Machine and deep learning methods for clinical outcome prediction based on physiological data of COVID-19 patients: a scoping review.

INTRODUCTION: Since the beginning of the COVID-19 pandemic, numerous machine and deep learning (MDL) methods have been proposed in the literature to analyze patient physiological data. The objective of this review is to summarize various aspects of these methods and assess their practical utility for predicting various clinical outcomes. METHODS: We searched PubMed, Scopus, and Cochrane Library, screened and selected the studies matching the inclusion criteria. The clinical analysis focused on the characteristics of the patient cohorts in the studies included in this review, the specific tasks in the context of the COVID-19 pandemic that machine and deep learning methods were used for, and their practical limitations. The technical analysis focused on the details of specific MDL methods and their performance. RESULTS: Analysis of the 48 selected studies revealed that the majority (∼54 %) of them examined the application of MDL methods for the prediction of survival/mortality-related patient outcomes, while a smaller fraction (∼13 %) of studies also examined applications to the prediction of patients' physiological outcomes and hospital resource utilization. 21 % of the studies examined the application of MDL methods to multiple clinical tasks. Machine and deep learning methods have been shown to be effective at predicting several outcomes of COVID-19 patients, such as disease severity, complications, intensive care unit (ICU) transfer, and mortality. MDL methods also achieved high accuracy in predicting the required number of ICU beds and ventilators. CONCLUSION: Machine and deep learning methods have been shown to be valuable tools for predicting disease severity, organ dysfunction and failure, patient outcomes, and hospital resource utilization during the COVID-19 pandemic. The discovered knowledge and our conclusions and recommendations can also be useful to healthcare professionals and artificial intelligence researchers in managing future pandemics.

Improving focusability of post-compressed PW laser pulses using a deformable mirror.

The use of the post-compression technique ensures gain in laser pulse peak power but at the same time degrades beam focusability due to the nonlinear wavefront distortions caused by a spatially nonuniform beam profile. In this paper a substantial focusability improvement of a post-compressed laser pulse by means of adaptive optics was demonstrated experimentally. The Strehl ratio increase from 0.16 to 0.43 was measured. Simulations showed that the peak intensity in this case reaches 0.52 of the theoretical limit.

Identifying Effective Motivational Interviewing Communication Sequences Using Automated Pattern Analysis.

Motivational Interviewing (MI) is an evidence-based strategy for communicating with patients about behavior change. Although there is strong empirical evidence linking "MI-consistent" counselor behaviors and patient motivational statements (i.e., "change talk"), the specific counselor communication behaviors effective for eliciting patient change talk vary by treatment context and, thus, are a subject of ongoing research. An integral part of this research is the sequential analysis of pre-coded MI transcripts. In this paper, we evaluate the empirical effectiveness of the Hidden Markov Model, a probabilistic generative model for sequence data, for modeling sequences of behavior codes and closed frequent pattern mining, a method to identify frequently occurring sequential patterns of behavior codes in MI communication sequences to inform MI practice. We conducted experiments with 1,360 communication sequences from 37 transcribed audio recordings of weight loss counseling sessions with African-American adolescents with obesity and their caregivers. Transcripts had been previously annotated with patient-counselor behavior codes using a specialized codebook. Empirical results indicate that Hidden Markov Model and closed frequent pattern mining techniques can identify counselor communication strategies that are effective at eliciting patients' motivational statements to guide clinical practice.

Deep Neural Architectures for Discourse Segmentation in E-Mail Based Behavioral Interventions.

Communication science approaches to develop effective behavior interventions, such as motivational interviewing (MI), are limited by traditional qualitative coding of communication exchanges, a very resource-intensive and time-consuming process. This study focuses on the analysis of e-Coaching sessions, behavior interventions delivered via email and grounded in the principles of MI. A critical step towards automated qualitative coding of e-Coaching sessions is segmentation of emails into fragments that correspond to MI behaviors. This study frames email segmentation task as a classification problem and utilizes word and punctuation mark embeddings in conjunction with part-of-speech features to address it. We evaluated the performance of conditional random fields (CRF) as well as multi-layer perceptron (MLP), bi-directional recurrent neural network (BRNN) and convolutional recurrent neural network (CRNN) for the task of email segmentation. Our results indicate that CRNN outperforms CRF, MLP and BRNN achieving 0.989 weighted macro-averaged F1-measure and 0.825 F1-measure for new segment detection.

