Multi-Label Classification Based on Multi-Objective Optimization

Multi-label classification refers to the task of predicting potentially multiple labels for a given instance. Conventional multi-label classification approaches focus on single objective setting, where the learning algorithm optimizes over a single performance criterion (e.g., Ranking Loss ) or a heuristic function. The basic assumption is that the optimization over one single objective can improve the overall performance of multi-label classification and meet the requirements of various applications. However, in many real applications, an optimal multi-label classifier may need to consider the trade-offs among multiple inconsistent objectives, such as minimizing Hamming Loss while maximizing Micro F1 . In this article, we study the problem of multi-objective multi-label classification and propose a novel solution (called M oml ) to optimize over multiple objectives simultaneously. Note that optimization objectives may be inconsistent, even conflicting, thus one cannot identify a single solution that is optimal on all objectives. Our M oml algorithm finds a set of non-dominated solutions which are optimal according to different trade-offs among multiple objectives. So users can flexibly construct various predictive models from the solution set, which provides more meaningful classification results in different application scenarios. Empirical studies on real-world tasks demonstrate that the M oml can effectively boost the overall performance of multi-label classification by optimizing over multiple objectives simultaneously.

Clarifying the role of psychological pain in the risks of suicidal ideation and suicidal acts among patients with major depressive episodes

The role of psychological pain in the risk of suicide was explored using a three-dimensional psychological pain model (pain arousal, painful feelings, pain avoidance). The sample consisted of 111 outpatients with major depressive episodes, including 28 individuals with suicidal histories. They completed the Chinese version of the Beck Scale for Suicide Ideation (BSI), the Beck Depression Inventory (BDI), the Psychache Scale, and the three-dimensional Psychological Pain Scale (TDPPS). A structured clinical interview was conducted to assess the history of suicidal acts. Significant correlations were found among BDI, BSI, and TDPPS scores (p < .01). Stepwise regression analyses showed that only pain avoidance scores significantly predicted suicide ideation at one's worst point (β = .79, p < .001) and suicidal acts (β = .46, p < .001). Pain avoidance was also a better predictor of current suicidal ideation (β = .37, p = .001) than were BDI scores (β = .31, p < .01). Increased levels of pain avoidance during a major depressive episode may be a dominant component of the motivation for suicide. Future clinical assessments for populations at high risk of suicide should include measures of psychological pain to reduce the incidence of suicide.

Ozone catalytic oxidation of adsorbed benzene over AgMn/HZSM-5 catalysts at room temperature

Investigation of ozone catalytic oxidation of adsorbed benzene over AgMn/HZSM-5 to provide insight into plasma catalytic oxidation of adsorbed benzene in the cycled storage–discharge process.

Synthesis and characterization of two-dimensional Nb4C3 (MXene)

By etching Nb₄AlC₃ powders in hydrofluoric acid, a phase-pure, highly conductive, Nb₄C₃ MXene – the second with formula M₄X₃ – was produced. The latter's structure was investigated using pair distribution function analysis.

In situ synthesis of Au–Pd bimetallic nanoparticles on amine-functionalized SiO2 for the aqueous-phase hydrodechlorination of chlorobenzene

The homogeneous Au–Pd NPs have been prepared through a facile in situ reduction method. The optimal AuPd_1.0/SiO₂ catalyst could complete the conversion of chlorobenzene due to the high dispersion and modified electronic properties of Pd.

An ecologically valid performance-based social functioning assessment battery for schizophrenia

Psychiatrists pay more attention to the social functioning outcome of schizophrenia nowadays. How to evaluate the real world function among schizophrenia is a challenging task due to culture difference, there is no such kind of instrument in terms of the Chinese setting. This study aimed to report the validation of an ecologically valid performance-based everyday functioning assessment for schizophrenia, namely the Beijing Performance-based Functional Ecological Test (BJ-PERFECT). Fifty community-dwelling adults with schizophrenia and 37 healthy controls were recruited. Fifteen of the healthy controls were re-tested one week later. All participants were administered the University of California, San Diego, Performance-based Skill Assessment-Brief version (UPSA-B) and the MATRICS Consensus Cognitive Battery (MCCB). The finalized assessment included three subdomains: transportation, financial management and work ability. The test-retest and inter-rater reliabilities were good. The total score significantly correlated with the UPSA-B. The performance of individuals with schizophrenia was significantly more impaired than healthy controls, especially in the domain of work ability. Among individuals with schizophrenia, functional outcome was influenced by premorbid functioning, negative symptoms and neurocognition such as processing speed, visual learning and attention/vigilance.

Anhedonia and pain avoidance in the suicidal mind: behavioral evidence for motivational manifestations of suicidal ideation in patients with major depressive disorder

OBJECTIVES

Psychological pain may be helpful in conceptualizing suicidal behavior, in that high motivation to avoid pain combined with painful feelings may contribute to an increased risk of suicide. However, no experimental study has tested this hypothesis. The aim of the present study is to provide empirical evidence for the relationship between anhedonia, pain avoidance motivation, and suicidal ideation.

