Towards deep phenotyping pregnancy: a systematic review on artificial intelligence and machine learning methods to improve pregnancy outcomes.

OBJECTIVE: Development of novel informatics methods focused on improving pregnancy outcomes remains an active area of research. The purpose of this study is to systematically review the ways that artificial intelligence (AI) and machine learning (ML), including deep learning (DL), methodologies can inform patient care during pregnancy and improve outcomes. MATERIALS AND METHODS: We searched English articles on EMBASE, PubMed and SCOPUS. Search terms included ML, AI, pregnancy and informatics. We included research articles and book chapters, excluding conference papers, editorials and notes. RESULTS: We identified 127 distinct studies from our queries that were relevant to our topic and included in the review. We found that supervised learning methods were more popular (n = 69) than unsupervised methods (n = 9). Popular methods included support vector machines (n = 30), artificial neural networks (n = 22), regression analysis (n = 17) and random forests (n = 16). Methods such as DL are beginning to gain traction (n = 13). Common areas within the pregnancy domain where AI and ML methods were used the most include prenatal care (e.g. fetal anomalies, placental functioning) (n = 73); perinatal care, birth and delivery (n = 20); and preterm birth (n = 13). Efforts to translate AI into clinical care include clinical decision support systems (n = 24) and mobile health applications (n = 9). CONCLUSIONS: Overall, we found that ML and AI methods are being employed to optimize pregnancy outcomes, including modern DL methods (n = 13). Future research should focus on less-studied pregnancy domain areas, including postnatal and postpartum care (n = 2). Also, more work on clinical adoption of AI methods and the ethical implications of such adoption is needed.

Predicting acute kidney injury following open partial nephrectomy treatment using SAT-pruned explainable machine learning model.

BACKGROUND: One of the most prevalent complications of Partial Nephrectomy (PN) is Acute Kidney Injury (AKI), which could have a negative impact on subsequent renal function and occurs in up to 24.3% of patients undergoing PN. The aim of this study was to predict the occurrence of AKI following PN using preoperative parameters by applying machine learning algorithms. METHODS: We included all adult patients (n = 723) who underwent open PN in our department since 1995 and on whom we have data on the pre-operative renal function. We developed a random forest (RF) model with Boolean satisfaction-based pruned decision trees for binary classification (AKI or non-AKI). Hyper-parameter grid search was performed to optimize the model's performance. Fivefold cross-validation was applied to evaluate the model. We implemented a RF model with greedy feature selection to binary classify AKI and non-AKI cases based on pre-operative data. RESULTS: The best model obtained a 0.69 precision and 0.69 recall in classifying the AKI and non-AKI groups on average (k = 5). In addition, the model's probability to correctly classify a new prediction is 0.75. The proposed model is available as an online calculator. CONCLUSIONS: Our model predicts the occurrence of AKI following open PN with (75%) accuracy. We plan to externally validate this model and modify it to minimally-invasive PN.

Brain tissues have single-voxel signatures in multi-spectral MRI.

Since the seminal works by Brodmann and contemporaries, it is well-known that different brain regions exhibit unique cytoarchitectonic and myeloarchitectonic features. Transferring the approach of classifying brain tissues - and other tissues - based on their intrinsic features to the realm of magnetic resonance (MR) is a longstanding endeavor. In the 1990s, atlas-based segmentation replaced earlier multi-spectral classification approaches because of the large overlap between the class distributions. Here, we explored the feasibility of performing global brain classification based on intrinsic MR features, and used several technological advances: ultra-high field MRI, q-space trajectory diffusion imaging revealing voxel-intrinsic diffusion properties, chemical exchange saturation transfer and semi-solid magnetization transfer imaging as a marker of myelination and neurochemistry, and current neural network architectures to analyze the data. In particular, we used the raw image data as well to increase the number of input features. We found that a global brain classification of roughly 97 brain regions was feasible with gross classification accuracy of 60%; and that mapping from voxel-intrinsic MR data to the brain region to which the data belongs is possible. This indicates the presence of unique MR signals of different brain regions, similar to their cytoarchitectonic and myeloarchitectonic fingerprints.

