TCGA-Reports: A machine-readable pathology report resource for benchmarking text-based AI models

Cancer staging is an essential clinical attribute informing patient prognosis and clinical trial eligibility. However, it is not routinely recorded in structured electronic health records. Here, we present BB-TEN: Big Bird – TNM staging Extracted from Notes, a generalizable method for the automated classification of TNM stage directly from pathology report text. We train a BERT-based model using publicly available pathology reports across approximately 7000 patients and 23 cancer types. We explore the use of different model types, with differing input sizes, parameters, and model architectures. Our final model goes beyond term-extraction, inferring TNM stage from context when it is not included in the report text explicitly. As external validation, we test our model on almost 8000 pathology reports from Columbia University Medical Center, finding that our trained model achieved an AU-ROC of 0.815–0.942. This suggests that our model can be applied broadly to other institutions without additional institution-specific fine-tuning.

show abstract

Evolving Artificial Intelligence (AI) at the Crossroads: Potentiating Productive vs. Declining Disruptive Cancer Research

Sharma,

Sarode

2024

Cancers

View full text Add to dashboard Cite

Artificial intelligence (AI), encompassing several tools and platforms such as artificial “general” intelligence (AGI) and generative artificial intelligence (GenAI), has facilitated cancer research, enhancing productivity in terms of research publications and translational value for cancer patients. AGI tools, such as ChatGPT, assist preclinical and clinical scientists in identifying tumor heterogeneity, predicting therapy outcomes, and streamlining research publications. However, this perspective review also explores the potential of AI’s influence on cancer research with regard to its impact on disruptive sciences and discoveries by preclinical and clinical scientists. The increasing reliance on AI tools may compromise biological intelligence, disrupting abstraction, creativity, and critical thinking. This could contribute to the declining trend of disruptive sciences, hindering landmark discoveries and innovations. This perspective review narrates the role of different forms of AI in the potentiation of productive cancer research and the potential disruption of disruptive sciences due to AI’s influence.

show abstract

TCGA-Reports: A machine-readable pathology report resource for benchmarking text-based AI models

Cited by 3 publications

References 31 publications

Beyond Self-consistency: Ensemble Reasoning Boosts Consistency and Accuracy of LLMs in Cancer Staging

Beyond Self-consistency: Ensemble Reasoning Boosts Consistency and Accuracy of LLMs in Cancer Staging

Generalizable and automated classification of TNM stage from pathology reports with external validation

Evolving Artificial Intelligence (AI) at the Crossroads: Potentiating Productive vs. Declining Disruptive Cancer Research

Contact Info

Product

Resources

About