Transport proteins promoting Escherichia coli pathogenesis.

Escherichia coli is a genetically diverse species infecting hundreds of millions of people worldwide annually. We examined seven well-characterized E. coli pathogens causing urinary tract infections, gastroenteritis, pyelonephritis and haemorrhagic colitis. Their transport proteins were identified and compared with each other and a non-pathogenic E. coli K12 strain to identify transport proteins related to pathogenesis. Each pathogen possesses a unique set of protein secretion systems for export to the cell surface or for injecting effector proteins into host cells. Pathogens have increased numbers of iron siderophore receptors and ABC iron uptake transporters, but the numbers and types of low-affinity secondary iron carriers were uniform in all strains. The presence of outer membrane iron complex receptors and high-affinity ABC iron uptake systems correlated, suggesting co-evolution. Each pathovar encodes a different set of pore-forming toxins and virulence-related outer membrane proteins lacking in K12. Intracellular pathogens proved to have a characteristically distinctive set of nutrient uptake porters, different from those of extracellular pathogens. The results presented in this report provide information about transport systems relevant to various types of E. coli pathogenesis that can be exploited in future basic and applied studies.

Automatic Detection of Alzheimer's Disease Using Spontaneous Speech Only.

Alzheimer's disease (AD) is a neurodegenerative syndrome which affects tens of millions of elders worldwide. Although there is no treatment currently available, early recognition can improve the lives of people with AD and their caretakers and families. To find a cost-effective and easy-to-use method for dementia detection and address the dementia classification task of InterSpeech 2021 ADReSSo (Alzheimer's' Dementia Recognition through Spontaneous Speech only) challenge, we conduct a systematic comparison of approaches to detection of cognitive impairment based on spontaneous speech. We investigated the characteristics of acoustic modality and linguistic modality directly based on the audio recordings of narrative speech, and explored a variety of modality fusion strategies. With an ensemble over top-10 classifiers on the training set, we achieved an accuracy of 81.69% compared to the baseline of 78.87% on the test set. The results suggest that although transcription errors will be introduced through automatic speech recognition, integrating textual information generally improves classification performance. Besides, ensemble methods can boost both the accuracy and the robustness of models.

The Joint Effects of Acoustic and Linguistic Markers for Early Identification of Mild Cognitive Impairment.

In recent years, behavioral markers such as spoken language and lexical preferences have been studied in the early detection of mild cognitive impairment (MCI) using conversations. While the combination of linguistic and acoustic signals have been shown to be effective in detecting MCI, they have generally been restricted to structured conversations in which the interviewee responds to fixed prompts. In this study, we show that linguistic and acoustic features can be combined synergistically to identify MCI in semi-structured conversations. Using conversational data from an on-going clinical trial (Clinicaltrials.gov: NCT02871921), we find that the combination of linguistic and acoustic features on semi-structured conversations achieves a mean AUC of 82.7, significantly (p < 0.01) out-performing linguistic-only (74.9 mean AUC) or acoustic-only (65.0 mean AUC) detections on hold-out data. Additionally, features (linguistic, acoustic and combination) obtained from semi-structured conversations outperform their counterparts obtained from structured weekly conversations in identifying MCI. Some linguistic categories are significantly better at predicting MCI status (e.g., death, home) than others.

Scalable diagnostic screening of mild cognitive impairment using AI dialogue agent.

The search for early biomarkers of mild cognitive impairment (MCI) has been central to the Alzheimer's Disease (AD) and dementia research community in recent years. To identify MCI status at the earliest possible point, recent studies have shown that linguistic markers such as word choice, utterance and sentence structures can potentially serve as preclinical behavioral markers. Here we present an adaptive dialogue algorithm (an AI-enabled dialogue agent) to identify sequences of questions (a dialogue policy) that distinguish MCI from normal (NL) cognitive status. Our AI agent adapts its questioning strategy based on the user's previous responses to reach an individualized conversational strategy per user. Because the AI agent is adaptive and scales favorably with additional data, our method provides a potential avenue for large-scale preclinical screening of neurocognitive decline as a new digital biomarker, as well as longitudinal tracking of aging patterns in the outpatient setting.

