ABSTRACT
BACKGROUND: Recommender systems have great potential in mental health care to personalize self-guided content for patients, allowing them to supplement their mental health treatment in a scalable way. OBJECTIVE: In this paper, we describe and evaluate 2 knowledge-based content recommendation systems as parts of Ginger, an on-demand mental health platform, to bolster engagement in self-guided mental health content. METHODS: We developed two algorithms to provide content recommendations in the Ginger mental health smartphone app: (1) one that uses users' responses to app onboarding questions to recommend content cards and (2) one that uses the semantic similarity between the transcript of a coaching conversation and the description of content cards to make recommendations after every session. As a measure of success for these recommendation algorithms, we examined the relevance of content cards to users' conversations with their coach and completion rates of selected content within the app measured over 14,018 users. RESULTS: In a real-world setting, content consumed in the recommendations section (or "Explore" in the app) had the highest completion rates (3353/7871, 42.6%) compared to other sections of the app, which had an average completion rate of 37.35% (21,982/58,614; P<.001). Within the app's recommendations section, conversation-based content recommendations had 11.4% (1108/2364) higher completion rates per card than onboarding response-based recommendations (1712/4067; P=.003) and 26.1% higher than random recommendations (534/1440; P=.005). Studied via subject matter experts' annotations, conversation-based recommendations had a 16.1% higher relevance rate for the top 5 recommended cards, averaged across sessions of varying lengths, compared to a random control (110 conversational sessions). Finally, it was observed that both age and gender variables were sensitive to different recommendation methods, with responsiveness to personalized recommendations being higher if the users were older than 35 years or identified as male. CONCLUSIONS: Recommender systems can help scale and supplement digital mental health care with personalized content and self-care recommendations. Onboarding-based recommendations are ideal for "cold starting" the process of recommending content for new users and users that tend to use the app just for content but not for therapy or coaching. The conversation-based recommendation algorithm allows for dynamic recommendations based on information gathered during coaching sessions, which is a critical capability, given the changing nature of mental health needs during treatment. The proposed algorithms are just one step toward the direction of outcome-driven personalization in mental health. Our future work will involve a robust causal evaluation of these algorithms using randomized controlled trials, along with consumer feedback-driven improvement of these algorithms, to drive better clinical outcomes.
ABSTRACT
In light of the National Institute of Mental Health (NIMH)'s Research Domain Criteria (RDoC), the advent of functional neuroimaging, novel technologies and methods provide new opportunities to develop precise and personalized prognosis and diagnosis of mental disorders. Machine learning (ML) and artificial intelligence (AI) technologies are playing an increasingly critical role in the new era of precision psychiatry. Combining ML/AI with neuromodulation technologies can potentially provide explainable solutions in clinical practice and effective therapeutic treatment. Advanced wearable and mobile technologies also call for the new role of ML/AI for digital phenotyping in mobile mental health. In this review, we provide a comprehensive review of ML methodologies and applications by combining neuroimaging, neuromodulation, and advanced mobile technologies in psychiatry practice. We further review the role of ML in molecular phenotyping and cross-species biomarker identification in precision psychiatry. We also discuss explainable AI (XAI) and neuromodulation in a closed human-in-the-loop manner and highlight the ML potential in multi-media information extraction and multi-modal data fusion. Finally, we discuss conceptual and practical challenges in precision psychiatry and highlight ML opportunities in future research.
