Computer-Assisted Cohort Identification in Practice

The standard approach to expert-in-the-loop machine learning is active learning, where, repeatedly, an expert is asked to annotate one or more records and the machine finds a classifier that respects all annotations made until that point. We propose an alternative approach, IQRef , in which the expert iteratively designs a classifier and the machine helps him or her to determine how well it is performing and, importantly, when to stop, by reporting statistics on a fixed, hold-out sample of annotated records. We justify our approach based on prior work giving a theoretical model of how to re-use hold-out data. We compare the two approaches in the context of identifying a cohort of EHRs and examine their strengths and weaknesses through a case study arising from an optometric research problem. We conclude that both approaches are complementary, and we recommend that they both be employed in conjunction to address the problem of cohort identification in health research.

Explaining Machine Learning Models for Clinical Gait Analysis

Machine Learning (ML) is increasingly used to support decision-making in the healthcare sector. While ML approaches provide promising results with regard to their classification performance, most share a central limitation, their black-box character. This article investigates the usefulness of Explainable Artificial Intelligence (XAI) methods to increase transparency in automated clinical gait classification based on time series. For this purpose, predictions of state-of-the-art classification methods are explained with a XAI method called Layer-wise Relevance Propagation (LRP). Our main contribution is an approach that explains class-specific characteristics learned by ML models that are trained for gait classification. We investigate several gait classification tasks and employ different classification methods, i.e., Convolutional Neural Network, Support Vector Machine, and Multi-layer Perceptron. We propose to evaluate the obtained explanations with two complementary approaches: a statistical analysis of the underlying data using Statistical Parametric Mapping and a qualitative evaluation by two clinical experts. A gait dataset comprising ground reaction force measurements from 132 patients with different lower-body gait disorders and 62 healthy controls is utilized. Our experiments show that explanations obtained by LRP exhibit promising statistical properties concerning inter-class discriminativity and are also in line with clinically relevant biomechanical gait characteristics.

Emotion Recognition Robust to Indoor Environmental Distortions and Non-targeted Emotions Using Out-of-distribution Detection

The rapid development of machine learning on acoustic signal processing has resulted in many solutions for detecting emotions from speech. Early works were developed for clean and acted speech and for a fixed set of emotions. Importantly, the datasets and solutions assumed that a person only exhibited one of these emotions. More recent work has continually been adding realism to emotion detection by considering issues such as reverberation, de-amplification, and background noise, but often considering one dataset at a time, and also assuming all emotions are accounted for in the model. We significantly improve realistic considerations for emotion detection by (i) more comprehensively assessing different situations by combining the five common publicly available datasets as one and enhancing the new dataset with data augmentation that considers reverberation and de-amplification, (ii) incorporating 11 typical home noises into the acoustics, and (iii) considering that in real situations a person may be exhibiting many emotions that are not currently of interest and they should not have to fit into a pre-fixed category nor be improperly labeled. Our novel solution combines CNN with out-of-data distribution detection. Our solution increases the situations where emotions can be effectively detected and outperforms a state-of-the-art baseline.

A Minimalist Method Toward Severity Assessment and Progression Monitoring of Obstructive Sleep Apnea on the Edge

Artificial Intelligence-enabled applications on edge devices have the potential to revolutionize disease detection and monitoring in future smart health (sHealth) systems. In this study, we investigated a minimalist approach for the severity classification, severity estimation, and progression monitoring of obstructive sleep apnea (OSA) in a home environment using wearables. We used the recursive feature elimination technique to select the best feature set of 70 features from a total of 200 features extracted from polysomnogram. We used a multi-layer perceptron model to investigate the performance of OSA severity classification with all the ranked features to a subset of features available from either Electroencephalography or Heart Rate Variability (HRV) and time duration of SpO2 level. The results indicate that using only computationally inexpensive features from HRV and SpO2, an area under the curve of 0.91 and an accuracy of 83.97% can be achieved for the severity classification of OSA. For estimation of the apnea-hypopnea index, the accuracy of RMSE = 4.6 and R-squared value = 0.71 have been achieved in the test set using only ranked HRV and SpO2 features. The Wilcoxon-signed-rank test indicates a significant change (p < 0.05) in the selected feature values for a progression in the disease over 2.5 years. The method has the potential for integration with edge computing for deployment on everyday wearables. This may facilitate the preliminary severity estimation, monitoring, and management of OSA patients and reduce associated healthcare costs as well as the prevalence of untreated OSA.