Predictive Pattern Classification Can Distinguish Gender Identity Subtypes from Behavior and Brain Imaging.

The exact neurobiological underpinnings of gender identity (i.e., the subjective perception of oneself belonging to a certain gender) still remain unknown. Combining both resting-state functional connectivity and behavioral data, we examined gender identity in cisgender and transgender persons using a data-driven machine learning strategy. Intrinsic functional connectivity and questionnaire data were obtained from cisgender (men/women) and transgender (trans men/trans women) individuals. Machine learning algorithms reliably detected gender identity with high prediction accuracy in each of the four groups based on connectivity signatures alone. The four normative gender groups were classified with accuracies ranging from 48% to 62% (exceeding chance level at 25%). These connectivity-based classification accuracies exceeded those obtained from a widely established behavioral instrument for gender identity. Using canonical correlation analyses, functional brain measurements and questionnaire data were then integrated to delineate nine canonical vectors (i.e., brain-gender axes), providing a multilevel window into the conventional sex dichotomy. Our dimensional gender perspective captures four distinguishable brain phenotypes for gender identity, advocating a biologically grounded reconceptualization of gender dimorphism. We hope to pave the way towards objective, data-driven diagnostic markers for gender identity and transgender, taking into account neurobiological and behavioral differences in an integrative modeling approach.

Predictive value of ATN biomarker profiles in estimating disease progression in Alzheimer's disease dementia.

We aimed to evaluate the value of ATN biomarker classification system (amyloid beta [A], pathologic tau [T], and neurodegeneration [N]) for predicting conversion from mild cognitive impairment (MCI) to dementia. In a sample of people with MCI (n = 415) we assessed predictive performance of ATN classification using empirical knowledge-based cut-offs for each component of ATN and compared it to two data-driven approaches, logistic regression and RUSBoost machine learning classifiers, which used continuous clinical or biomarker scores. In data-driven approaches, we identified ATN features that distinguish normals from individuals with dementia and used them to classify persons with MCI into dementia-like and normal groups. Both data-driven classification methods performed better than the empirical cut-offs for ATN biomarkers in predicting conversion to dementia. Classifiers that used clinical features performed as well as classifiers that used ATN biomarkers for prediction of progression to dementia. We discuss that data-driven modeling approaches can improve our ability to predict disease progression and might have implications in future clinical trials.

Classification of Current Procedural Terminology Codes from Electronic Health Record Data Using Machine Learning.

BACKGROUND: Accurate anesthesiology procedure code data are essential to quality improvement, research, and reimbursement tasks within anesthesiology practices. Advanced data science techniques, including machine learning and natural language processing, offer opportunities to develop classification tools for Current Procedural Terminology codes across anesthesia procedures. METHODS: Models were created using a Train/Test dataset including 1,164,343 procedures from 16 academic and private hospitals. Five supervised machine learning models were created to classify anesthesiology Current Procedural Terminology codes, with accuracy defined as first choice classification matching the institutional-assigned code existing in the perioperative database. The two best performing models were further refined and tested on a Holdout dataset from a single institution distinct from Train/Test. A tunable confidence parameter was created to identify cases for which models were highly accurate, with the goal of at least 95% accuracy, above the reported 2018 Centers for Medicare and Medicaid Services (Baltimore, Maryland) fee-for-service accuracy. Actual submitted claim data from billing specialists were used as a reference standard. RESULTS: Support vector machine and neural network label-embedding attentive models were the best performing models, respectively, demonstrating overall accuracies of 87.9% and 84.2% (single best code), and 96.8% and 94.0% (within top three). Classification accuracy was 96.4% in 47.0% of cases using support vector machine and 94.4% in 62.2% of cases using label-embedding attentive model within the Train/Test dataset. In the Holdout dataset, respective classification accuracies were 93.1% in 58.0% of cases and 95.0% among 62.0%. The most important feature in model training was procedure text. CONCLUSIONS: Through application of machine learning and natural language processing techniques, highly accurate real-time models were created for anesthesiology Current Procedural Terminology code classification. The increased processing speed and a priori targeted accuracy of this classification approach may provide performance optimization and cost reduction for quality improvement, research, and reimbursement tasks reliant on anesthesiology procedure codes.