ABSTRACT
BACKGROUND: Psychiatry is nearly entirely reliant on patient self-reporting, and there are few objective and reliable tests or sources of collateral information available to help diagnostic and assessment procedures. Technology offers opportunities to collect objective digital data to complement patient experience and facilitate more informed treatment decisions. OBJECTIVE: We aimed to develop computational algorithms based on internet search activity designed to support diagnostic procedures and relapse identification in individuals with schizophrenia spectrum disorders. METHODS: We extracted 32,733 time-stamped search queries across 42 participants with schizophrenia spectrum disorders and 74 healthy volunteers between the ages of 15 and 35 (mean 24.4 years, 44.0% male), and built machine-learning diagnostic and relapse classifiers utilizing the timing, frequency, and content of online search activity. RESULTS: Classifiers predicted a diagnosis of schizophrenia spectrum disorders with an area under the curve value of 0.74 and predicted a psychotic relapse in individuals with schizophrenia spectrum disorders with an area under the curve of 0.71. Compared with healthy participants, those with schizophrenia spectrum disorders made fewer searches and their searches consisted of fewer words. Prior to a relapse hospitalization, participants with schizophrenia spectrum disorders were more likely to use words related to hearing, perception, and anger, and were less likely to use words related to health. CONCLUSIONS: Online search activity holds promise for gathering objective and easily accessed indicators of psychiatric symptoms. Utilizing search activity as collateral behavioral health information would represent a major advancement in efforts to capitalize on objective digital data to improve mental health monitoring.
ABSTRACT
(1) One strategy to improve the outcome of orthopedic implants is to use porous implants with the addition of a coating with an antibacterial biomolecule. In this study, we aimed to produce and test the biocompatibility, the osteopromotive (both under normal conditions and under a bacterial challenge with lipopolysaccharide (LPS)) and antibacterial activities of a porous Ti-6Al-4V implant coated with the flavonoid quercitrin in vitro. (2) Porous Ti-6Al-4V implants were produced by 3D printing and further functionalized with quercitrin by wet chemistry. Implants were characterized in terms of porosity and mechanical testing, and the coating with quercitrin by fluorescence staining. Implant biocompatibility and bioactivity was tested using MC3T3-E1 preosteoblasts by analyzing cytotoxicity, cell adhesion, osteocalcin production, and alkaline phosphatase (ALP) activity under control and under bacterial challenging conditions using lipopolysaccharide (LPS). Finally, the antibacterial properties of the implants were studied using Staphylococcus epidermidis by measuring bacterial viability and adhesion. (3) Porous implants showed pore size of about 500 µm and a porosity of 52%. The coating was homogeneous over all the 3D surface and did not alter the mechanical properties of the Young modulus. Quercitrin-coated implants showed higher biocompatibility, cell adhesion, and osteocalcin production compared with control implants. Moreover, higher ALP activity was observed for the quercitrin group under both normal and bacterial challenging conditions. Finally, S. epidermidis live/dead ratio and adhesion after 4 h of incubation was lower on quercitrin implants compared with the control. (4) Quercitrin-functionalized porous Ti-6Al-4V implants present a great potential as an orthopedic porous implant that decreases bacterial adhesion and viability while promoting bone cell growth and differentiation.
Subject(s)
Chronic Pain/diagnosis , Chronic Pain/therapy , Pain Management/methods , Sleep/physiology , Acoustics , Algorithms , Animals , Biomarkers/metabolism , Diagnosis, Computer-Assisted , Electroencephalography , Hyperalgesia/physiopathology , Inflammation , Male , Neural Networks, Computer , Pattern Recognition, Automated , Rats , Rats, Sprague-Dawley , Sleep StagesABSTRACT
PURPOSE: Biomarkers for disease-specific survival (DSS) in early-stage melanoma are needed to select patients for adjuvant immunotherapy and accelerate clinical trial design. We present a pathology-based computational method using a deep neural network architecture for DSS prediction. EXPERIMENTAL DESIGN: The model was trained on 108 patients from four institutions and tested on 104 patients from Yale School of Medicine (YSM, New Haven, CT). A receiver operating characteristic (ROC) curve was generated on the basis of vote aggregation of individual image sequences, an optimized cutoff was selected, and the computational model was tested on a third independent population of 51 patients from Geisinger Health Systems (GHS). RESULTS: Area under the curve (AUC) in the YSM patients was 0.905 (P < 0.0001). AUC in the GHS patients was 0.880 (P < 0.0001). Using the cutoff selected in the YSM cohort, the computational model predicted DSS in the GHS cohort based on Kaplan-Meier (KM) analysis (P < 0.0001). CONCLUSIONS: The novel method presented is applicable to digital images, obviating the need for sample shipment and manipulation and representing a practical advance over current genetic and IHC-based methods.