Automatic Parotid Gland Segmentation in MVCT Using Deep Convolutional Neural Networks

Radiation-induced xerostomia, as a major problem in radiation treatment of the head and neck cancer, is mainly due to the overdose irradiation injury to the parotid glands. Helical Tomotherapy-based megavoltage computed tomography (MVCT) imaging during the Tomotherapy treatment can be applied to monitor the successive variations in the parotid glands. While manual segmentation is time consuming, laborious, and subjective, automatic segmentation is quite challenging due to the complicated anatomical environment of head and neck as well as noises in MVCT images. In this article, we propose a localization-refinement scheme to segment the parotid gland in MVCT. After data pre-processing we use mask region convolutional neural network (Mask R-CNN) in the localization stage after data pre-processing, and design a modified U-Net in the following fine segmentation stage. To the best of our knowledge, this study is a pioneering work of deep learning on MVCT segmentation. Comprehensive experiments based on different data distribution of head and neck MVCTs and different segmentation models have demonstrated the superiority of our approach in terms of accuracy, effectiveness, flexibility, and practicability. Our method can be adopted as a powerful tool for radiation-induced injury studies, where accurate organ segmentation is crucial.

Attention-based Unsupervised Keyphrase Extraction and Phrase Graph for COVID-19 Medical Literature Retrieval

Searching, reading, and finding information from the massive medical text collections are challenging. A typical biomedical search engine is not feasible to navigate each article to find critical information or keyphrases. Moreover, few tools provide a visualization of the relevant phrases to the query. However, there is a need to extract the keyphrases from each document for indexing and efficient search. The transformer-based neural networks—BERT has been used for various natural language processing tasks. The built-in self-attention mechanism can capture the associations between words and phrases in a sentence. This research investigates whether the self-attentions can be utilized to extract keyphrases from a document in an unsupervised manner and identify relevancy between phrases to construct a query relevancy phrase graph to visualize the search corpus phrases on their relevancy and importance. The comparison with six baseline methods shows that the self-attention-based unsupervised keyphrase extraction works well on a medical literature dataset. This unsupervised keyphrase extraction model can also be applied to other text data. The query relevancy graph model is applied to the COVID-19 literature dataset and to demonstrate that the attention-based phrase graph can successfully identify the medical phrases relevant to the query terms.

Supporting Personalized Health Care With Social Media Analytics: An Application to Hypothyroidism

Social media analytics can considerably contribute to understanding health conditions beyond clinical practice, by capturing patients’ discussions and feelings about their quality of life in relation to disease treatments. In this article, we propose a methodology to support a detailed analysis of the therapeutic experience in patients affected by a specific disease, as it emerges from health forums. As a use case to test the proposed methodology, we analyze the experience of patients affected by hypothyroidism and their reactions to standard therapies. Our approach is based on a data extraction and filtering pipeline, a novel topic detection model named Generative Text Compression with Agglomerative Clustering Summarization ( GTCACS ), and an in-depth data analytic process. We advance the state of the art on automated detection of adverse drug reactions ( ADRs ) since, rather than simply detecting and classifying positive or negative reactions to a therapy, we are capable of providing a fine characterization of patients along different dimensions, such as co-morbidities, symptoms, and emotional states.

GeCoAgent: A Conversational Agent for Empowering Genomic Data Extraction and Analysis

With the availability of reliable and low-cost DNA sequencing, human genomics is relevant to a growing number of end-users, including biologists and clinicians. Typical interactions require applying comparative data analysis to huge repositories of genomic information for building new knowledge, taking advantage of the latest findings in applied genomics for healthcare. Powerful technology for data extraction and analysis is available, but broad use of the technology is hampered by the complexity of accessing such methods and tools. This work presents GeCoAgent, a big-data service for clinicians and biologists. GeCoAgent uses a dialogic interface, animated by a chatbot, for supporting the end-users’ interaction with computational tools accompanied by multi-modal support. While the dialogue progresses, the user is accompanied in extracting the relevant data from repositories and then performing data analysis, which often requires the use of statistical methods or machine learning. Results are returned using simple representations (spreadsheets and graphics), while at the end of a session the dialogue is summarized in textual format. The innovation presented in this article is concerned with not only the delivery of a new tool but also our novel approach to conversational technologies, potentially extensible to other healthcare domains or to general data science.

Evaluation of U-Net Based Architectures for Automatic Aortic Dissection Segmentation

Segmentation and reconstruction of arteries is important for a variety of medical and engineering fields, such as surgical planning and physiological modeling. However, manual methods can be laborious and subject to a high degree of human variability. In this work, we developed various convolutional neural network ( CNN ) architectures to segment Stanford type B aortic dissections ( TBADs ), characterized by a tear in the descending aortic wall creating a normal channel of blood flow called a true lumen and a pathologic channel within the wall called a false lumen. We introduced several variations to the two-dimensional ( 2D ) and three-dimensional (3 D ) U-Net, where small stacks of slices were inputted into the networks instead of individual slices or whole geometries. We compared these variations with a variety of CNN segmentation architectures and found that stacking the input data slices in the upward direction with 2D U-Net improved segmentation accuracy, as measured by the Dice similarity coefficient ( DC ) and point-by-point average distance ( AVD ), by more than 15\% . Our optimal architecture produced DC scores of 0.94, 0.88, and 0.90 and AVD values of 0.074, 0.22, and 0.11 in the whole aorta, true lumen, and false lumen, respectively. Altogether, the predicted reconstructions closely matched manual reconstructions.