The role of artificial intelligence is growing in healthcare and disease prediction. Because of its potential impact and demographic disparities that have been identified in machine learning models for disease prediction, there are growing concerns about transparency, accountability and fairness of these predictive models. However, very little research has investigated methods for improving model fairness in disease prediction, particularly when the sensitive attribute is multivariate and when the distribution of sensitive attribute groups is highly skewed. In this work, we explore algorithmic fairness when predicting heart disease and Alzheimer's Disease and Related Dementias (ADRD). We propose a fine tuning approach to improve model fairness that takes advantage of observations from the majority groups to build a pre-trained model and uses observations from each underrepresented subgroup to fine tune the pre-trained model, thereby incorporating additional specific knowledge about each subgroup. We find that our fine tuning approach performs better than other algorithmic fairness fixing methods across all subgroups even if the subgroup distribution is very imbalanced and some subgroups are very small. This is an important step toward understanding approaches for improving fairness for healthcare and disease prediction.
Adults with serious mental illnesses are disproportionately affected by chronic health conditions that are linked to inadequately managed medical and psychiatric illnesses and are associated with poor lifestyle behaviors. Emerging intervention models emphasize the value of peer specialists (certified individuals who offer emotional, social, and practical assistance to those with similar lived experiences) in promoting better illness management and meaningful community rehabilitation. Over the last few years, there has been an increasing uptake in the use of digital services and online platforms for the dissemination of various peer services. However, current literature cannot scale current service delivery approaches through audio recording of all interactions to monitor and ensure fidelity at scale. This research aims to understand the individual components of digital peer support to develop a corpus and use natural language processing to classify high-fidelity evidence-based techniques used by peer support specialists in novel datasets. The research hypothesizes that a binary classifier can be developed with an accuracy of 70% through the analysis of digital peer support data.
Colonoscopy plays a critical role in screening of colorectal carcinomas (CC). Unfortunately, the data related to this procedure are stored in disparate documents, colonoscopy, pathology, and radiology reports respectively. The lack of integrated standardized documentation is impeding accurate reporting of quality metrics and clinical and translational research. Natural language processing (NLP) has been used as an alternative to manual data abstraction. Performance of Machine Learning (ML) based NLP solutions is heavily dependent on the accuracy of annotated corpora. Availability of large volume annotated corpora is limited due to data privacy laws and the cost and effort required. In addition, the manual annotation process is error-prone, making the lack of quality annotated corpora the largest bottleneck in deploying ML solutions. The objective of this study is to identify clinical entities critical to colonoscopy quality, and build a high-quality annotated corpus using domain specific taxonomies following standardized annotation guidelines. The annotated corpus can be used to train ML models for a variety of downstream tasks.
Clinical named entity recognition (NER) is an essential building block for many downstream natural language processing (NLP) applications such as information extraction and de-identification. Recently, deep learning (DL) methods that utilize word embeddings have become popular in clinical NLP tasks. However, there has been little work on evaluating and combining the word embeddings trained from different domains. The goal of this study is to improve the performance of NER in clinical discharge summaries by developing a DL model that combines different embeddings and investigate the combination of standard and contextual embeddings from the general and clinical domains. We developed: 1) A human-annotated high-quality internal corpus with discharge summaries and 2) A NER model with an input embedding layer that combines different embeddings: standard word embeddings, context-based word embeddings, a character-level word embedding using a convolutional neural network (CNN), and an external knowledge sources along with word features as one-hot vectors. Embedding was followed by bidirectional long short-term memory (Bi-LSTM) and conditional random field (CRF) layers. The proposed model reaches or overcomes state-of-the-art performance on two publicly available data sets and an F1 score of 94.31% on an internal corpus. After incorporating mixed-domain clinically pre-trained contextual embeddings, the F1 score further improved to 95.36% on the internal corpus. This study demonstrated an efficient way of combining different embeddings that will improve the recognition performance aiding the downstream de-identification of clinical notes.
Colonoscopy is a screening and diagnostic procedure for detection of colorectal carcinomas with specific quality metrics that monitor and improve adenoma detection rates. These quality metrics are stored in disparate documents i.e., colonoscopy, pathology, and radiology reports. The lack of integrated standardized documentation is impeding colorectal cancer research. Clinical concept extraction using Natural Language Processing (NLP) and Machine Learning (ML) techniques is an alternative to manual data abstraction. Contextual word embedding models such as BERT (Bidirectional Encoder Representations from Transformers) and FLAIR have enhanced performance of NLP tasks. Combining multiple clinically-trained embeddings can improve word representations and boost the performance of the clinical NLP systems. The objective of this study is to extract comprehensive clinical concepts from the consolidated colonoscopy documents using concatenated clinical embeddings. We built high-quality annotated corpora for three report types. BERT and FLAIR embeddings were trained on unlabeled colonoscopy related documents. We built a hybrid Artificial Neural Network (h-ANN) to concatenate and fine-tune BERT and FLAIR embeddings. To extract concepts of interest from three report types, 3 models were initialized from the h-ANN and fine-tuned using the annotated corpora. The models achieved best F1-scores of 91.76%, 92.25%, and 88.55% for colonoscopy, pathology, and radiology reports respectively.