METHOD

The sample comprised 40 depressed outpatients and 20 healthy control subjects. All participants completed the Beck Scale for Suicide Ideation (BSS), Beck Depression Inventory, Psychache Scale, Three-Dimensional Psychological Pain Scale, the monetary incentive delay (MID), and affective incentive delay (AID) tasks. Based on BSS scores, clinical participants were divided into high suicidal ideation (HSI) and low suicidal ideation (LSI) groups.

RESULTS

In the AID task, the HSI group had longer response times (RTs) under the reward condition than those under the punishment condition (p = .002). The LSI and control groups had shorter RTs under the reward condition compared with those under the neural condition (p <.001 and p = .008, respectively). The LSI group also had shorter RTs under the reward condition than under the punishment condition (p = .003). Pain arousal (r = -.33, p <.01) and BSS scores were significantly negatively correlated with differences in RTs between neutral and reward conditions. Pain avoidance (r = .35, p <.01) and BSS scores were positively correlated with differences in RTs between neutral and punishment conditions.

CONCLUSIONS

The AID task was more sensitive than the MID task for the detection of participants' motivation in approaching hedonic experiences and avoiding pain. A suicidal mindset is manifested as decreased motivation to experience hedonia and increased motivation to avoid pain, which could be strong predictors of suicidal behavior.

Direct targeting of SUZ12/ROCK2 by miR-200b/c inhibits cholangiocarcinoma tumourigenesis and metastasis

Background:

The multidrug resistance and distant metastasis of cholangiocarcinoma result in high postoperative recurrence and low long-term survival rates. It has been demonstrated that the ectopic expression of miR-200 suppresses the multidrug resistance and metastasis of cancer. However, the expression and function of miR-200 in cholangiocarcinoma has not yet been described.

Methods:

In this study, we identified dysregulated microRNAs (miRNAs, miR) in cholangiocarcinoma tissue by microarray analysis, and subsequent real-time PCR and northern blot analyses validated the expression of candidate miR. We performed functional analyses and investigated the relationship between miR-200b/c expression and the properties of cholangiocarcinoma cells. A dual luciferase assay was applied to examine the effect of miRNAs on the 3′-UTR of target genes, and we demonstrated the function of the target gene by siRNA transfection identifying the downstream pathway via western blotting.

Results:

We found significantly downregulated expression of four miR-200 family members (miR-200a/b/c/429) and then confirmed that ectopic miR-200b/200c inhibits the migration and invasion of cholangiocarcinoma cells both in vitro and in vivo. We found that miR-200b/c influenced the tumourigenesis of cholangiocarcinoma cells including their tumour-initiating capacity, sphere formation, and drug resistance. We further found that miR-200b/c regulated migration and invasion capacities by directly targeting rho-kinase 2 and regulated tumorigenic properties by directly targeting SUZ12 (a subunit of a polycomb repressor complex).

Conclusion:

Our study shows that miR-200b/c has a critical role in the regulation of the tumorigenic and metastatic capacity of cholangiocarcinoma and reveals the probable underlying mechanisms.

Modeling CO2 reduction on Pt(111)

Density functional theory was used to model the electrochemical reduction of CO2 on Pt(111) with an explicit solvation layer and the presence of extra hydrogen atoms to represent a negatively charged electrode. We focused on the electronic energy barriers for the first four lowest energy proton-electron transfer steps for reducing CO2 on Pt(111) beginning with adsorbed *CO2 and continuing with *COOH, *CO + H2O, *COH, and ending with *C + H2O. We find that simple elementary steps in which a proton is transferred to an adsorbate (such as the protonation of *CO to *COH) have small barriers on the order of 0.1 eV. Elementary steps in which a proton is transferred and a C-O bond is simultaneously cleaved show barriers on the order of 0.5 eV. All barriers calculated for these steps show no sign of being insurmountable at room temperature. To explain why these barriers are so small, we analyze the charge density and the density of states plots to see that first, the electron transfer is decoupled from the proton transfer so that in the initial state, the surface and adsorbate are already charged up and can easily accept the proton from solution. Also, we see that in the cases where barriers are on the order of 0.1 eV, electron density in the initial state localizes on the oxygen end of the adsorbate, while electron density is more spread out on the surface for initial states of the C-O bond cleaving elementary steps.

A gEUD-based inverse planning technique for HDR prostate brachytherapy: feasibility study

PURPOSE

The purpose of this work was to study the feasibility of a new inverse planning technique based on the generalized equivalent uniform dose for image-guided high dose rate (HDR) prostate cancer brachytherapy in comparison to conventional dose-volume based optimization.

