MyMedR

Displaying all 18 publications

Abstract:

Sort:

A generalized subspace approach for estimating visual evoked potentials

Kamel N, Yusoff MZ

Annu Int Conf IEEE Eng Med Biol Soc, 2008;2008:5208-11.
PMID: 19163891 DOI: 10.1109/IEMBS.2008.4650388

A "single-trial" signal subspace approach for extracting visual evoked potential (VEP) from the ongoing 'colored' electroencephalogram (EEG) noise is proposed. The algorithm applies the generalized eigendecomposition on the covariance matrices of the VEP and noise to transform them jointly into diagonal matrices in order to avoid a pre-whitening stage. The proposed generalized subspace approach (GSA) decomposes the corrupted VEP space into a signal subspace and noise subspace. Enhancement is achieved by removing the noise subspace and estimating the clean VEPs only from the signal subspace. The validity and effectiveness of the proposed GSA scheme in estimating the latencies of P100's (used in objective assessment of visual pathways) are evaluated using real data collected from Selayang Hospital in Kuala Lumpur. The performance of GSA is compared with the recently proposed single-trial technique called the Third Order Correlation (TOC).
Fulltext Motor Imagery Classification for Brain Computer Interface Using Deep Convolutional Neural Networks and Mixup Augmentation

Alwasiti H, Yusoff MZ

IEEE Open J Eng Med Biol, 2022;3:171-177.
PMID: 36578777 DOI: 10.1109/OJEMB.2022.3220150

Goal: Building a DL model that can be trained on small EEG training set of a single subject presents an interesting challenge that this work is trying to address. In particular, this study is trying to avoid the need for long EEG data collection sessions, and without combining multiple subjects training datasets, which has a detrimental effect on the classification performance due to the inter-individual variability among subjects. Methods: A customized Convolutional Neural Network with mixup augmentation was trained with [Formula: see text]120 EEG trials for only one subject per model. Results: Modified ResNet18 and DenseNet121 models with mixup augmentation achieved 0.920 (95% Confidence Interval: 0.908, 0.933) and 0.933 (95% Confidence Interval: 0.922, 0.945) classification accuracy, respectively. Conclusions: We show that the designed classifiers resulted in a higher classification performance in comparison to other DL classifiers of previous studies on the same dataset, despite the limited training dataset used in this work.
Fulltext Medication adherence in schizophrenia: a comparison between outpatients and relapse cases

Razali SM, Yusoff MZ

East Asian Arch Psychiatry, 2014 Jun;24(2):68-74.
PMID: 24986201

Objective: Adherence to medication is essential for maximising the outcomes of patients with schizophrenia as the consequences of poor adherence are devastating. The study aimed to compare medication adherence between patients with relapse schizophrenia and those attending psychiatric follow-up clinics, and to determine the factors affecting adherence.
Methods: This was a cross-sectional study involving 70 patients with schizophrenia who were divided equally into 2 groups. Medication adherence was assessed with the Medication Adherence Rating Scale. Appropriate instruments were used to measure insight, social support, and psychopathology. Various socio-demographic and clinical variables were explored to find associations with medication adherence.
Results: Medication adherence among patients with schizophrenia was poor; 51% of the patients did not adhere to a medication regimen. Adherence was better in outpatients with schizophrenia (61%) than in relapse cases (39%), although the difference was not statistically significant (t = 1.70; p = 0.09). Besides, relapse patients had significant higher number of admission (X2 = 22.95; p < 0.05) and severe psychopathology (t = –29.96; p < 0.05), while perceived social support was significantly better in outpatients with schizophrenia (t = 2.90; p < 0.05). Frequency of admission (adjusted b = –0.55; 95% confidence interval [CI], -0.99 to -0.10; p < 0.05) and psychopathology (adjusted b = –0.12; 95% CI, -0.24 to -0.01; p < 0.05) were also significantly associated with medication adherence.
Conclusion: Medication adherence among both groups of patients with schizophrenia was poor. If adherence is addressed appropriately, the number of admissions and severity of psychopathology could be improved.
Key words: Patient compliance; Psychopathology; Schizophrenia; Social support
Study site: Psychiatric clinic, Hospital Universiti Sains Malaysia (HUSM)
Single-trial subspace-based approach for VEP extraction

