Deep active learning for classifying cancer pathology reports.


De Angeli K(1)(2), Gao S(3), Alawad M(1), Yoon HJ(1), Schaefferkoetter N(1), Wu XC(4), Durbin EB(5), Doherty J(6), Stroup A(7), Coyle L(8), Penberthy L(9), Tourassi G(1).
Author information:
(1)Oak Ridge National Lab, Oak Ridge, TN, USA.
(2)The Bredesen Center, The University of Tennessee, Knoxville, TN, US.
(3)Oak Ridge National Lab, Oak Ridge, TN, USA. [Email]
(4)Louisiana Tumor Registry, Louisiana State University Health Sciences Center, School of Public Health, New Orleans, LA, USA.
(5)College of Medicine, University of Kentucky, Lexington, KY, USA.
(6)Utah Cancer Registry, University of Utah School of Medicine, Salt Lake City, UT, USA.
(7)New Jersey State Cancer Registry, New Jersey Department of Health, Trenton, NJ, USA.
(8)Information Management Services Inc., Calverton, MD, USA.
(9)Surveillance Research Program, Division of Cancer Control and Population Sciences, National Cancer Institute, Bethesda, MD, USA.


BACKGROUND: Automated text classification has many important applications in the clinical setting; however, obtaining labelled data for training machine learning and deep learning models is often difficult and expensive. Active learning techniques may mitigate this challenge by reducing the amount of labelled data required to effectively train a model. In this study, we analyze the effectiveness of 11 active learning algorithms on classifying subsite and histology from cancer pathology reports using a Convolutional Neural Network as the text classification model. RESULTS: We compare the performance of each active learning strategy using two differently sized datasets and two different classification tasks. Our results show that on all tasks and dataset sizes, all active learning strategies except diversity-sampling strategies outperformed random sampling, i.e., no active learning. On our large dataset (15K initial labelled samples, adding 15K additional labelled samples each iteration of active learning), there was no clear winner between the different active learning strategies. On our small dataset (1K initial labelled samples, adding 1K additional labelled samples each iteration of active learning), marginal and ratio uncertainty sampling performed better than all other active learning techniques. We found that compared to random sampling, active learning strongly helps performance on rare classes by focusing on underrepresented classes. CONCLUSIONS: Active learning can save annotation cost by helping human annotators efficiently and intelligently select which samples to label. Our results show that a dataset constructed using effective active learning techniques requires less than half the amount of labelled data to achieve the same performance as a dataset constructed using random sampling.