Machine-learning analysis outperforms conventional statistical models and CT classification systems in predicting 6-month outcomes in pediatric patients sustaining traumatic brain injury.

OBJECTIVEModern surgical planning and prognostication requires the most accurate outcomes data to practice evidence-based medicine. For clinicians treating children following traumatic brain injury (TBI) these data are severely lacking. The first aim of this study was to assess published CT classification systems in the authors' pediatric cohort. A pediatric-specific machine-learning algorithm called an artificial neural network (ANN) was then created that robustly outperformed traditional CT classification systems in predicting TBI outcomes in children.METHODSThe clinical records of children under the age of 18 who suffered a TBI and underwent head CT within 24 hours after TBI (n = 565) were retrospectively reviewed.RESULTS"Favorable" outcome (alive with Glasgow Outcome Scale [GOS] score ≥ 4 at 6 months postinjury, n = 533) and "unfavorable" outcome (death at 6 months or GOS score ≤ 3 at 6 months postinjury, n = 32) were used as the primary outcomes. The area under the receiver operating characteristic (ROC) curve (AUC) was used to delineate the strength of each CT grading system in predicting survival (Helsinki, 0.814; Rotterdam, 0.838; and Marshall, 0.781). The AUC for CT score in predicting GOS score ≤ 3, a measure of overall functionality, was similarly predictive (Helsinki, 0.717; Rotterdam, 0.748; and Marshall, 0.663). An ANN was then constructed that was able to predict 6-month outcomes with profound accuracy (AUC = 0.9462 ± 0.0422).CONCLUSIONSThis study showed that machine-learning can be leveraged to more accurately predict TBI outcomes in children.

Automatic fish sounds classification.

The work presented in this paper focuses on the use of acoustic systems for passive acoustic monitoring of ocean vitality for fish populations. Specifically, it focuses on the use of acoustic systems for passive acoustic monitoring of ocean vitality for fish populations. To this end, various indicators can be used to monitor marine areas such as both the geographical and temporal evolution of fish populations. A discriminative model is built using supervised machine learning (random-forest and support-vector machines). Each acquisition is represented in a feature space, in which the patterns belonging to different semantic classes are as separable as possible. The set of features proposed for describing the acquisitions come from an extensive state of the art in various domains in which classification of acoustic signals is performed, including speech, music, and environmental acoustics. Furthermore, this study proposes to extract features from three representations of the data (time, frequency, and cepstral domains). The proposed classification scheme is tested on real fish sounds recorded on several areas, and achieves 96.9% correct classification compared to 72.5% when using reference state of the art features as descriptors. The classification scheme is also validated on continuous underwater recordings, thereby illustrating that it can be used to both detect and classify fish sounds in operational scenarios.

A survey on modern trainable activation functions.

In neural networks literature, there is a strong interest in identifying and defining activation functions which can improve neural network performance. In recent years there has been a renovated interest in the scientific community in investigating activation functions which can be trained during the learning process, usually referred to as trainable, learnable or adaptable activation functions. They appear to lead to better network performance. Diverse and heterogeneous models of trainable activation function have been proposed in the literature. In this paper, we present a survey of these models. Starting from a discussion on the use of the term "activation function" in literature, we propose a taxonomy of trainable activation functions, highlight common and distinctive proprieties of recent and past models, and discuss main advantages and limitations of this type of approach. We show that many of the proposed approaches are equivalent to adding neuron layers which use fixed (non-trainable) activation functions and some simple local rule that constrains the corresponding weight layers.

Self-augmentation: Generalizing deep networks to unseen classes for few-shot learning.

Few-shot learning aims to classify unseen classes with a few training examples. While recent works have shown that standard mini-batch training with carefully designed training strategies can improve generalization ability for unseen classes, well-known problems in deep networks such as memorizing training statistics have been less explored for few-shot learning. To tackle this issue, we propose self-augmentation that consolidates self-mix and self-distillation. Specifically, we propose a regional dropout technique called self-mix, in which a patch of an image is substituted into other values in the same image. With this dropout effect, we show that the generalization ability of deep networks can be improved as it prevents us from learning specific structures of a dataset. Then, we employ a backbone network that has auxiliary branches with its own classifier to enforce knowledge sharing. This sharing of knowledge forces each branch to learn diverse optimal points during training. Additionally, we present a local representation learner to further exploit a few training examples of unseen classes by generating fake queries and novel weights. Experimental results show that the proposed method outperforms the state-of-the-art methods for prevalent few-shot benchmarks and improves the generalization ability.

