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

BACKGROUND: Major depressive disorder (MDD) affects a substantial number of individuals worldwide. New approaches are required to improve the diagnosis of MDD, which relies heavily on subjective reports of depression-related symptoms. AIM: Establish an objective measurement and evaluation of MDD. METHODS: Functional near-infrared spectroscopy (fNIRS) was used to investigate the brain activity of MDD patients and healthy controls (HCs). Leveraging a sizeable fNIRS dataset of 263 HCs and 251 patients with MDD, including mild to moderate MDD (mMDD; n = 139) and severe MDD (sMDD; n = 77), we developed an interpretable deep learning model for screening MDD and staging its severity. RESULTS: The proposed deep learning model achieved an accuracy of 80.9% in diagnostic classification and 78.6% in severity staging for MDD. We discerned five channels with the most significant contribution to MDD identification through Shapley additive explanations (SHAP), located in the right medial prefrontal cortex, right dorsolateral prefrontal cortex, right superior temporal gyrus, and left posterior superior frontal cortex. The findings corresponded closely to the features of haemoglobin responses between HCs and individuals with MDD, as we obtained a good discriminative ability for MDD using cortical channels that are related to the disorder, namely the frontal and temporal cortical channels with areas under the curve of 0.78 and 0.81, respectively. CONCLUSION: Our study demonstrated the potential of integrating the fNIRS system with artificial intelligence algorithms to classify and stage MDD in clinical settings using a large dataset. This approach can potentially enhance MDD assessment and provide insights for clinical diagnosis and intervention.

Differentiating people with schizophrenia from healthy controls in a developing Country: An evaluation of portable functional near infrared spectroscopy (fNIRS) as an adjunct diagnostic tool.

Introduction: This study aimed to evaluate portable functional near-infrared spectroscopy (fNIRS) device as an adjunct diagnostic tool in Vietnam to assess hemodynamics when people with schizophrenia and healthy controls performed cognitive tasks. Methods: One hundred fifty-seven participants were divided into schizophrenia (n = 110) and healthy controls group (n = 47), which were recruited by match of age, and gender. Hemodynamic responses in the frontal cortex were monitored with a 48-channel portable device during the Stroop Color-Word Test (SCWT) and Verbal Fluency Test (VFT). General linear model compared the differences in oxyhemoglobin (HbO2) levels between the two groups. The Receiver Operating Characteristic (ROC) graph was generated for each neuroanatomical area. Results: People with schizophrenia did not show significant activation in the frontal lobe during the SCWT and VFT as compared to pre-task. During the VFT, the area under the ROC curve of the bilateral dorsolateral prefrontal cortex, bilateral orbitofrontal cortex, bilateral frontopolar prefrontal cortex, and bilateral ventrolateral prefrontal cortex were greater than 0.7 (p < 0.001). The area under the ROC curve (AUC) for the right orbitofrontal cortex was maximal during the VFT (AUC = 0.802, 95%CI = 0.731-0.872). The Youden's index reached a peak (0.57) at the optimal cut-point value (HbO2 cutoff <0.209 µmol/ml for schizophrenia) in which the sensitivity was 85%; specificity was 72%; positive predictive value (PPV) was 0.88; negative predictive value (NPV) was 0.68 and correct classification rate was 76%. Discussion: Assessing hemodynamics during VFT by portable fNIRS offers the potential as an adjunct diagnostic tool for schizophrenia in developing countries.

Utility of portable functional near-infrared spectroscopy (fNIRS) in patients with bipolar and unipolar disorders: A comparison with healthy controls.