Subject(s)
Deep Learning/standards , Image Processing, Computer-Assisted/standards , Melanoma/mortality , Melanoma/pathology , Neoplasm Recurrence, Local/mortality , Neoplasm Recurrence, Local/pathology , Staining and Labeling/methods , Adult , Aged , Aged, 80 and over , Algorithms , Area Under Curve , Biopsy/methods , Disease Progression , Female , Follow-Up Studies , Humans , Image Processing, Computer-Assisted/methods , Male , Middle Aged , Neural Networks, Computer , Retrospective Studies , Risk Factors , Survival Rate , Young AdultABSTRACT
Pain is a complex multidimensional experience encompassing sensory-discriminative, affective-motivational and cognitive-emotional components mediated by different neural mechanisms. Investigations of neurophysiological signals from simultaneous recordings of two or more cortical circuits may reveal important circuit mechanisms on cortical pain processing. The anterior cingulate cortex (ACC) and primary somatosensory cortex (S1) represent two most important cortical circuits related to sensory and affective processing of pain. Here, we recorded in vivo extracellular activity of the ACC and S1 simultaneously from male adult Sprague-Dale rats (n = 5), while repetitive noxious laser stimulations were delivered to animalÕs hindpaw during pain experiments. We identified spontaneous pain-like events based on stereotyped pain behaviors in rats. We further conducted systematic analyses of spike and local field potential (LFP) recordings from both ACC and S1 during evoked and spontaneous pain episodes. From LFP recordings, we found stronger phase-amplitude coupling (theta phase vs. gamma amplitude) in the S1 than the ACC (n = 10 sessions), in both evoked (p = 0.058) and spontaneous pain-like behaviors (p = 0.017, paired signed rank test). In addition, pain-modulated ACC and S1 neuronal firing correlated with the amplitude of stimulus-induced event-related potentials (ERPs) during evoked pain episodes. We further designed statistical and machine learning methods to detect pain signals by integrating ACC and S1 ensemble spikes and LFPs. Together, these results reveal differential coding roles between the ACC and S1 in cortical pain processing, as well as point to distinct neural mechanisms between evoked and putative spontaneous pain at both LFP and cellular levels.
ABSTRACT
OBJECTIVE: Sleep spindles have been implicated in memory consolidation and synaptic plasticity during NREM sleep. Detection accuracy and latency in automatic spindle detection are critical for real-time applications. APPROACH: Here we propose a novel deep learning strategy (SpindleNet) to detect sleep spindles based on a single EEG channel. While the majority of spindle detection methods are used for off-line applications, our method is well suited for online applications. MAIN RESULTS: Compared with other spindle detection methods, SpindleNet achieves superior detection accuracy and speed, as demonstrated in two publicly available expert-validated EEG sleep spindle datasets. Our real-time detection of spindle onset achieves detection latencies of 150-350 ms (~two-three spindle cycles) and retains excellent performance under low EEG sampling frequencies and low signal-to-noise ratios. SpindleNet has good generalization across different sleep datasets from various subject groups of different ages and species. SIGNIFICANCE: SpindleNet is ultra-fast and scalable to multichannel EEG recordings, with an accuracy level comparable to human experts, making it appealing for long-term sleep monitoring and closed-loop neuroscience experiments.
Subject(s)
Computer Systems , Deep Learning , Neural Networks, Computer , Sleep Stages/physiology , Adolescent , Adult , Aged , Cohort Studies , Computer Systems/statistics & numerical data , Databases, Factual/statistics & numerical data , Deep Learning/statistics & numerical data , Female , Humans , Male , Middle Aged , Young AdultABSTRACT
Noninvasive transcranial brain stimulation has been widely used in experimental and clinical applications to perturb the brain activity, aiming at promoting synaptic plasticity or enhancing functional connectivity within targeted brain regions. However, there are different types of neurostimulations and various choices of stimulation parameters; how these choices influence the intermediate neurophysiological effects and brain connectivity remain incompletely understood. We propose several quantitative methods to investigate the brain connectivity of an epileptic patient before and after transcranial alternating/direct current stimulation (tACS/tDCS). The neuro-feedback derived from our analyses may provide useful cues for the effectiveness of neurostimulation.