Modality independent adversarial network for generalized zero shot image classification.

Zero Shot Learning (ZSL) aims to classify images of unseen target classes by transferring knowledge from source classes through semantic embeddings. The core of ZSL research is to embed both visual representation of object instance and semantic description of object class into a joint latent space and learn cross-modal (visual and semantic) latent representations. However, the learned representations by existing efforts often fail to fully capture the underlying cross-modal semantic consistency, and some of the representations are very similar and less discriminative. To circumvent these issues, in this paper, we propose a novel deep framework, called Modality Independent Adversarial Network (MIANet) for Generalized Zero Shot Learning (GZSL), which is an end-to-end deep architecture with three submodules. First, both visual feature and semantic description are embedded into a latent hyper-spherical space, where two orthogonal constraints are employed to ensure the learned latent representations discriminative. Second, a modality adversarial submodule is employed to make the latent representations independent of modalities to make the shared representations grab more cross-modal high-level semantic information during training. Third, a cross reconstruction submodule is proposed to reconstruct latent representations into the counterparts instead of themselves to make them capture more modality irrelevant information. Comprehensive experiments on five widely used benchmark datasets are conducted on both GZSL and standard ZSL settings, and the results show the effectiveness of our proposed method.

Pattern classification as decision support tool in antipsychotic treatment algorithms.

Pattern classification aims to establish a new approach in personalized treatment. The scope is to tailor treatment on individual characteristics during all phases of care including prevention, diagnosis, treatment, and clinical outcome. In psychotic disorders, this need results from the fact that a third of patients with psychotic symptoms do not respond to antipsychotic treatment and are described as having treatment-resistant disorders. This, in addition to the high variability of treatment responses among patients, enhances the need of applying advanced classification algorithms to identify antipsychotic treatment patterns. This review comprehensively summarizes advancements and challenges of pattern classification in antipsychotic treatment response to date and aims to introduce clinicians and researchers to the challenges of including pattern classification into antipsychotic treatment decision algorithms.

Relationship between Machine-Learning Image Classification of T2-Weighted Intramedullary Hypointensity on 3 Tesla Magnetic Resonance Imaging and Clinical Outcome in Dogs with Severe Spinal Cord Injury.

Early prognostic information in cases of severe spinal cord injury can aid treatment planning and stratification for clinical trials. Analysis of intraparenchymal signal change on magnetic resonance imaging has been suggested to inform outcome prediction in traumatic spinal cord injury. We hypothesized that intraparenchymal T2-weighted hypointensity would be associated with a lower potential for functional recovery and a higher risk of progressive neurological deterioration in dogs with acute, severe, naturally occurring spinal cord injury. Our objectives were to: 1) demonstrate capacity for machine-learning criteria to identify clinically relevant regions of hypointensity and 2) compare clinical outcomes for cases with and without such regions. A total of 95 dogs with complete spinal cord injury were evaluated. An image classification system, based on Speeded-Up Robust Features (SURF), was trained to recognize individual axial T2-weighted slices that contained hypointensity. The presence of such slices in a given transverse series was correlated with a lower chance of functional recovery (odds ratio [OR], 0.08; confidence interval [CI], 0.02-0.38; p < 10-3) and with a higher risk of neurological deterioration (OR, 0.14; 95% CI, 0.05-0.42; p < 10-3). Identification of intraparenchymal T2-weighted hypointensity in severe, naturally occurring spinal cord injury may be assisted by an image classification tool and is correlated with functional recovery.

Adaptive transfer learning for EEG motor imagery classification with deep Convolutional Neural Network.