Retaining Privileged Information for Multi-Task Learning.

Knowledge transfer has been of great interest in current machine learning research, as many have speculated its importance in modeling the human ability to rapidly generalize learned models to new scenarios. Particularly in cases where training samples are limited, knowledge transfer shows improvement on both the learning speed and generalization performance of related tasks. Recently, Learning Using Privileged Information (LUPI) has presented a new direction in knowledge transfer by modeling the transfer of prior knowledge as a Teacher-Student interaction process. Under LUPI, a Teacher model uses Privileged Information (PI) that is only available at training time to improve the sample complexity required to train a Student learner for a given task. In this work, we present a LUPI formulation that allows privileged information to be retained in a multi-task learning setting. We propose a novel feature matching algorithm that projects samples from the original feature space and the privilege information space into a joint latent space in a way that informs similarity between training samples. Our experiments show that useful knowledge from PI is maintained in the latent space and greatly improves the sample efficiency of other related learning tasks. We also provide an analysis of sample complexity of the proposed LUPI method, which under some favorable assumptions can achieve a greater sample efficiency than brute force methods.

MetaPred: Meta-Learning for Clinical Risk Prediction with Limited Patient Electronic Health Records.

In recent years, large amounts of health data, such as patient Electronic Health Records (EHR), are becoming readily available. This provides an unprecedented opportunity for knowledge discovery and data mining algorithms to dig insights from them, which can, later on, be helpful to the improvement of the quality of care delivery. Predictive modeling of clinical risks, including in-hospital mortality, hospital readmission, chronic disease onset, condition exacerbation, etc., from patient EHR, is one of the health data analytic problems that attract lots of the interests. The reason is not only because the problem is important in clinical settings, but also is challenging when working with EHR such as sparsity, irregularity, temporality, etc. Different from applications in other domains such as computer vision and natural language processing, the data samples in medicine (patients) are relatively limited, which creates lots of troubles for building effective predictive models, especially for complicated ones such as deep learning. In this paper, we propose MetaPred, a meta-learning framework for clinical risk prediction from longitudinal patient EHR. In particular, in order to predict the target risk with limited data samples, we train a meta-learner from a set of related risk prediction tasks which learns how a good predictor is trained. The meta-learned can then be directly used in target risk prediction, and the limited available samples in the target domain can be used for further fine-tuning the model performance. The effectiveness of MetaPred is tested on a real patient EHR repository from Oregon Health & Science University. We are able to demonstrate that with Convolutional Neural Network (CNN) and Recurrent Neural Network (RNN) as base predictors, MetaPred can achieve much better performance for predicting target risk with low resources comparing with the predictor trained on the limited samples available for this risk alone.

Predictive modeling in urgent care: a comparative study of machine learning approaches.

OBJECTIVE: The growing availability of rich clinical data such as patients' electronic health records provide great opportunities to address a broad range of real-world questions in medicine. At the same time, artificial intelligence and machine learning (ML)-based approaches have shown great premise on extracting insights from those data and helping with various clinical problems. The goal of this study is to conduct a systematic comparative study of different ML algorithms for several predictive modeling problems in urgent care. DESIGN: We assess the performance of 4 benchmark prediction tasks (eg mortality and prediction, differential diagnostics, and disease marker discovery) using medical histories, physiological time-series, and demographics data from the Medical Information Mart for Intensive Care (MIMIC-III) database. MEASUREMENTS: For each given task, performance was estimated using standard measures including the area under the receiver operating characteristic (AUC) curve, F-1 score, sensitivity, and specificity. Microaveraged AUC was used for multiclass classification models. RESULTS AND DISCUSSION: Our results suggest that recurrent neural networks show the most promise in mortality prediction where temporal patterns in physiologic features alone can capture in-hospital mortality risk (AUC > 0.90). Temporal models did not provide additional benefit compared to deep models in differential diagnostics. When comparing the training-testing behaviors of readmission and mortality models, we illustrate that readmission risk may be independent of patient stability at discharge. We also introduce a multiclass prediction scheme for length of stay which preserves sensitivity and AUC with outliers of increasing duration despite decrease in sample size.