Subject(s)
Brain , Transcranial Direct Current Stimulation , Brain Mapping , Humans , Neuronal Plasticity , NeurophysiologyABSTRACT
BACKGROUND: India's corporate sector has grown steadily over the past decade, and it is providing a lot of work opportunities to Indian youth. Around 20% of employees in the corporate sector in India smoke cigarettes. In general, addictive behaviors including smoking are associated with certain personality dimensions. Hence, we conducted a study with the aims to assess the level of nicotine dependence in tobacco smokers (working in corporate sector), study their personality profile, and association of their personality traits with continuing smoking behavior. MATERIALS AND METHODS: The study proposal along with its intended aims and objectives was cleared by the Institutional Ethical Review Board. It was a cross-sectional study. We used FTND for level of nicotine dependence and NEO FFI 3 for personality profile along with a structured proforma. RESULTS: Most of the clients were of very low to low level of nicotine dependence. As high as 40% of the clients did not even attempt to quit smoking, most common reason for attempt at quitting was health concerns. Major causes of relapse were friends, people at workplace, and nature of work. Clients were high on neuroticism, average on extraversion and openness, and low on agreeableness and conscientiousness. Neuroticism was significantly associated with the level of nicotine dependence. Extraversion and openness were associated with health concerns, while agreeableness and conscientiousness were associated with social factors as a reason to quit. Extraversion and agreeableness were associated with occupational factors and social factors as reasons to relapse. CONCLUSION: Understanding one's personality would be helpful to identify health-enhancing (which help to attempt at quitting) and health-destructive (which were responsible for relapse) behaviors. This can further help in framing interventions that particularly target these personality traits and behaviors.
ABSTRACT
WHAT WE ALREADY KNOW ABOUT THIS TOPIC: WHAT THIS ARTICLE TELLS US THAT IS NEW: BACKGROUND:: Hypotension is a risk factor for adverse perioperative outcomes. Machine-learning methods allow large amounts of data for development of robust predictive analytics. The authors hypothesized that machine-learning methods can provide prediction for the risk of postinduction hypotension. METHODS: Data was extracted from the electronic health record of a single quaternary care center from November 2015 to May 2016 for patients over age 12 that underwent general anesthesia, without procedure exclusions. Multiple supervised machine-learning classification techniques were attempted, with postinduction hypotension (mean arterial pressure less than 55 mmHg within 10 min of induction by any measurement) as primary outcome, and preoperative medications, medical comorbidities, induction medications, and intraoperative vital signs as features. Discrimination was assessed using cross-validated area under the receiver operating characteristic curve. The best performing model was tuned and final performance assessed using split-set validation. RESULTS: Out of 13,323 cases, 1,185 (8.9%) experienced postinduction hypotension. Area under the receiver operating characteristic curve using logistic regression was 0.71 (95% CI, 0.70 to 0.72), support vector machines was 0.63 (95% CI, 0.58 to 0.60), naive Bayes was 0.69 (95% CI, 0.67 to 0.69), k-nearest neighbor was 0.64 (95% CI, 0.63 to 0.65), linear discriminant analysis was 0.72 (95% CI, 0.71 to 0.73), random forest was 0.74 (95% CI, 0.73 to 0.75), neural nets 0.71 (95% CI, 0.69 to 0.71), and gradient boosting machine 0.76 (95% CI, 0.75 to 0.77). Test set area for the gradient boosting machine was 0.74 (95% CI, 0.72 to 0.77). CONCLUSIONS: The success of this technique in predicting postinduction hypotension demonstrates feasibility of machine-learning models for predictive analytics in the field of anesthesiology, with performance dependent on model selection and appropriate tuning.
