Optimal vocabulary selection approaches for privacy-preserving deep NLP model training for information extraction and cancer epidemiology

Abstract

BACKGROUND: With the use of artificial intelligence and machine learning techniques for biomedical informatics, security and privacy concerns over the data and subject identities have also become an important issue and essential research topic. Without intentional safeguards, machine learning models may find patterns and features to improve task performance that are associated with private personal information. OBJECTIVE: The privacy vulnerability of deep learning models for information extraction from medical textural contents needs to be quantified since the models are exposed to private health information and personally identifiable information. The objective of the study is to quantify the privacy vulnerability of the deep learning models for natural language processing and explore a proper way of securing patients’ information to mitigate confidentiality breaches. METHODS: The target model is the multitask convolutional neural network for information extraction from cancer pathology reports, where the data for training the model are from multiple state population-based cancer registries. This study proposes the following schemes to collect vocabularies from the cancer pathology reports; (a) words appearing in multiple registries, and (b) words that have higher mutual information. We performed membership inference attacks on the models in high-performance computing environments. RESULTS: The comparison outcomes suggest that the proposed vocabulary selection methods resulted in lower privacy vulnerability while maintaining the same level of clinical task performance.

Precision information extraction for rare disease epidemiology at scale

“…Automated approaches to clinical epidemiology include “Data extraction for epidemiological research” or DExtER [ 35 ]. NLP approaches to information extraction include analyzing social media to analyze drug abuse epidemiology[ 36 ] and detecting cancer cases to calculate epidemiologic prevalence [ 37 , 38 ].…”

Cancer registries and data protection in the age of health digital interoperability in Europe: The perspective of the Italian Network of Cancer Registries (AIRTUM)

“…Also, the sharing of good practices for data linkage within EU member states has been proposed ( 41 ). Although information technology has played a positive role in quality improvement and facility of cancer registries, more strict restriction strategies, such as identifying authentication levels, controlling and coding data approaches, tailored de-identification methods, and other technical measures, had to be developed to secure the patients’ privacy ( 22 , 42 , 43 ).…”

Artificial intelligence assists precision medicine in cancer treatment

“…In addition, data from patients are governed by privacy laws ( 194 ). The lack of supervision of these data may lead to breaching patient privacy rules; therefore, appropriate intervention and the improvement of laws and regulations are required.…”

“…Certainly, studies have focused on solving the privacy problem with regards to patient data. Under the same performance, the privacy vulnerability is reduced by vocabulary selection means ( 194 ). At the same time, fusing these data should comply with the principles of medical ethics.…”