Chronology of illness in dual diagnosis heroin addicts: The role of mood disorders.

BACKGROUND: Recent celebrity deaths have been widely reported in the media and turned the public attention to the coexistence of mood, psychiatric and substance-abuse disorders. These tragic and untimely deaths motivated us to examine the scientific and clinical data, including our own work in this area. The self-medication hypothesis states that individuals with psychiatric illness tend to use heroin to alleviate their symptoms. This study examined the correlations between heroin use, mood and psychiatric disorders, and their chronology in the context of dual diagnosis. METHODS: Out of 506 dual diagnosed heroin addicts, 362 patients were implicated in heroin abuse with an onset of at least one year prior to the associated mental disorder (HER-PR), and 144 patients were diagnosed of mental illness at least one year prior to the associated onset of heroin use disorder (MI-PR). The retrospective cross-sectional analysis of the two groups compared their demographic, clinical and diagnostic characteristics at univariate and multivariate levels. RESULTS: Dual diagnosis heroin addicts whose heroin dependences existed one year prior to their diagnoses (HER-PR) reported more frequent somatic comorbidity (p≤0.001), less major problems at work (p=0.003), more legal problems (p=0.004) and more failed treatment for their heroin dependence (p<0.001) in the past. More than 2/3 reached the third stage of heroin addiction (p=<0.001). Their length of dependence was longer (p=0.004). HER-PR patients were diagnosed more frequently as affected by mood disorders and less frequently as affected by psychosis (p=0.004). At the multivariate level, HER-PR patients were characterized by having reached stage 3 of heroin dependence (OR=2.45), diagnosis of mood disorder (OR=2.25), unsuccessful treatment (OR=2.07) and low education (OR=1.79). LIMITATIONS: The main limitation is its retrospective nature. Nonetheless, it does shed light on what needs to be done from a clinical and public health perspective and especially prevention. CONCLUSIONS: The data emerging from this study, does not allow us to determine a causal relation between heroin use and mental illness onset. However, this data, even if requiring longitudinal studies, suggest that self-medication theory, in these patients, can be applied only for chronic psychoses, but should not be applied to patients with mood disorders using heroin.

Resveratrol induces cell death and inhibits human herpesvirus 8 replication in primary effusion lymphoma cells.

Resveratrol (3,4',5-trihydroxy-trans-stilbene) has been reported to inhibit proliferation of various cancer cells. However, the effects of resveratrol on the human herpesvirus 8 (HHV8) harboring primary effusion lymphoma (PEL) cells remains unclear. The anti-proliferation effects and possible mechanisms of resveratrol in the HHV8 harboring PEL cells were examined in this study. Results showed that resveratrol induced caspase-3 activation and the formation of acidic vacuoles in the HHV8 harboring PEL cells, indicating resveratrol treatment could cause apoptosis and autophagy in PEL cells. In addition, resveratrol treatment increased ROS generation but did not lead to HHV8 reactivation. ROS scavenger (N-acetyl cysteine, NAC) could attenuate both the resveratrol induced caspase-3 activity and the formation of acidic vacuoles, but failed to attenuate resveratrol induced PEL cell death. Caspase inhibitor, autophagy inhibitors and necroptosis inhibitor could not block resveratrol induced PEL cell death. Moreover, resveratrol disrupted HHV8 latent infection, inhibited HHV8 lytic gene expression and decreased virus progeny production. Overexpression of HHV8-encoded viral FLICE inhibitory protein (vFLIP) could partially block resveratrol induced cell death in PEL cells. These data suggest that resveratrol-induced cell death in PEL cells may be mediated by disruption of HHV8 replication. Resveratrol may be a potential anti-HHV8 drug and an effective treatment for HHV8-related tumors.