Green approach for the fabrication of dual-functional S/N doped graphene tagged ZnO nanograins for in vitro bioimaging and water pollutant remediation.

Dual-functional S/N (sulfur and nitrogen) doped graphene-tagged zinc oxide nanograins were synthesized for bioimaging applications and light-dependent photocatalytic activity. Applying the green synthesis approach, graphene was synthesized from kimchi cabbage through a hydrothermal process followed by tagging it with synthesized zinc oxide nanoparticles (ZnO-NPs). The 2D/0D heterostructure prepared by combining both exhibited exceptional advantages. Comprehensive characterizations such as TEM, SEM, XRD, FTIR, XPS, and UV-Vis spectra have been performed to confirm the structures and explore the properties of the synthesized nanocomposite. The graphene/ZnO-NP composite produced exhibited more intense fluorescence, greater chemical stability and biocompatibility, lower cytotoxicity, and better durability than ZnO NPs conferring them with potential applications in cellular imaging. While tagging the ZnO NPs with carbon derived from a natural source containing hydroxyl, sulfur, and nitrogen-containing functional group, the S/N doped graphene/ZnO heterostructure evidences the high photocatalytic activity under UV and visible irradiation which is 3.2 and 3.8 times higher than the as-prepared ZnO-NPs. It also demonstrated significant antibacterial activity which confers its application in removing pathogenic contaminant bacteria in water bodies. In addition, the composite had better optical properties and biocompatibility, and lower toxicity than ZnO NPs. Our findings indicate that the synthesized nanocomposite will be suitable for various biomedical and pollutant remediation due to its bright light-emitting properties and stable fluorescence.

Exploiting Missing Value Patterns for a Backdoor Attack on Machine Learning Models of Electronic Health Records: Development and Validation Study.

BACKGROUND: A backdoor attack controls the output of a machine learning model in 2 stages. First, the attacker poisons the training data set, introducing a back door into the victim's trained model. Second, during test time, the attacker adds an imperceptible pattern called a trigger to the input values, which forces the victim's model to output the attacker's intended values instead of true predictions or decisions. While backdoor attacks pose a serious threat to the reliability of machine learning-based medical diagnostics, existing backdoor attacks that directly change the input values are detectable relatively easily. OBJECTIVE: The goal of this study was to propose and study a robust backdoor attack on mortality-prediction machine learning models that use electronic health records. We showed that our backdoor attack grants attackers full control over classification outcomes for safety-critical tasks such as mortality prediction, highlighting the importance of undertaking safe artificial intelligence research in the medical field. METHODS: We present a trigger generation method based on missing patterns in electronic health record data. Compared to existing approaches, which introduce noise into the medical record, the proposed backdoor attack makes it simple to construct backdoor triggers without prior knowledge. To effectively avoid detection by manual inspectors, we employ variational autoencoders to learn the missing patterns in normal electronic health record data and produce trigger data that appears similar to this data. RESULTS: We experimented with the proposed backdoor attack on 4 machine learning models (linear regression, multilayer perceptron, long short-term memory, and gated recurrent units) that predict in-hospital mortality using a public electronic health record data set. The results showed that the proposed technique achieved a significant drop in the victim's discrimination performance (reducing the area under the precision-recall curve by at most 0.45), with a low poisoning rate (2%) in the training data set. In addition, the impact of the attack on general classification performance was negligible (it reduced the area under the precision-recall curve by an average of 0.01025), which makes it difficult to detect the presence of poison. CONCLUSIONS: To the best of our knowledge, this is the first study to propose a backdoor attack that uses missing information from tabular data as a trigger. Through extensive experiments, we demonstrated that our backdoor attack can inflict severe damage on medical machine learning classifiers in practice.

Effects of postconditioning with N,N,N'N'-tetrakis-[2-pyridylmethyl]-ethylenediamine in isolated rat hearts.

BACKGROUND: It was reported that N,N,N'N'-tetrakis-[2-pyridylmethyl]-ethylenediamine (TPEN), a transition metal chelator, confers cardioprotection against myocardial ischemic injury. In this study, we investigated the effect of TPEN targeting reperfusion period in isolated rat hearts. METHODS: Langendorff perfused rat hearts were subjected to 30 min of regional ischemia and 2 h of reperfusion. Hearts were randomly assigned to either control (n = 9) or 10 microM of TPEN (n = 8) groups. TPEN was perfused for a period of 5 min before and 30 min after reperfusion. RESULTS: The ratio of infarct area/ischemic area (AN/AR) was significantly reduced in TPEN treated hearts (6.9 +/- 1.7%, P < 0.001) compared to control hearts (29.5 +/- 3.2%). Recovery of left ventricular developed pressure (LVDP), rate-pressure product (RPP), +dP/dt(max), and -dP/dt(min) in the control group after reperfusion were 53.8 +/- 6.2%, 51.0 +/- 6.3%, 51.9 +/- 5.7%, and 51.4 +/- 5.7%, respectively, of the baseline levels. In the TPEN group, LVDP, RPP, +dP/dt(max), and -dP/dt(min) returned to 58.5 +/- 4.6%, 54.8 +/- 6.4%, 61.7 +/- 4.9%, and 53.4 +/- 3.9%, respectively, of the baseline levels. There were no significant differences in the cardiodynamic variables between the two groups (P > 0.05). CONCLUSIONS: Pharmacological postconditioning with TPEN reduces myocardial infarction however, TPEN does not modify post-ischemic systolic dysfunction in isolated rat hearts.