ASB16-AS1 in liver hepatocellular carcinoma (LIHC). A variety of immune cell content evaluation methods were used to mutually verify the correlation between ASB16-AS1 and immune infiltration. Finally, the relationships between ASB16-AS1 and molecular characteristics were further explored. In terms of comprehensive analysis, compared with non-tumor tissues, ASB16-AS1 was highly expressed in tumor tissues, and indicated the value of poor prognosis in multiple cancer types. Functional assays, such as counting kit-8 assay, transwell assay and scratch-wound assay verified that high ASB16-AS1 expression promoted tumor progression in LIHC. ASB16-AS1 was positively correlated with B cells, T cells CD4+ and T cells CD8+ in most cancer types, and negatively correlated with macrophages, dendritic cells and neutrophils in some cancer types. In addition, there were different interaction modes between ASB16-AS1 and molecular features, such as the relationship with oncogenic signaling pathways, showing that the high ASB16-AS1 expression was related to alterations in oncogenic signaling pathways. Our study emphasizes that ASB16-AS1 is a potential pan-cancer prognostic marker, whichs is associated with the immune infiltration in multiple cancer types.

show abstract

Explainable Artificial Intelligence Reveals Novel Insight into Tumor Microenvironment Conditions Linked with Better Prognosis in Patients with Breast Cancer

Cited by 33 publications

References 67 publications

Factors Predicting Surgical Effort Using Explainable Artificial Intelligence in Advanced Stage Epithelial Ovarian Cancer

Factors Predicting Surgical Effort Using Explainable Artificial Intelligence in Advanced Stage Epithelial Ovarian Cancer

Mapping intellectual structures and research hotspots in the application of artificial intelligence in cancer: A bibliometric analysis

Expression, prognosis value, and immune infiltration of lncRNA ASB16-AS1 identified by pan-cancer analysis

Contact Info

Product

Resources

About