Kamel N, Yusoff MZ, Hani AF

IEEE Trans Biomed Eng, 2011 May;58(5):1383-93.
PMID: 21177154 DOI: 10.1109/TBME.2010.2101073

A signal subspace approach for extracting visual evoked potentials (VEPs) from the background electroencephalogram (EEG) colored noise without the need for a prewhitening stage is proposed. Linear estimation of the clean signal is performed by minimizing signal distortion while maintaining the residual noise energy below some given threshold. The generalized eigendecomposition of the covariance matrices of a VEP signal and brain background EEG noise is used to transform them jointly to diagonal matrices. The generalized subspace is then decomposed into signal subspace and noise subspace. Enhancement is performed by nulling the components in the noise subspace and retaining the components in the signal subspace. The performance of the proposed algorithm is tested with simulated and real data, and compared with the recently proposed signal subspace techniques. With the simulated data, the algorithms are used to estimate the latencies of P(100), P(200), and P(300) of VEP signals corrupted by additive colored noise at different values of SNR. With the real data, the VEP signals are collected at Selayang Hospital, Kuala Lumpur, Malaysia, and the capability of the proposed algorithm in detecting the latency of P(100) is obtained and compared with other subspace techniques. The ensemble averaging technique is used as a baseline for this comparison. The results indicated significant improvement by the proposed technique in terms of better accuracy and less failure rate.
Fulltext A review of norms and normative multiagent systems

Mahmoud MA, Ahmad MS, Yusoff MZ, Mustapha A

ScientificWorldJournal, 2014;2014:684587.
PMID: 25110739 DOI: 10.1155/2014/684587

Norms and normative multiagent systems have become the subjects of interest for many researchers. Such interest is caused by the need for agents to exploit the norms in enhancing their performance in a community. The term norm is used to characterize the behaviours of community members. The concept of normative multiagent systems is used to facilitate collaboration and coordination among social groups of agents. Many researches have been conducted on norms that investigate the fundamental concepts, definitions, classification, and types of norms and normative multiagent systems including normative architectures and normative processes. However, very few researches have been found to comprehensively study and analyze the literature in advancing the current state of norms and normative multiagent systems. Consequently, this paper attempts to present the current state of research on norms and normative multiagent systems and propose a norm's life cycle model based on the review of the literature. Subsequently, this paper highlights the significant areas for future work.
Fulltext Four products from Escherichia coli pseudogenes increase hydrogen production

Mohd Yusoff MZ, Hashiguchi Y, Maeda T, Wood TK

Biochem Biophys Res Commun, 2013 Oct 4;439(4):576-9.
PMID: 24025676 DOI: 10.1016/j.bbrc.2013.09.016

Pseudogenes are considered to be nonfunctional genes that lack a physiological role. By screening 3985 Escherichia coli mutants using chemochromic membranes, we found four pseudogenes involved in hydrogen metabolism. Knockouts of pseudogenes ydfW and ypdJ had a defective hydrogen phenotype on glucose and formate, respectively. Also, the knockout of pseudogene yqiG formed hydrogen from formate but not from glucose. For the yqiG mutant, 100% hydrogen recovery was obtained by the complementation of YqiG via a plasmid. The knockout of pseudogene ylcE showed hydrogen deficiency in minimal media which suggested that the role of YlcE is associated with cell growth. Hence, the products of these four pseudogenes play an important physiological role in hydrogen production in E. coli.
Fulltext Deep learning architectures analysis for age-related macular degeneration segmentation on optical coherence tomography scans

Alsaih K, Yusoff MZ, Tang TB, Faye I, Mériaudeau F

Comput Methods Programs Biomed, 2020 Oct;195:105566.
PMID: 32504911 DOI: 10.1016/j.cmpb.2020.105566