In recent years, deep learning has emerged as a powerful tool for developing Brain-Computer Interface (BCI) systems. However, for deep learning models trained entirely on the data from a specific individual, the performance increase has only been marginal owing to the limited availability of subject-specific data. To overcome this, many transfer-based approaches have been proposed, in which deep networks are trained using pre-existing data from other subjects and evaluated on new target subjects. This mode of transfer learning however faces the challenge of substantial inter-subject variability in brain data. Addressing this, in this paper, we propose 5 schemes for adaptation of a deep convolutional neural network (CNN) based electroencephalography (EEG)-BCI system for decoding hand motor imagery (MI). Each scheme fine-tunes an extensively trained, pre-trained model and adapt it to enhance the evaluation performance on a target subject. We report the highest subject-independent performance with an average (N=54) accuracy of 84.19% (±9.98%) for two-class motor imagery, while the best accuracy on this dataset is 74.15% (±15.83%) in the literature. Further, we obtain a statistically significant improvement (p=0.005) in classification using the proposed adaptation schemes compared to the baseline subject-independent model.

Named entity recognition in electronic health records using transfer learning bootstrapped Neural Networks.

Neural networks (NNs) have become the state of the art in many machine learning applications, such as image, sound (LeCun et al., 2015) and natural language processing (Young et al., 2017; Linggard et al., 2012). However, the success of NNs remains dependent on the availability of large labelled datasets, such as in the case of electronic health records (EHRs). With scarce data, NNs are unlikely to be able to extract this hidden information with practical accuracy. In this study, we develop an approach that solves these problems for named entity recognition, obtaining 94.6 F1 score in I2B2 2009 Medical Extraction Challenge (Uzuner et al., 2010), 4.3 above the architecture that won the competition. To achieve this, we bootstrap our NN models through transfer learning by pretraining word embeddings on a secondary task performed on a large pool of unannotated EHRs and using the output embeddings as a foundation of a range of NN architectures. Beyond the official I2B2 challenge, we further achieve 82.4 F1 on extracting relationships between medical terms using attention-based seq2seq models bootstrapped in the same manner.

Decoding of single-trial EEG reveals unique states of functional brain connectivity that drive rapid speech categorization decisions.

OBJECTIVE: Categorical perception (CP) is an inherent property of speech perception. The response time (RT) of listeners' perceptual speech identification is highly sensitive to individual differences. While the neural correlates of CP have been well studied in terms of the regional contributions of the brain to behavior, functional connectivity patterns that signify individual differences in listeners' speed (RT) for speech categorization is less clear. In this study, we introduce a novel approach to address these questions. APPROACH: We applied several computational approaches to the EEG, including graph mining, machine learning (i.e., support vector machine), and stability selection to investigate the unique brain states (functional neural connectivity) that predict the speed of listeners' behavioral decisions. MAIN RESULTS: We infer that (i) the listeners' perceptual speed is directly related to dynamic variations in their brain connectomics, (ii) global network assortativity and efficiency distinguished fast, medium, and slow RTs, (iii) the functional network underlying speeded decisions increases in negative assortativity (i.e., became disassortative) for slower RTs, (iv) slower categorical speech decisions cause excessive use of neural resources and more aberrant information flow within the CP circuitry, (v) slower responders tended to utilize functional brain networks excessively (or inappropriately) whereas fast responders (with lower global efficiency) utilized the same neural pathways but with more restricted organization. SIGNIFICANCE: Findings show that neural classifiers (SVM) coupled with stability selection correctly classify behavioral RTs from functional connectivity alone with over 92% accuracy (AUC = 0.9). Our results corroborate previous studies by supporting the engagement of similar temporal (STG), parietal, motor, and prefrontal regions in CP using an entirely data-driven approach.

Characterization of specific and distinct patient types in clinical trials of acute schizophrenia using an uncorrelated PANSS score matrix transform (UPSM).