METHODS

The quality of 12 clinical HDR brachytherapy implants for prostate utilizing HIPO (Hybrid Inverse Planning Optimization) is compared with alternative plans, which were produced through inverse planning using the generalized equivalent uniform dose (gEUD). All the common dose-volume indices for the prostate and the organs at risk were considered together with radiobiological measures. The clinical effectiveness of the different dose distributions was investigated by comparing dose volume histogram and gEUD evaluators.

RESULTS

Our results demonstrate the feasibility of gEUD-based inverse planning in HDR brachytherapy implants for prostate. A statistically significant decrease in D10 or/and final gEUD values for the organs at risk (urethra, bladder, and rectum) was found while improving dose homogeneity or dose conformity of the target volume.

CONCLUSIONS

Following the promising results of gEUD-based optimization in intensity modulated radiation therapy treatment optimization, as reported in the literature, the implementation of a similar model in HDR brachytherapy treatment plan optimization is suggested by this study. The potential of improved sparing of organs at risk was shown for various gEUD-based optimization parameter protocols, which indicates the ability of this method to adapt to the user's preferences.

CpG7909 adjuvant enhanced immunogenicity efficacy in mice immunized with ESAT6-Ag85A fusion protein, but does not confer significant protection against Mycobacterium tuberculosis infection

AIMS

This study aimed to investigate the ability of CpG7909 adjuvant to enhance immunogenicity and protective efficacy of a subunit vaccine composed of ESAT6-Ag85A fusion protein (Pe685a) of Mycobacterium tuberculosis.

METHODS AND RESULTS

ELISA was used to detect specific antibody and IFN-γ expression in sera; ELISPOT, to detect IFN-γ expression in splenocytes; MTT assay and FACS, to detect T-lymphocytes proliferation in spleens; and RT-PCR, to detect cytokines expression in lungs of mice after immunization. Bacterial load and histopathological lesions in lungs or spleens of mice challenged with Myco. tuberculosis H37Rv strain were analysed. Compared with incomplete Freund's adjuvant, CpG7909 induced more potent production of Pe685a-specific IgG2a/IgG1 antibody and higher expression of IFN-γ in sera, stimulated more generation of antigen-specific IFN-γ-secreting splenocytes, enhanced frequencies of CD3(+) CD4(+) and CD3(+) CD8(+) T-lymphocytes in spleen and increased transcription of TNF-α, IFN-γ, IL-6 and TLR9 in lung. However, lower bacterial load in lung and less severe lung pathology were not observed in CpG7909 group mice.

CONCLUSIONS

CpG7909 is able to enhance immunological effects of Pe685a subunit vaccine, but does not confer significant protective efficacy against Myco. tuberculosis infection.

SIGNIFICANCE AND IMPACT OF THE STUDY

CpG7909 as an adjuvant of subunit vaccine against Myco. tuberculosis is worthy of further investigation.

Comparison of prediction equations to estimate glomerular filtration rate in Chinese patients with chronic kidney disease

AIM

We sought to evaluate the applicability of formulae based on serum creatinine (SC) levels in Chinese patients with chronic kidney disease (CKD).

MATERIALS AND METHODS

Three hundred and twenty-seven patients with CKD who had undergone (99m) Tc-DTPA glomerular filtration rate (GFR) estimation were enrolled. The Cockcroft-Gault equation, SC-reciprocal equation, Gate equation, Hull equation, Jelliffe-1973 equation, Jelliffe-1971 equation, Mawer equation, Bjornsson equation, reexpressed 6-variable MDRD equation and reexpressed 4-variable MDRD equation were compared. Using the (99m) Tc-DTPA GFR as the standard GFR (sGFR), the accuracy of estimated GFR was compared with sGFR in various stages of CKD.

RESULTS

Median per cents of the absolute difference ranged from 28.16% to 39.39%, accuracy with a deviation less than 30% ranging from 39.4% to 53.5%, accuracy with a deviation less than 50% ranging from 63.0% to 80.7%. None of the equations had accuracy up to the 70% level with a deviation less than 30% from sGFR. Bland-Altman analysis demonstrated that mean difference ranged from -2.42 to 16.39 mL/min/1.73 m(2), whereas precision ranged from 82.66 to 106.15 mL/min/1.73 m(2). However, the agreement limits of all the equations exceeded the prior acceptable tolerances defined as 60 mL/min/1.73 m(2). Linear regression showed that the slopes of regression line ranged from 0.37 to 0.54 and intercepts ranged from -12.10 to 3.86. When the overall performance as well as bias and accuracy were compared in different stages of CKD, GFR estimated by Jelliffe-1973 equation, Cockcroft-Gault equation and Bjornsson equation showed promising results.

CONCLUSION

When SC was measured by the enzymatic method, GFR estimation equations showed great bias in Chinese CKD patients. At present, the Jelliffe-1973 equation and Cockcroft-Gault equation may be more accurate in the Chinese ethnic group.