BACKGROUND AND OBJECTIVES: Aged people usually are more to be diagnosed with retinal diseases in developed countries. Retinal capillaries leakage into the retina swells and causes an acute vision loss, which is called age-related macular degeneration (AMD). The disease can not be adequately diagnosed solely using fundus images as depth information is not available. The variations in retina volume assist in monitoring ophthalmological abnormalities. Therefore, high-fidelity AMD segmentation in optical coherence tomography (OCT) imaging modality has raised the attention of researchers as well as those of the medical doctors. Many methods across the years encompassing machine learning approaches and convolutional neural networks (CNN) strategies have been proposed for object detection and image segmentation.
METHODS: In this paper, we analyze four wide-spread deep learning models designed for the segmentation of three retinal fluids outputting dense predictions in the RETOUCH challenge data. We aim to demonstrate how a patch-based approach could push the performance for each method. Besides, we also evaluate the methods using the OPTIMA challenge dataset for generalizing network performance. The analysis is driven into two sections: the comparison between the four approaches and the significance of patching the images.
RESULTS: The performance of networks trained on the RETOUCH dataset is higher than human performance. The analysis further generalized the performance of the best network obtained by fine-tuning it and achieved a mean Dice similarity coefficient (DSC) of 0.85. Out of the three types of fluids, intraretinal fluid (IRF) is more recognized, and the highest DSC value of 0.922 is achieved using Spectralis dataset. Additionally, the highest average DSC score is 0.84, which is achieved by PaDeeplabv3+ model using Cirrus dataset.
CONCLUSIONS: The proposed method segments the three fluids in the retina with high DSC value. Fine-tuning the networks trained on the RETOUCH dataset makes the network perform better and faster than training from scratch. Enriching the networks with inputting a variety of shapes by extracting patches helped to segment the fluids better than using a full image.
Fulltext Isolation and characterization of Burkholderia sp. strain CCA53 exhibiting ligninolytic potential

Akita H, Kimura Z, Mohd Yusoff MZ, Nakashima N, Hoshino T

Springerplus, 2016;5:596.
PMID: 27247892 DOI: 10.1186/s40064-016-2237-y

Microbial degradation of lignin releases fermentable sugars, effective utilization of which could support biofuel production from lignocellulosic biomass. In the present study, a lignin-degrading bacterium was isolated from leaf soil and identified as Burkholderia sp. based on 16S rRNA gene sequencing. This strain was named CCA53, and its lignin-degrading capability was assessed by observing its growth on medium containing alkali lignin or lignin-associated aromatic monomers as the sole carbon source. Alkali lignin and at least eight lignin-associated aromatic monomers supported growth of this strain, and the most effective utilization was observed for p-hydroxybenzene monomers. These findings indicate that Burkholderia sp. strain CCA53 has fragmentary activity for lignin degradation.
Fulltext Subcritical Water-Carbon Dioxide Pretreatment of Oil Palm Mesocarp Fiber for Xylooligosaccharide and Glucose Production

Ahmad N, Zakaria MR, Mohd Yusoff MZ, Fujimoto S, Inoue H, Ariffin H, et al.

Molecules, 2018 May 30;23(6).
PMID: 29848973 DOI: 10.3390/molecules23061310

The present work aimed to investigate the pretreatment of oil palm mesocarp fiber (OPMF) in subcritical H₂O-CO₂ at a temperature range from 150⁻200 °C and 20⁻180 min with CO₂ pressure from 3⁻5 MPa. The pretreated solids and liquids from this process were separated by filtration and characterized. Xylooligosaccharides (XOs), sugar monomers, acids, furans and phenols in the pretreated liquids were analyzed by using HPLC. XOs with a degree of polymerization X2⁻X4 comprising xylobiose, xylotriose, xylotetraose were analyzed by using HPAEC-PAD. Enzymatic hydrolysis was performed on cellulose-rich pretreated solids to observe xylose and glucose production. An optimal condition for XOs production was achieved at 180 °C, 60 min, 3 MPa and the highest XOs obtained was 81.60 mg/g which corresponded to 36.59% of XOs yield from total xylan of OPMF. The highest xylose and glucose yields obtained from pretreated solids were 29.96% and 84.65%, respectively at cellulase loading of 10 FPU/g-substrate.
Ecotoxicological assessment of palm oil mill effluent final discharge by zebrafish (Danio rerio) embryonic assay

Hashiguchi Y, Zakaria MR, Toshinari M, Mohd Yusoff MZ, Shirai Y, Hassan MA

Environ Pollut, 2021 May 15;277:116780.
PMID: 33640825 DOI: 10.1016/j.envpol.2021.116780

