Deep learning for prediction of depressive symptoms in a large textual dataset

Neural Computing and Applications, Год журнала: 2021, Номер 34(1), С. 721 - 744

Опубликована: Авг. 27, 2021

Abstract Depression is a common illness worldwide with potentially severe implications. Early identification of depressive symptoms crucial first step towards assessment, intervention, and relapse prevention. With an increase in data sets relevance for depression, the advancement machine learning, there potential to develop intelligent systems detect depression written material. This work proposes efficient approach using Long Short-Term Memory (LSTM)-based Recurrent Neural Network (RNN) identify texts describing self-perceived depression. The applied on large dataset from public online information channel young people Norway. consists youth’s own text-based questions this channel. Features are then provided one-hot process robust features extracted reflection possible pre-defined by medical psychological experts. better than conventional approaches, which mostly based word frequencies (i.e., some topmost frequent words chosen as whole text model underlying events any message) rather symptoms. Then, deep learning RNN) train time-sequential discriminating posts no such descriptions (non-depression posts). Finally, trained RNN used automatically predict posts. system compared against approaches where it achieved superior performance others. linear discriminant space clearly reveals robustness generating clustering other traditional features. Besides, since may generate meaningful explanations decision models explainable Artificial Intelligence (XAI) algorithm called Local Interpretable Model-Agnostic Explanations (LIME). proposed symptom feature-based shows general frequency-based frequency gets more importance specific Although Norwegian dataset, similar can be datasets developed languages proper annotations symptom-based feature extraction. Thus, prediction adopted contribute mental health care technologies chatbots.

Язык: Английский

---

Deep learning in fNIRS: a review

Neurophotonics, Год журнала: 2022, Номер 9(04)

Опубликована: Июль 20, 2022

Significance: Optical neuroimaging has become a well-established clinical and research tool to monitor cortical activations in the human brain. It is notable that outcomes of functional near-infrared spectroscopy (fNIRS) studies depend heavily on data processing pipeline classification model employed. Recently, deep learning (DL) methodologies have demonstrated fast accurate performances tasks across many biomedical fields. Aim: We aim review emerging DL applications fNIRS studies. Approach: first introduce some commonly used techniques. Then, summarizes current work most active areas this field, including brain-computer interface, neuro-impairment diagnosis, neuroscience discovery. Results: Of 63 papers considered review, 32 report comparative study techniques traditional machine where 26 been shown outperforming latter terms accuracy. In addition, eight also utilize reduce amount preprocessing typically done with or increase via augmentation. Conclusions: The application mitigate hurdles present such as lengthy small sample sizes while achieving comparable improved

Язык: Английский

---

Interpretable deep learning model for major depressive disorder assessment based on functional near-infrared spectroscopy

Asian Journal of Psychiatry, Год журнала: 2024, Номер 92, С. 103901 - 103901

Опубликована: Янв. 3, 2024

Язык: Английский

---

Deep Multi-Modal Network Based Automated Depression Severity Estimation

IEEE Transactions on Affective Computing, Год журнала: 2022, Номер 14(3), С. 2153 - 2167

Опубликована: Июнь 1, 2022

Depression is a severe mental illness that impairs person's capacity to function normally in personal and professional life. The assessment of depression usually requires comprehensive examination by an expert professional. Recently, machine learning-based automatic has received considerable attention for reliable efficient diagnosis. Various techniques automated detection were developed; however, certain concerns still need be investigated. In this work, we propose novel deep multi-modal framework effectively utilizes facial verbal cues assessment. Specifically, first partition the audio video data into fixed-length segments. Then, these segments are fed Spatio-Temporal Networks as input, which captures both spatial temporal features well assigns higher weights contribute most. addition, Volume Local Directional Structural Pattern (VLDSP) based dynamic feature descriptor introduced extract dynamics encoding structural aspects. Afterwards, employ Temporal Attentive Pooling (TAP) approach summarize segment-level data. Finally, factorized bilinear pooling (MFB) strategy applied fuse effectively. An extensive experimental study reveals proposed method outperforms state-of-the-art approaches.

Язык: Английский

---

SKEAFN: Sentiment Knowledge Enhanced Attention Fusion Network for multimodal sentiment analysis

Information Fusion, Год журнала: 2023, Номер 100, С. 101958 - 101958

Опубликована: Авг. 2, 2023

Язык: Английский

---

GNN-Based Depression Recognition Using Spatio-Temporal Information: A fNIRS Study

IEEE Journal of Biomedical and Health Informatics, Год журнала: 2022, Номер 26(10), С. 4925 - 4935

Опубликована: Июль 29, 2022

In recent years, depression has become an increasingly serious problem globally. Previous studies of automatic recognition based on functional near-Infrared spectroscopy (fNIRS) or other brain imaging techniques have shown potential to serve as auxiliary diagnosis methods that provide assistance medical professionals. Recently, some found that, besides directly using the data themselves (temporal data), use connectivity among channels (spatial data) also can be effective. this paper, we propose a method Graph Neural Network (GNN) combines both temporal and spatial features fNIRS for recognition. Specifically, 96 subjects were collected pre-processed. Basic statistical metrics each channel extracted features, (coherence correlation) calculated features. Point-biserial analysis was conducted these labels data-driven motivation. For classification, considered subject graph, with node edge weights. The graphs fed into GNNs training testing. Experimental results showed our GNN-based realized best performance compared classical machine-learning regarding accuracy, F1 score, precision, especially in score over 10%.

Язык: Английский

---

Feature extraction based on sparse graphs embedding for automatic depression detection

Biomedical Signal Processing and Control, Год журнала: 2023, Номер 86, С. 105257 - 105257

Опубликована: Июль 20, 2023

Язык: Английский

---

Diagnostic machine learning applications on clinical populations using functional near infrared spectroscopy: a review

Reviews in the Neurosciences, Год журнала: 2024, Номер 35(4), С. 421 - 449

Опубликована: Фев. 3, 2024

Abstract Functional near-infrared spectroscopy (fNIRS) and its interaction with machine learning (ML) is a popular research topic for the diagnostic classification of clinical disorders due to lack robust objective biomarkers. This review provides an overview on psychiatric diseases by using fNIRS ML. Article search was carried out 45 studies were evaluated considering their sample sizes, used features, ML methodology, reported accuracy. To our best knowledge, this first that reports applications fNIRS. We found there has been increasing trend perform fNIRS-based biomarker since 2010. The most studied populations are schizophrenia ( n = 12), attention deficit hyperactivity disorder 7), autism spectrum 6) populations. There significant negative correlation between size (>21) accuracy values. Support vector (SVM) deep (DL) approaches classifier (SVM 20) (DL 10). Eight these recruited number participants more than 100 classification. Concentration changes in oxy-hemoglobin (ΔHbO) based features concentration deoxy-hemoglobin (ΔHb) ones ΔHbO-based mean ΔHbO 11) functional connections 11). Using data might be promising approach reveal specific biomarkers

Язык: Английский

---

A review of detection techniques for depression and bipolar disorder

Smart Health, Год журнала: 2022, Номер 24, С. 100282 - 100282

Опубликована: Апрель 9, 2022

Язык: Английский

---

Robust discriminant feature extraction for automatic depression recognition

Biomedical Signal Processing and Control, Год журнала: 2023, Номер 82, С. 104505 - 104505

Опубликована: Янв. 18, 2023

Язык: Английский

---