BACKGROUND: This study aimed to evaluate portable functional near-infrared spectroscopy (fNIRS) device as an adjunct diagnostic tool for bipolar and unipolar disorders while performing cognitive tasks. METHODS: 150 participants were divided into three groups including bipolar, unipolar disorder, and healthy controls (50:50:50), matched by age, gender, and family history of mood disorder. Hemodynamics in the frontal cortex were monitored by fNIRS during the Stroop Color-Word Test and Verbal Fluency Test. The GLM compared the differences in oxy-hemoglobin levels between the two groups. The Receiver Operating Characteristic (ROC) graph was generated for each neuroanatomical area. RESULTS: For people with BD group, the area under the ROC curve (AUC) for the left orbitofrontal cortex was maximal during the VFT [AUC = 0.727, 95%CI = 0.617-0.824]. The Youden's index reached a peak (0.40) at the optimal cut-point value (HbO2 cutoff <0.180 µmol/ml for BD) in which the sensitivity was 82 %; specificity was 58 %; PPV was 0.66; NPV was 0.76 and correct classification rate was 70 %. Regarding the UD group, during VFT, the highest value AUC [AUC = 0.822, 95%CI = 0.740-0.903] was recorded in the left dorsolateral prefrontal cortex with the optimal cut-off value (HbO2cutoff ≥0.163 µmol/ml for healthy controls; <0.163 for unipolar disorder), the sensitivity was 72 %; specificity was 82 %; PPV was 0.80; NPV was 0.75, correct classification rate was 77 %, and the Youden's index was 0.54. CONCLUSION: Assessing hemodynamics during VFT using portable fNIRS offers the potential as an adjunct diagnostic tool for mood disorders in low-resource environments.

Identifying neuroimaging biomarkers of major depressive disorder from cortical hemodynamic responses using machine learning approaches.

BACKGROUND: Early diagnosis of major depressive disorder (MDD) could enable timely interventions and effective management which subsequently improve clinical outcomes. However, quantitative and objective assessment tools for the suspected cases who present with depressive symptoms have not been fully established. METHODS: Based on a large-scale dataset (n = 363 subjects) collected with functional near-infrared spectroscopy (fNIRS) measurements during the verbal fluency task (VFT), this study proposed a data representation method for extracting spatiotemporal characteristics of NIRS signals, which emerged as candidate predictors in a two-phase machine learning framework to detect distinctive biomarkers for MDD. Supervised classifiers (e.g., support vector machine (SVM), k-nearest neighbors (KNN)) cooperated with cross-validation were implemented to evaluate the predictive capability of selected features in a training set. Another test set that was not involved in developing the algorithms enabled the independent assessment of the model's generalization. FINDINGS: For the classification with the optimal fusion features, the SVM classifier achieved the highest accuracy of 75.6% ± 4.7% in the nested cross-validation, and the correct prediction rate of 78.0% with a sensitivity of 75.0% and a specificity of 81.4% in the test set. Moreover, the multiway ANOVA test on clinical and demographic factors confirmed that twenty out of 39 optimal features were significantly correlated with the MDD-distinctive consequence. INTERPRETATION: The abnormal prefrontal activity of MDD may be quantified as diminished relative intensity and inappropriate activation timing of hemodynamic response, resulting in an objectively measurable biomarker for assessing cognitive deficits and screening MDD at the early stage. FUNDING: This study was funded by NUS iHeathtech Other Operating Expenses (R-722-000-004-731).

Integration of the Cortical Haemodynamic Response Measured by Functional Near-Infrared Spectroscopy and Amino Acid Analysis to Aid in the Diagnosis of Major Depressive Disorder.

BACKGROUND: Major depressive disorder (MDD) is a debilitating condition with a high disease burden and medical comorbidities. There are currently few to no validated biomarkers to guide the diagnosis and treatment of MDD. In the present study, we evaluated the differences between MDD patients and healthy controls (HCs) in terms of cortical haemodynamic responses during a verbal fluency test (VFT) using functional near-infrared spectroscopy (fNIRS) and serum amino acid profiles, and ascertained if these parameters were correlated with clinical characteristics. METHODS: Twenty-five (25) patients with MDD and 25 age-, gender-, and ethnicity-matched HCs were recruited for the study. Real-time monitoring of the haemodynamic response during completion of a VFT was quantified using a 52-channel NIRS system. Serum samples were analysed and quantified by liquid chromatography-mass spectrometry for amino acid profiling. Receiver-operating characteristic (ROC) curves were used to classify potential candidate biomarkers. RESULTS: The MDD patients had lower prefrontal and temporal activation during completion of the VFT than HCs. The MDD patients had lower mean concentrations of oxy-Hb in the left orbitofrontal cortex (OFC), and lower serum histidine levels. When the oxy-haemoglobin response was combined with the histidine concentration, the sensitivity and specificity of results improved significantly from 66.7% to 73.3% and from 65.0% to 90.0% respectively, as compared to results based only on the NIRS response. CONCLUSIONS: These findings demonstrate the use of combination biomarkers to aid in the diagnosis of MDD. This technique could be a useful approach to detect MDD with greater precision, but additional studies are required to validate the methodology.