Most palm oil mills adopted conventional ponding system, including anaerobic, aerobic, facultative and algae ponds, for the treatment of palm oil mill effluent (POME). Only a few mills installed a bio-polishing plant to treat POME further before its final discharge. The present study aims to determine the quality and toxicity levels of POME final discharge from three different mills by using conventional chemical analyses and fish (Danio rerio) embryo toxicity (FET) test. The effluent derived from mill A which installed with a bio-polishing plant had lower values of BOD, COD and TSS at 45 mg/L, 104 mg/L, and 27 mg/L, respectively. Only mill A nearly met the industrial effluent discharge standard for BOD. In FET test, effluent from mill A recorded low lethality and most of the embryos were malformed after hatching (half-maximal effective concentration (EC50) = 20%). The highest toxicity was observed from the effluent of mill B and all embryos were coagulated after 24 h in samples greater than 75% of effluent (38% of half-maximal lethal concentration (LC50) at 96 h). The embryos in the effluent from mill C recorded high mortality after hatching, and the survivors were malformed after 96 h exposure (LC50 = 26%). Elemental analysis of POME final discharge samples showed Cu, Zn, and Fe concentrations were in the range of 0.10-0.32 mg/L, 0.01-0.99 mg/L, and 0.94-4.54 mg/L, respectively and all values were below the effluent permissible discharge limits. However, the present study found these metals inhibited D. rerio embryonic development at 0.12 mg/L of Cu, and 4.9 mg/L of Fe for 96 h-EC50. The present study found that bio-polishing plant installed in mill A effectively removing pollutants especially BOD and the FET test was a useful method to monitor quality and toxicity of the POME final discharge samples.
Fulltext Frontal Electroencephalogram Alpha Asymmetry during Mental Stress Related to Workplace Noise

Alyan E, Saad NM, Kamel N, Yusoff MZ, Zakariya MA, Rahman MA, et al.

Sensors (Basel), 2021 Mar 11;21(6).
PMID: 33799722 DOI: 10.3390/s21061968

This study aims to investigate the effects of workplace noise on neural activity and alpha asymmetries of the prefrontal cortex (PFC) during mental stress conditions. Workplace noise exposure is a pervasive environmental pollutant and is negatively linked to cognitive effects and selective attention. Generally, the stress theory is assumed to underlie the impact of noise on health. Evidence for the impacts of workplace noise on mental stress is lacking. Fifteen healthy volunteer subjects performed the Montreal imaging stress task in quiet and noisy workplaces while their brain activity was recorded using electroencephalography. The salivary alpha-amylase (sAA) was measured before and immediately after each tested workplace to evaluate the stress level. The results showed a decrease in alpha rhythms, or an increase in cortical activity, of the PFC for all participants at the noisy workplace. Further analysis of alpha asymmetry revealed a greater significant relative right frontal activation of the noisy workplace group at electrode pairs F4-F3 but not F8-F7. Furthermore, a significant increase in sAA activity was observed in all participants at the noisy workplace, demonstrating the presence of stress. The findings provide critical information on the effects of workplace noise-related stress that might be neglected during mental stress evaluations.
Fulltext Draft Genome Sequence of Burkholderia sp. Strain CCA53, Isolated from Leaf Soil

Akita H, Kimura Z, Yusoff MZ, Nakashima N, Hoshino T

Genome Announc, 2016;4(4).
PMID: 27389268 DOI: 10.1128/genomeA.00630-16

Burkholderia sp. strain CCA53 was isolated from leaf soil collected in Higashi-Hiroshima City in Hiroshima Prefecture, Japan. Here, we present a draft genome sequence of this strain, which consists of a total of 4 contigs containing 6,647,893 bp, with a G+C content of 67.0% and comprising 9,329 predicted coding sequences.
Fulltext Pseudogene product YqiG is important for pflB expression and biohydrogen production in Escherichia coli BW25113