Subject(s)
Anesthesia, General/adverse effects , Hypotension/diagnosis , Postoperative Complications/diagnosis , Supervised Machine Learning , Adult , Aged , Female , Humans , Hypotension/epidemiology , Hypotension/physiopathology , Male , Middle Aged , Postoperative Complications/epidemiology , Postoperative Complications/physiopathology , Predictive Value of TestsABSTRACT
Pain is a complex sensory and affective experience. The current definition for pain relies on verbal reports in clinical settings and behavioral assays in animal models. These definitions can be subjective and do not take into consideration signals in the neural system. Local field potentials (LFPs) represent summed electrical currents from multiple neurons in a defined brain area. Although single neuronal spike activity has been shown to modulate the acute pain, it is not yet clear how ensemble activities in the form of LFPs can be used to decode the precise timing and intensity of pain. The anterior cingulate cortex (ACC) is known to play a role in the affective-aversive component of pain in human and animal studies. Few studies, however, have examined how neural activities in the ACC can be used to interpret or predict acute noxious inputs. Here, we recorded in vivo extracellular activity in the ACC from freely behaving rats after stimulus with non-noxious, low-intensity noxious, and high-intensity noxious stimuli, both in the absence and chronic pain. Using a supervised machine learning classifier with selected LFP features, we predicted the intensity and the onset of acute nociceptive signals with high degree of precision. These results suggest the potential to use LFPs to decode acute pain.
Subject(s)
Action Potentials/physiology , Acute Pain/physiopathology , Neurons/physiology , Animals , Male , Rats , Rats, Sprague-Dawley , Severity of Illness IndexABSTRACT
Exercise is increasingly being used as a treatment for alcohol use disorders (AUD), but the interactive effects of alcohol and exercise on the brain remain largely unexplored. Alcohol damages the brain, in part by altering glial functioning. In contrast, exercise promotes glial health and plasticity. In the present study, we investigated whether binge alcohol would attenuate the effects of subsequent exercise on glia. We focused on the medial prefrontal cortex (mPFC), an alcohol-vulnerable region that also undergoes neuroplastic changes in response to exercise. Adult female Long-Evans rats were gavaged with ethanol (25% w/v) every 8h for 4days. Control animals received an isocaloric, non-alcohol diet. After 7days of abstinence, rats remained sedentary or exercised for 4weeks. Immunofluorescence was then used to label microglia, astrocytes, and neurons in serial tissue sections through the mPFC. Confocal microscope images were processed using FARSIGHT, a computational image analysis toolkit capable of automated analysis of cell number and morphology. We found that exercise increased the number of microglia in the mPFC in control animals. Binged animals that exercised, however, had significantly fewer microglia. Furthermore, computational arbor analytics revealed that the binged animals (regardless of exercise) had microglia with thicker, shorter arbors and significantly less branching, suggestive of partial activation. We found no changes in the number or morphology of mPFC astrocytes. We conclude that binge alcohol exerts a prolonged effect on morphology of mPFC microglia and limits the capacity of exercise to increase their numbers.
Subject(s)
Binge Drinking/physiopathology , Microglia/physiology , Motor Activity/physiology , Neuronal Plasticity/physiology , Prefrontal Cortex/physiopathology , Animals , Astrocytes/drug effects , Astrocytes/pathology , Astrocytes/physiology , Automation, Laboratory , Binge Drinking/pathology , Binge Drinking/therapy , Cell Count , Central Nervous System Depressants/toxicity , Disease Models, Animal , Ethanol/toxicity , Exercise Therapy , Female , Fluorescent Antibody Technique , Image Processing, Computer-Assisted , Microglia/drug effects , Microglia/pathology , Microscopy, Confocal , Neuronal Plasticity/drug effects , Neurons/drug effects , Neurons/pathology , Neurons/physiology , Prefrontal Cortex/drug effects , Prefrontal Cortex/pathology , Random Allocation , Rats, Long-Evans , Sedentary BehaviorABSTRACT
A 33-year-old multidrug-resistant tuberculosis female patient diagnosed as cycloserine-induced psychosis developed several neuroleptic side effects such as extrapyramidal reaction, neuroleptic malignant syndrome, and drug-induced parkinsonism while she was being treated with initially haloperidol and then olanzapine over a period of 2 months. Patient's antipsychotic medications were withdrawn, and treatment with bromocriptine showed prompt recovery. The multiple neurological adverse effects which the patient developed had implications on the management of the complications as well as her illness.