Bayesian approach to incorporating different types of biomedical knowledge bases into information retrieval systems for clinical decision support in precision medicine.

By providing clinicians with information regarding treatment options for molecular sub-types of complex diseases with genetic origin, such as cancer, information retrieval (IR) systems play an important role in precision medicine. In this paper, we propose Bayesian Precision Medicine (BPM), a novel probabilistic framework for query expansion in information retrieval systems for Clinical Decision Support (CDS) in Precision Medicine (PM). Such systems can assist clinicians with selecting personalized treatment of complex diseases based on the patients' genomic data, such as gene mutations. In particular, we focus on a clinical decision support scenario in which clinicians provide two types of information in their queries: (1) short description of a patient's case, which may contain information regarding the type of cancer that a patient has as well as symptoms and demographics, and (2) gene mutations, which may contain gene names, mutation code and type of mutation. The goal of an IR system in this scenario is to rank biomedical articles from a large collection, such as the MEDLINE, based on their relevance to the provided query. One of the main challenges faced by IR systems in this scenario is semantic matching of heterogeneous information (gene names, medical terminology and other query keywords) in queries and relevant biomedical articles. To address this challenge, we propose a probabilistic framework that enables mapping gene mutations provided in a given query onto the biomedical concepts that are related to the entire query and can be effectively utilized for query expansion. The BPM obtains candidate query expansion concepts from biomedical knowledge bases, the Unified Medical Language System (UMLS) and the Drug-Gene Interaction Database (DGIdb), as well as the top-ranked MEDLINE articles retrieved for the original query. The BPM then utilizes information from the Catalog of Somatic Mutations in Cancer (COSMIC) and co-occurrence statistics in MEDLINE to assess the relatedness of candidate query expansion concepts to gene mutations and other information provided in a query. Experimental evaluation of the BPM was conducted on a large subset of MEDLINE articles as well as abstracts from the American Association for Cancer Research (AACR) and American Society of Clinical Oncology (ASCO) proceedings. Experimental results on a publicly available benchmark provided by the 2017 TREC precision medicine track indicate that the proposed probabilistic framework is effective at utilizing both genomic and textual information in queries to improve the accuracy of IR systems for CDS in PM through query expansion.

Developing Machine Learning Models for Behavioral Coding.

OBJECTIVE: The goal of this research is to develop a machine learning supervised classification model to automatically code clinical encounter transcripts using a behavioral code scheme. METHODS: We first evaluated the efficacy of eight state-of-the-art machine learning classification models to recognize patient-provider communication behaviors operationalized by the motivational interviewing framework. Data were collected during the course of a single weight loss intervention session with 37 African American adolescents and their caregivers. We then tested the transferability of the model to a novel treatment context, 80 patient-provider interactions during routine human immunodeficiency virus (HIV) clinic visits. RESULTS: Of the eight models tested, the support vector machine model demonstrated the best performance, achieving a .680 F1-score (a function of model precision and recall) in adolescent and .639 in caregiver sessions. Adding semantic and contextual features improved accuracy with 75.1% of utterances in adolescent and 73.8% in caregiver sessions correctly coded. With no modification, the model correctly classified 72.0% of patient-provider utterances in HIV clinical encounters with reliability comparable to human coders (k = .639). CONCLUSIONS: The development of a validated approach for automatic behavioral coding offers an efficient alternative to traditional, resource-intensive methods with the potential to dramatically accelerate the pace of outcomes-oriented behavioral research. The knowledge gained from computer-driven behavioral research can inform clinical practice by providing clinicians with empirically supported communication strategies to tailor their conversations with patients. Lastly, automatic behavioral coding is a critical first step toward fully automated eHealth/mHealth (electronic/mobile Health) behavioral interventions.

Predicting the Outcome of Patient-Provider Communication Sequences using Recurrent Neural Networks and Probabilistic Models.