Functional near-infrared spectroscopy during a decision-making task in patients with major depressive disorder.

OBJECTIVE: Patients with major depressive disorder tend to exhibit poorer decision-making capacity than the general population, but neurobiological evidence is lacking. Functional near-infrared spectroscopy monitors changes in oxy-haemoglobin concentration in the cerebral cortex. It may provide an objective assessment of neurophysiological responses during decision-making processes. Thus, this study investigated the effect of major depressive disorder diagnosis and severity on prefrontal cortex activity during the Iowa gambling task. METHODS: Right-handed healthy controls (n = 25) and patients with major depressive disorder (n = 25) were matched for age, gender, ethnicity and years of education in this cross-sectional study. Functional near-infrared spectroscopy signals and the responses made during a computerised Iowa gambling task were recorded. In addition, demographics, clinical history and symptom severity were noted. RESULTS: Compared to healthy controls, patients with major depressive disorder had reduced haemodynamic response in several cortical regions of the frontal lobe (Hedge's g range from 0.71 to 1.52; p values range from ⩽0.001 to 0.041). Among patients, mean oxy-haemoglobin declined with major depressive disorder severity in the right orbitofrontal cortex (Pearson's r = -0.423; p = 0.024). CONCLUSION: Haemodynamic dysfunction of the prefrontal cortex during decision-making processes is associated with major depressive disorder diagnosis and severity. These neurophysiological alterations may have a role in the decision-making capacity of patients with major depressive disorder.

Cortical haemodynamic response measured by functional near infrared spectroscopy during a verbal fluency task in patients with major depression and borderline personality disorder.

BACKGROUND: Functional near infrared spectroscopy (fNIRS) provides a direct and quantitative assessment of cortical haemodynamic function during a cognitive task. This functional neuroimaging modality may be used to elucidate the pathophysiology of psychiatric disorders, and identify neurophysiological differences between co-occurring psychiatric disorders. However, fNIRS research on borderline personality disorder (BPD) has been limited. Hence, this study aimed to compare cerebral haemodynamic function in healthy controls (HC), patients with major depressive disorder (MDD) and patients with BPD. METHODS: fNIRS signals during a verbal fluency task designed for clinical assessment was recorded for all participants. Demographics, clinical history and symptom severity were also noted. FINDINGS: Compared to HCs (n = 31), both patient groups (MDD, n = 31; BPD, n = 31) displayed diminished haemodynamic response in the frontal, temporal and parietal cortices. Moreover, haemodynamic response in the right frontal cortex is markedly lower in patients with MDD compared to patients with BPD. INTERPRETATION: Normal cortical function in patients with BPD is disrupted, but not as extensively as in patients with MDD. These results provide further neurophysiological evidence for the distinction of patients with MDD from patients with BPD.

IL-1ß, IL-6, TNF- α and CRP in Elderly Patients with Depression or Alzheimer's disease: Systematic Review and Meta-Analysis.