Subject(s)
Antibiotics, Antitubercular/adverse effects , Antipsychotic Agents/adverse effects , Cycloserine/adverse effects , Neuroleptic Malignant Syndrome/complications , Psychoses, Substance-Induced/complications , Psychoses, Substance-Induced/drug therapy , Adult , Antiparkinson Agents/therapeutic use , Bromocriptine/therapeutic use , Female , Humans , Neuroleptic Malignant Syndrome/drug therapyABSTRACT
BACKGROUND: There is a need for effective computational methods for quantifying the three-dimensional (3-D) spatial distribution, cellular arbor morphologies, and the morphological diversity of brain astrocytes to support quantitative studies of astrocytes in health, injury, and disease. NEW METHOD: Confocal fluorescence microscopy of multiplex-labeled (GFAP, DAPI) brain tissue is used to perform imaging of astrocytes in their tissue context. The proposed computational method identifies the astrocyte cell nuclei, and reconstructs their arbors using a local priority based parallel (LPP) tracing algorithm. Quantitative arbor measurements are extracted using Scorcioni's L-measure, and profiled by unsupervised harmonic co-clustering to reveal the morphological diversity. RESULTS: The proposed method identifies astrocyte nuclei, generates 3-D reconstructions of their arbors, and extracts quantitative arbor measurements, enabling a morphological grouping of the cell population. COMPARISON WITH EXISTING METHODS: Our method enables comprehensive spatial and morphological profiling of astrocyte populations in brain tissue for the first time, and overcomes limitations of prior methods. Visual proofreading of the results indicate a >95% accuracy in identifying astrocyte nuclei. The arbor reconstructions exhibited 3.2% fewer erroneous jumps in tracing, and 17.7% fewer false segments compared to the widely used fast-marching method that resulted in 9% jumps and 20.8% false segments. CONCLUSIONS: The proposed method can be used for large-scale quantitative studies of brain astrocyte distribution and morphology.
Subject(s)
Astrocytes/metabolism , Glial Fibrillary Acidic Protein/metabolism , Imaging, Three-Dimensional , Microscopy, Confocal , Prefrontal Cortex/cytology , Animals , Astrocytes/ultrastructure , Nerve Tissue Proteins/metabolism , RatsABSTRACT
BACKGROUND: Robust reconstructions of the three-dimensional network of blood vessels in developing embryos imaged by optical coherence tomography (OCT) are needed for quantifying the longitudinal development of vascular networks in live mammalian embryos, in support of developmental cardiovascular research. Past computational methods [such as speckle variance (SV)] have demonstrated the feasibility of vascular reconstruction, but multiple challenges remain including: the presence of vessel structures at multiple spatial scales, thin blood vessels with weak flow, and artifacts resulting from bulk tissue motion (BTM). METHODS: In order to overcome these challenges, this paper introduces a robust and scalable reconstruction algorithm based on a combination of anomaly detection algorithms and a parametric dictionary based sparse representation of blood vessels from structural OCT data. RESULTS: Validation results using confocal data as the baseline demonstrate that the proposed method enables the detection of vessel segments that are either partially missed or weakly reconstructed using the SV method. Finally, quantitative measurements of vessel reconstruction quality indicate an overall higher quality of vessel reconstruction with the proposed method. CONCLUSIONS: Results suggest that sparsity-integrated speckle anomaly detection (SSAD) is potentially a valuable tool for performing accurate quantification of the progression of vascular development in the mammalian embryonic yolk sac as imaged using OCT.
ABSTRACT
Optical coherence tomography (OCT) is an important mode of biomedical imaging for the diagnosis and management of ocular disease. Here we report on the construction of a synthetic retinal OCT image data set that may be used for quantitative analysis of image processing methods. Synthetic image data were generated from statistical characteristics of real images (n = 14). Features include: multiple stratified layers with representative thickness, boundary gradients, contour, and intensity distributions derived from real data. The synthetic data also include retinal vasculature with typical signal obscuration beneath vessels. This synthetic retinal image can provide a realistic simulated data set to help quantify the performance of image processing algorithms.