Zakaria MA, Mohd Yusoff MZ, Zakaria MR, Hassan MA, Wood TK, Maeda T

3 Biotech, 2018 Oct;8(10):435.
PMID: 30306004 DOI: 10.1007/s13205-018-1461-2

Pseudogenes in the Escherichia coli genome are assumed to be non-functional. In this study, Keio collection BW25113∆yqiG and YqiG-producing strain (BW25113/pCA24N-YqiG) were used to evaluate the importance of pseudogene yqiG in hydrogen metabolism. Our results show pseudogene protein YqiG was identified as an essential protein in the production of biohydrogen from glucose. The mutant yqiG decreased biohydrogen production from 37 µmol mg-1 protein to 6 µmol mg-1 protein compared to the wild-type strain, and glucose consumption was reduced by 80%. Through transcriptional analysis, we found that the yqiG mutation represses pflB transcription tenfold; pflB encodes pyruvate-formate lyase, one of the key enzymes in the anaerobic metabolism of E. coli. Moreover, production of YqiG stimulated glycolysis and increased biohydrogen productivity 1.5-fold compared to that of the wild-type strain. Thus, YqiG is important for the central glycolysis reaction and is able to influence hydrogen metabolism activity in E. coli.
Effective Connectivity for Decoding Electroencephalographic Motor Imagery Using a Probabilistic Neural Network

Awais MA, Yusoff MZ, Khan DM, Yahya N, Kamel N, Ebrahim M

Sensors (Basel), 2021 Sep 30;21(19).
PMID: 34640888 DOI: 10.3390/s21196570

Motor imagery (MI)-based brain-computer interfaces have gained much attention in the last few years. They provide the ability to control external devices, such as prosthetic arms and wheelchairs, by using brain activities. Several researchers have reported the inter-communication of multiple brain regions during motor tasks, thus making it difficult to isolate one or two brain regions in which motor activities take place. Therefore, a deeper understanding of the brain's neural patterns is important for BCI in order to provide more useful and insightful features. Thus, brain connectivity provides a promising approach to solving the stated shortcomings by considering inter-channel/region relationships during motor imagination. This study used effective connectivity in the brain in terms of the partial directed coherence (PDC) and directed transfer function (DTF) as intensively unconventional feature sets for motor imagery (MI) classification. MANOVA-based analysis was performed to identify statistically significant connectivity pairs. Furthermore, the study sought to predict MI patterns by using four classification algorithms-an SVM, KNN, decision tree, and probabilistic neural network. The study provides a comparative analysis of all of the classification methods using two-class MI data extracted from the PhysioNet EEG database. The proposed techniques based on a probabilistic neural network (PNN) as a classifier and PDC as a feature set outperformed the other classification and feature extraction techniques with a superior classification accuracy and a lower error rate. The research findings indicate that when the PDC was used as a feature set, the PNN attained the greatest overall average accuracy of 98.65%, whereas the same classifier was used to attain the greatest accuracy of 82.81% with the DTF. This study validates the activation of multiple brain regions during a motor task by achieving better classification outcomes through brain connectivity as compared to conventional features. Since the PDC outperformed the DTF as a feature set with its superior classification accuracy and low error rate, it has great potential for application in MI-based brain-computer interfaces.
Production of acetoin from hydrothermally pretreated oil mesocarp fiber using metabolically engineered Escherichia coli in a bioreactor system

Mohd Yusoff MZ, Akita H, Hassan MA, Fujimoto S, Yoshida M, Nakashima N, et al.

Bioresour Technol, 2017 Dec;245(Pt A):1040-1048.
PMID: 28946206 DOI: 10.1016/j.biortech.2017.08.131

Acetoin is used in the biochemical, chemical and pharmaceutical industries. Several effective methods for acetoin production from petroleum-based substrates have been developed, but they all have an environmental impact and do not meet sustainability criteria. Here we describe a simple and efficient method for acetoin production from oil palm mesocarp fiber hydrolysate using engineered Escherichia coli. An optimization of culture conditions for acetoin production was carried out using reagent-grade chemicals. The final concentration reached 29.9gL(-1) with a theoretical yield of 79%. The optimal pretreatment conditions for preparing hydrolysate with higher sugar yields were then determined. When acetoin was produced using hydrolysate fortified with yeast extract, the theoretical yield reached 97% with an acetoin concentration of 15.5gL(-1). The acetoin productivity was 10-fold higher than that obtained using reagent-grade sugars. This approach makes use of a compromise strategy for effective utilization of oil palm biomass towards industrial application.
A one-step self-sustained low temperature carbonization of coconut shell biomass produced a high specific surface area biochar-derived nano-adsorbent

Samsudin MH, Hassan MA, Idris J, Ramli N, Mohd Yusoff MZ, Ibrahim I, et al.