The problem of analyzing temporally ordered sequences of observations generated by molecular, physiological or psychological processes to make predictions about the outcome of these processes arises in many domains of clinical informatics. In this paper, we focus on predicting the outcome of patient-provider communication sequences in the context of the clinical dialog. Specifically, we consider prediction of the motivational interview success (i.e. eliciting a particular type of patient behavioral response) based on an observed sequence of coded patient-provider communication exchanges as a sequence classification problem. We propose two solutions to this problem, one that is based on Recurrent Neural Networks (RNNs) and another that is based on Markov Chain (MC) and Hidden Markov Model (HMM), and compare the accuracy of these solutions using communication sequences annotated with behavior codes from the real-life motivational interviews. Our experiments indicate that the deep learning-based approach is significantly more accurate than the approach based on probabilistic models in predicting the success of motivational interviews (0.8677 versus 0.7038 and 0.6067 F1-score by RNN, MC and HMM, respectively, when using undersampling to correct for class imbalance, and 0.8381 versus 0.7775 and 0.7520 F1-score by RNN, MC and HMM, respectively, when using over-sampling). These results indicate that the proposed method can be used for real-time monitoring of progression of clinical interviews and more efficient identification of effective provider communication strategies, which in turn can significantly decrease the effort required to develop behavioral interventions and increase their effectiveness.

The Conducting Spin-Crossover Compound Combining Fe(II) Cation Complex with TCNQ in a Fractional Reduction State.

The radical anion salt [Fe{HC(pz)3}2](TCNQ)3 demonstrates conductivity and spin-crossover (SCO) transition associated with Fe(II) complex cation subsystem. It was synthesized and structurally characterized at temperatures 100, 300, 400, and 450 K. The compound demonstrates unusual for 7,7,8,8,-tetracyanoquinodimethane (TCNQ)-based salts quasi-two-dimensional conductivity. Pronounced changes of the in-plane direct-current resistivity and intensity of the electron paramagnetic resonance (EPR) signal, originated from TCNQ subsystem, precede the SCO transition at the midpoint T* = 445 K. The boltzmannian growth of the total magnetic response and structural changes in the vicinity of T* uniquely show that half [Fe{HC(pz)3}2] cations exist in high-spin state. Robust broadening of the EPR signal triggered by the SCO transition is interpreted in terms of cross relaxation between the TCNQ and Fe(II) spin subsystems.

A study of the effectiveness of machine learning methods for classification of clinical interview fragments into a large number of categories.

This study examines the effectiveness of state-of-the-art supervised machine learning methods in conjunction with different feature types for the task of automatic annotation of fragments of clinical text based on codebooks with a large number of categories. We used a collection of motivational interview transcripts consisting of 11,353 utterances, which were manually annotated by two human coders as the gold standard, and experimented with state-of-art classifiers, including Naïve Bayes, J48 Decision Tree, Support Vector Machine (SVM), Random Forest (RF), AdaBoost, DiscLDA, Conditional Random Fields (CRF) and Convolutional Neural Network (CNN) in conjunction with lexical, contextual (label of the previous utterance) and semantic (distribution of words in the utterance across the Linguistic Inquiry and Word Count dictionaries) features. We found out that, when the number of classes is large, the performance of CNN and CRF is inferior to SVM. When only lexical features were used, interview transcripts were automatically annotated by SVM with the highest classification accuracy among all classifiers of 70.8%, 61% and 53.7% based on the codebooks consisting of 17, 20 and 41 codes, respectively. Using contextual and semantic features, as well as their combination, in addition to lexical ones, improved the accuracy of SVM for annotation of utterances in motivational interview transcripts with a codebook consisting of 17 classes to 71.5%, 74.2%, and 75.1%, respectively. Our results demonstrate the potential of using machine learning methods in conjunction with lexical, semantic and contextual features for automatic annotation of clinical interview transcripts with near-human accuracy.

Interpretable Probabilistic Latent Variable Models for Automatic Annotation of Clinical Text.

We propose Latent Class Allocation (LCA) and Discriminative Labeled Latent Dirichlet Allocation (DL-LDA), two novel interpretable probabilistic latent variable models for automatic annotation of clinical text. Both models separate the terms that are highly characteristic of textual fragments annotated with a given set of labels from other non-discriminative terms, but rely on generative processes with different structure of latent variables. LCA directly learns class-specific multinomials, while DL-LDA breaks them down into topics (clusters of semantically related words). Extensive experimental evaluation indicates that the proposed models outperform Naïve Bayes, a standard probabilistic classifier, and Labeled LDA, a state-of-the-art topic model for labeled corpora, on the task of automatic annotation of transcripts of motivational interviews, while the output of the proposed models can be easily interpreted by clinical practitioners.