Understanding the specificity of symptom change in schizophrenia can facilitate the evaluation antipsychotic efficacy for different symptom domains. Previous work identified a transform of PANSS using an uncorrelated PANSS score matrix (UPSM) to reduce pseudospecificity among symptom domains during clinical trials of schizophrenia. Here we used UPSM-transformed factor scores to identify 5 distinct patient types, each having elevated and specific severity among each of 5 symptom domains. Subjects from placebo-controlled clinical trials of acute schizophrenia were clustered (baseline) and classified (post-baseline) by a machine-learning algorithm. At baseline, all 5 patient types were similar in PANSS total score. Post-baseline, subjects' memberships among the 5 UPSM patient types were relatively stable over treatment duration and were relatively insensitive to overall improvements in symptoms, in contrast to other methods based on untransformed PANSS items. Using UPSM-transformed PANSS, drug treatment effect sizes versus placebo were doubly-dissociated for specificity across symptom domains and within specific patient types. This approach illustrates how broader clinical trial populations can nevertheless be utilized to characterize the specificity of new mechanisms across the dimensions of schizophrenia psychopathology.

Machine learning for clinical decision support in infectious diseases: a narrative review of current applications.

BACKGROUND: Machine learning (ML) is a growing field in medicine. This narrative review describes the current body of literature on ML for clinical decision support in infectious diseases (ID). OBJECTIVES: We aim to inform clinicians about the use of ML for diagnosis, classification, outcome prediction and antimicrobial management in ID. SOURCES: References for this review were identified through searches of MEDLINE/PubMed, EMBASE, Google Scholar, biorXiv, ACM Digital Library, arXiV and IEEE Xplore Digital Library up to July 2019. CONTENT: We found 60 unique ML-clinical decision support systems (ML-CDSS) aiming to assist ID clinicians. Overall, 37 (62%) focused on bacterial infections, 10 (17%) on viral infections, nine (15%) on tuberculosis and four (7%) on any kind of infection. Among them, 20 (33%) addressed the diagnosis of infection, 18 (30%) the prediction, early detection or stratification of sepsis, 13 (22%) the prediction of treatment response, four (7%) the prediction of antibiotic resistance, three (5%) the choice of antibiotic regimen and two (3%) the choice of a combination antiretroviral therapy. The ML-CDSS were developed for intensive care units (n = 24, 40%), ID consultation (n = 15, 25%), medical or surgical wards (n = 13, 20%), emergency department (n = 4, 7%), primary care (n = 3, 5%) and antimicrobial stewardship (n = 1, 2%). Fifty-three ML-CDSS (88%) were developed using data from high-income countries and seven (12%) with data from low- and middle-income countries (LMIC). The evaluation of ML-CDSS was limited to measures of performance (e.g. sensitivity, specificity) for 57 ML-CDSS (95%) and included data in clinical practice for three (5%). IMPLICATIONS: Considering comprehensive patient data from socioeconomically diverse healthcare settings, including primary care and LMICs, may improve the ability of ML-CDSS to suggest decisions adapted to various clinical contexts. Currents gaps identified in the evaluation of ML-CDSS must also be addressed in order to know the potential impact of such tools for clinicians and patients.

Utility of MemTrax and Machine Learning Modeling in Classification of Mild Cognitive Impairment.

BACKGROUND: The widespread incidence and prevalence of Alzheimer's disease and mild cognitive impairment (MCI) has prompted an urgent call for research to validate early detection cognitive screening and assessment. OBJECTIVE: Our primary research aim was to determine if selected MemTrax performance metrics and relevant demographics and health profile characteristics can be effectively utilized in predictive models developed with machine learning to classify cognitive health (normal versus MCI), as would be indicated by the Montreal Cognitive Assessment (MoCA). METHODS: We conducted a cross-sectional study on 259 neurology, memory clinic, and internal medicine adult patients recruited from two hospitals in China. Each patient was given the Chinese-language MoCA and self-administered the continuous recognition MemTrax online episodic memory test on the same day. Predictive classification models were built using machine learning with 10-fold cross validation, and model performance was measured using Area Under the Receiver Operating Characteristic Curve (AUC). Models were built using two MemTrax performance metrics (percent correct, response time), along with the eight common demographic and personal history features. RESULTS: Comparing the learners across selected combinations of MoCA scores and thresholds, Naïve Bayes was generally the top-performing learner with an overall classification performance of 0.9093. Further, among the top three learners, MemTrax-based classification performance overall was superior using just the top-ranked four features (0.9119) compared to using all 10 common features (0.8999). CONCLUSION: MemTrax performance can be effectively utilized in a machine learning classification predictive model screening application for detecting early stage cognitive impairment.