Waste Manag Res, 2019 May;37(5):551-555.
PMID: 30727859 DOI: 10.1177/0734242X18823953

A one-step self-sustained carbonization of coconut shell biomass, carried out in a brick reactor at a relatively low temperature of 300-500°C, successfully produced a biochar-derived adsorbent with 308 m2/g surface area, 2 nm pore diameter, and 0.15 cm3/g total pore volume. The coconut shell biochar qualifies as a nano-adsorbent, supported by scanning electron microscope images, which showed well-developed nano-pores on the surface of the biochar structure, even though there was no separate activation process. This is the first report whereby coconut shell can be converted to biochar-derived nano-adsorbent at a low carbonization temperature, without the need of the activation process. This is superior to previous reports on biochar produced from oil palm empty fruit bunch.
Fulltext Mercerization Optimization of Bamboo (Bambusa vulgaris) Fiber-Reinforced Epoxy Composite Structures Using a Box-Behnken Design

Hassan MZ, Roslan SA, Sapuan SM, Rasid ZA, Mohd Nor AF, Md Daud MY, et al.

Polymers (Basel), 2020 Jun 17;12(6).
PMID: 32560539 DOI: 10.3390/polym12061367

The objective of this research is to optimize the alkaline treatment variables, including sodium hydroxide (NaOH) concentration, soaking, and drying time, that influence the mechanical behavior of bamboo fiber-reinforced epoxy composites. In this study, a Box-Behnken design (BBD) of the response surface methodology (RSM) was employed to design an experiment to investigate the mercerization effect of bamboo fiber-reinforced epoxy composites. The evaluation of predicted tensile strength as a variable parameter of bamboo fiber (Bambusa vulgaris) reinforced epoxy composite structures was determined using analysis of variance (ANOVA) of the quadratic model. In this study, a total of 17 experiment runs were measured and a significant regression for the coefficient between the variables was obtained. Further, the triangular and square core structures made of treated and untreated bamboo fiber-reinforced epoxy composites were tested under compressive loading. It was found that the optimum mercerization condition lies at 5.81 wt.% of the NaOH, after a soaking time of 3.99 h and a drying time of 72 h. This optimum alkaline treatment once again had a great effect on the structures whereby all the treated composite cores with square and triangular structures impressively outperformed the untreated bamboo structures. The treated triangular core of bamboo reinforced composites gave an outstanding performance compared to the treated and untreated square core composite structures for compressive loading and specific energy absorbing capability.
Adsorption mechanism and effectiveness of phenol and tannic acid removal by biochar produced from oil palm frond using steam pyrolysis

Lawal AA, Hassan MA, Ahmad Farid MA, Tengku Yasim-Anuar TA, Samsudin MH, Mohd Yusoff MZ, et al.

Environ Pollut, 2021 Jan 15;269:116197.
PMID: 33316496 DOI: 10.1016/j.envpol.2020.116197

In order to meet the growing demand for adsorbents to treat wastewater effectively, there has been increased interest in using sustainable biomass feedstocks. In this present study, the dermal tissue of oil palm frond was pyrolyzed with superheated steam at 500 °C to produce nanoporous biochar as bioadsorbent. The effect of operating conditions was investigated to understand the adsorption mechanism and to enhance the adsorption of phenol and tannic acid. The biochar had a microporous structure with a Brunauer-Emmett-Teller surface area of 422 m2/g containing low polar groups. The adsorption capacity of 62.89 mg/g for phenol and 67.41 mg/g for tannic acid were obtained using 3 g/L biochar dosage after 8 h of treatment at solution pH of 6.5 and temperature of 45 °C. The Freundlich model had the best fit to the isotherm data of phenol (R2 of 0.9863), while the Langmuir model best elucidated the isotherm data of tannic acid (R2 of 0.9632). These indicated that the biochar-phenol interface was associated with a heterogeneous multilayer sorption mechanism, while the biochar-tannic acid interface had a nonspecific monolayer sorption mechanism. The residual concentration of 26.3 mg/L phenol and 23.1 mg/L tannic acid was achieved when treated from 260 mg/L three times consecutively with 1 g/L biochar dosage, compared to a reduction to 72.3 mg/L phenol and 69.9 mg/L tannic acid using 3 g/L biochar dosage in a single treatment. The biochar exhibited effective adsorption of phenol and tannic acid, making it possible to treat effluents that contain varieties of phenolic compounds.

Filters

Please provide feedback to Administrator (afdal@afpm.org.my)

External Links