We carried out systematic review and meta-analysis to evaluate whether peripheral levels of pro-inflammatory markers including Interleukin-1 beta (IL-1ß), Interleukin-6 (IL-6), Tumor Necrosis Factor-α (TNF- α) and C-Reactive Protein (CRP) are significantly higher in elderly with depression and Alzheimer's disease. We searched Pubmed, PsycINFO and Embase, and thirty-four relevant studies (2609 with Depression, 1645 with Alzheimer's disease and 14363 Controls) were included. Compared with controls, IL-1ß (pooled standardized mean difference [SMD]: 0.642; 95% confidence interval [CI]: 0.078-1.206; significant heterogeneity: I2 = 86.28%) and IL-6 (pooled SMD: 0.377; 95% CI: 0.156-0.598; significant heterogeneity: I2 = 88.75%) were significantly elevated in depression. There was no difference in TNF-α (p = 0.351) and CRP (p = 0.05) between those with depression and controls. Compared with controls, IL-1ß (pooled SMD: 1.37, 95% CI: 0.06-2.68, significant heterogeneity: I2 = 96.01%) was significantly elevated in Alzheimer's disease. There were no differences in IL-6 (p = 0.138), TNF-α (p = 0.451) and CRP (p = 0.07) between elderly with Alzheimer's disease and controls. After Bonferroni adjustment, only IL-6 remained significantly higher in depression. Elderly with depression have higher IL-6 than controls, while those with Alzheimer's disease did not have higher peripheral inflammatory markers.

Differentiating Medicated Patients Suffering from Major Depressive Disorder from Healthy Controls by Spot Urine Measurement of Monoamines and Steroid Hormones.

Introduction: Major Depressive Disorder (MDD) is a common psychiatric disorder. Currently, there is no objective, cost-effective and non-invasive method to measure biological markers related to the pathogenesis of MDD. Previous studies primarily focused on urinary metabolite markers which are not proximal to the pathogenesis of MDD. Herein, we compare urinary monoamines, steroid hormones and the derived ratios amongst MDD when compared to healthy controls. Methods: Morning urine samples of medicated patients suffering from MDD (n = 47) and healthy controls (n = 41) were collected. Enzyme-linked immunosorbent assay (ELISA) was performed to measure five biomarkers: cortisol, dopamine, noradrenaline, serotonin and sulphate derivative of dehydroepiandrosterone (DHEAS). The mean urinary levels and derived ratios of monoamines and steroid hormones were compared between patients and controls to identify potential biomarkers. The receiver operative characteristic curve (ROC) analysis was conducted to evaluate the diagnostic performance of potential biomarkers. Results: Medicated patients with MDD showed significantly higher spot urine ratio of DHEAS/serotonin (1.56 vs. 1.19, p = 0.004) and lower ratio of serotonin/dopamine (599.71 vs. 888.60, p = 0.008) than healthy controls. A spot urine serotonin/dopamine ratio cut-off of >667.38 had a sensitivity of 73.2% and specificity of 51.1%. Conclusions: Our results suggest that spot urine serotonin/dopamine ratio can be used as an objective diagnostic method for adults with MDD.

Analysis of print news media framing of ketamine treatment in the United States and Canada from 2000 to 2015.

OBJECTIVES: There are multifaceted views on the use of ketamine, a potentially addictive substance, to treat mental health problems. The past 15 years have seen growing media coverage of ketamine for medical and other purposes. This study examined the print news media coverage of medical and other uses of ketamine in North America to determine orientations and trends over time. METHODS: Print newspaper coverage of ketamine from 2000 to 2015 was reviewed, resulting in 43 print news articles from 28 North American newspapers. A 55-item structured coding instrument was applied to assess news reports of ketamine. Items captured negative and positive aspects, therapeutic use of ketamine, and adverse side effects. Chi-squares tested for changes in trends over time. RESULTS: In the 15-year reviewed period, the three most frequent themes related to ketamine were: abuse (68.2%), legal status (34.1%), and clinical use in anesthesia (31.8%). There was significant change in trends during two periods (2000-2007 and 2008-2015). In 2008-2015, print news media articles were significantly more likely to encourage clinical use of ketamine to treat depression (p = 0.002), to treat treatment resistant depression (p = 0.043), and to claim that ketamine is more effective than conventional antidepressants (p = 0.043). CONCLUSIONS: Our review found consistent positive changes in the portrayals of ketamine by the print news media as a therapeutic antidepressant that mirror the recent scientific publications. These changes in news media reporting might influence the popularity of ketamine use to treat clinical depression. Guidance is required for journalists on objective reporting of medical research findings, including limitations of current research evidence and potential risks of ketamine.