Monoclonal antibody to an endogenous bufadienolide, marinobufagenin, reverses preeclampsia-induced Na/K-ATPase inhibition and lowers blood pressure in NaCl-sensitive hypertension.

BACKGROUND: Levels of marinobufagenin (MBG), an endogenous bufadienolide Na/K-ATPase (NKA) inhibitor, increase in preeclampsia and in NaCl-sensitive hypertension. METHODS: We tested a 3E9 monoclonal anti-MBG antibody (mAb) for the ability to lower blood pressure (BP) in NaCl-sensitive hypertension and to reverse the preeclampsia-induced inhibition of erythrocyte NKA. Measurements of MBG were performed via immunoassay based on 4G4 anti-MBG mAb. RESULTS: In hypertensive Dahl-S rats, intraperitoneal administration of 50 microg/kg 3E9 mAb lowered BP by 32 mmHg and activated the Na/K-pump in the thoracic aorta by 51%. NaCl supplementation of pregnant rats (n = 16) produced a 37 mmHg increase in BP, a 3.5-fold rise in MBG excretion, and a 25% inhibition of the Na/K-pump in the thoracic aorta, compared with pregnant rats on a normal NaCl intake. In eight pregnant hypertensive rats, 3E9 mAb reduced the BP (21 mmHg) and restored the vascular Na/K-pump. In 14 patients with preeclampsia (mean BP, 126 +/- 3 mmHg; 26.9 +/- 1.4 years; gestational age, 37 +/- 0.8 weeks), plasma MBG was increased three-fold and erythrocyte NKA was inhibited compared with that of 12 normotensive pregnant women (mean BP, 71 +/- 3 mmHg) (1.5 +/- 0.1 vs. 3.1 +/- 0.2 micromol Pi/ml/h, respectively; P < 0.01). Ex-vivo 3E9 mAb restored NKA activity in erythrocytes from patients with preeclampsia. As compared with 3E9 mAb, Digibind, an affinity-purified antidigoxin antibody, was less active with respect to lowering BP in both hypertensive models and to restoration of NKA from erythrocytes from patients with preeclampsia. CONCLUSION: Anti-MBG mAbs may be a useful tool in studies of MBG in vitro and in vivo and may offer treatment of preeclampsia.

Signs of proinflammatory/genotoxic switch (adipogenotoxicosis) in mammary fat of breast cancer patients: role of menopausal status, estrogens and hyperglycemia.

The abundance of fat tissue surrounding normal and malignant epithelial mammary cells raises the questions whether such "adipose milieu" is important in the local proinflammatory/genotoxic shift, which apparently promotes tumor development and worsens prognosis, and what conditions stimulate this shift, or "adipogenotoxicosis." We studied 95 mammary fat samples from 70 postmenopausal and 25 premenopausal breast cancer (BC) patients at a distance of 1.5-2.0 cm from tumors. The levels of leptin, adiponectin, TNFalpha and IL-6 release after 4-hr incubation of the samples were evaluated with ELISA, nitric oxide (NO) production by Griess reaction and lipid peroxidation by determination of thiobarbiturate-reactive products (TBRP). Infiltration of fat with macrophages (CD68-positive cells) and expression of cytochrome P450 1B1/estrogen 4-hydroxylase (CYP1B1) were detected by immunohistochemistry. Aromatase (CYP19) activity in mammary fat was measured by (3)H(2)O release from (3)H-1beta-androstenedione. In the postmenopausal BC patients, NO and TNFalpha production by adipose tissue explants increased independent of BMI and in parallel with decreasing leptin and, especially, adiponectin release. In the premenopausal patients, higher CYP1B1 expression and TBRP level were found in mammary fat, while higher aromatase activity was combined with higher CYP1B1 expression as well as NO and IL-6 production. In the postmenopausal group, impaired glucose tolerance was associated with higher IL-6 release production by fat and with higher IL-6/adiponectin ratio. Thus, signs of adipogenotoxicosis in mammary fat can be found in both pre- and postmenopausal BC patients. This condition is likely being maintained through estrogen- and glucose-related factors and mechanisms presumably associated with less favorable types of hormonal carcinogenesis.

Circumventing recombination events encountered with production of a clinical-grade adenoviral vector with a double-expression cassette.

Delivery of multiple exogenous genes into target cells is important for a broad range of gene therapy applications, including combined therapeutic gene expression and noninvasive imaging. Previous studies ( Mol Ther 4: 223-231, 2001 ) have described the adenoviral vector RGDTKSSTR with a double-expression cassette that encodes herpes simplex virus thymidine kinase (HSVtk) for molecular chemotherapy and human somatostatin receptor subtype-2 (hSSTR2) for indirect imaging. In this vector, both genes are inserted in place of the E1 region of the adenoviral genome and expressed independently from two cytomegalovirus (CMV) promoters. During production of clinical-grade RGDTKSSTR, we found that the CMV promoters and simian virus 40 (SV40) poly(A) regions located in both expression cassettes provoked homologous recombination and deletion of one of the cassettes. To resolve this problem, we designed a strategy for substituting the duplicate promoters and poly(A) regions. We placed the hSSTR2 gene in the new Ad5.SSTR/TK.RGD vector under the control of a CMV promoter with a bovine growth hormone poly(A) region, whereas the SV40 promoter, enhancer, and poly(A) signal controlled HSVtk expression. This use of different regulatory sequences allowed independent expression of both transgenes from a single adenoviral vector and circumvented the recombination problem. Reconstruction of the vector with a double-expression cassette enables its use in human clinical trials.

Targeting of adenovirus via genetic modification of the viral capsid combined with a protein bridge.

A potential barrier to the development of genetically targeted adenovirus (Ad) vectors for cell-specific delivery of gene therapeutics lies in the fact that several types of targeting protein ligands require posttranslational modifications, such as the formation of disulfide bonds, which are not available to Ad capsid proteins due to their nuclear localization during assembly of the virion. To overcome this problem, we developed a new targeting strategy, which combines genetic modifications of the Ad capsid with a protein bridge approach, resulting in a vector-ligand targeting complex. The components of the complex associate by virtue of genetic modifications to both the Ad capsid and the targeting ligand. One component of this mechanism of association, the Fc-binding domain of Staphylococcus aureus protein A, is genetically incorporated into the Ad fiber protein. The ligand is comprised of a targeting component fused with the Fc domain of immunoglobulin, which serves as a docking moiety to bind to these genetically modified fibers during the formation of the Ad-ligand complex. The modular design of the ligand solves the problem of structural and biosynthetic compatibility with the Ad and thus facilitates targeting of the vector to a variety of cellular receptors. Our study shows that targeting ligands incorporating the Fc domain and either an anti-CD40 single-chain antibody or CD40L form stable complexes with protein A-modified Ad vectors, resulting in significant augmentation of gene delivery to CD40-positive target cells. Since this gene transfer is independent of the expression of the native Ad5 receptor by the target cells, this strategy results in the derivation of truly targeted Ad vectors suitable for tissue-specific gene therapy.

Biological Activity of Peptide SCV-07 Against Murine Tuberculosis.

SCV-07 (gamma-glutamyl-tryptophan) is a new immunomodulatory compound that was developed and patented both for composition and immunomodulatory use. SCV-07 was shown to have a broad spectrum of immunostimulatory activities both in vitro and in vivo. In the present study we investigated the biological activity of SCV-07 in a murine model of experimental tuberculosis (TB) induced with M. bovis-bovinus 8 strain. Therapy with SCV-07 at doses of 0.01, 0.1, and 1 &mgr;g/kg (5 daily injections) decreased the lung damage index compared to untreated controls and to those treated with isoniazid alone. The growth of M. bovis-bovinus 8 in spleen culture was decreased. Cytokine studies showed that on the 24th day after the treatment with SCV-07 the production of IL-2 was restored to the level seen in uninfected animals. Proliferative responses for both thymic and spleen cells were nearly restored to the responses observed in uninfected animals. IFN-gamma production by both thymic and spleen cells, as well as its circulating levels in serum, was increased by the SCV-07 treatment. Concurrently, IL-4 production was decreased in the same cell types and the serum. These changes suggest that SCV-07 is stimulating a shift of T helper cells to a Th1-like immune response. SCV-07 treatment also stimulated the macrophage functions, which had been decreased by tuberculosis infection and isoniazid therapy, with an improved phagocytosis activity of peritoneal macrophage. The obtained results suggest that SCV-07 treatment increases the efficacy of anti-tuberculosis therapy as well as the strength of the immune response. Thus, SCV-07 is a prospective immunomodulator for a complex therapy of TB.