Prognostic prediction of sepsis patient using transformer with skip connected token for tabular data.

Sepsis is known as a common syndrome in intensive care units (ICU), and severe sepsis and septic shock are among the leading causes of death worldwide. The purpose of this study is to develop a deep learning model that supports clinicians in efficiently managing sepsis patients in the ICU by predicting mortality, ICU length of stay (>14 days), and hospital length of stay (>30 days). The proposed model was developed using 591 retrospective data with 16 tabular data related to a sequential organ failure assessment (SOFA) score. To analyze tabular data, we designed the modified architecture of the transformer that has achieved extraordinary success in the field of languages and computer vision tasks in recent years. The main idea of the proposed model is to use a skip-connected token, which combines both local (feature-wise token) and global (classification token) information as the output of a transformer encoder. The proposed model was compared with four machine learning models (ElasticNet, Extreme Gradient Boosting [XGBoost]), and Random Forest) and three deep learning models (Multi-Layer Perceptron [MLP], transformer, and Feature-Tokenizer transformer [FT-Transformer]) and achieved the best performance (mortality, area under the receiver operating characteristic (AUROC) 0.8047; ICU length of stay, AUROC 0.8314; hospital length of stay, AUROC 0.7342). We anticipate that the proposed model architecture will provide a promising approach to predict the various clinical endpoints using tabular data such as electronic health and medical records.

Automated diagnosis of atrial fibrillation using ECG component-aware transformer.

Atrial fibrillation (AF) is the most common sustained arrhythmia worldwide and imposes a substantial economic burden on the public healthcare system due to its high morbidity and mortality. Early detection of AF is crucial in providing timely treatment and preventing complications such as stroke and other thromboembolism. For AF diagnosis, the 12-lead electrocardiogram (ECG) has been established as the gold standard. However, it requires the clinical experiences of cardiologists and may be vulnerable to inter-observer variability. Although automated AF diagnostic techniques based on deep neural networks (DNN) have been proposed, most studies were conducted using small-scale datasets, resulting in the over-fitting problem. Furthermore, they have not fully exploited ECG components such as P-wave, QRS-complex, and T-wave contrary to the approach adopted by cardiologists who interpret ECG by considering its components. To overcome these limitations, this study presents the component-aware transformer (CAT), which segments the ECG waveform into each component, vectorizes them with length and types information into one vector, and used it as the input of the transformer. We conducted extensive experiments to evaluate the CAT using a large-scale dataset called Shaoxing Hospital Zhejiang University School of Medicine database (AF: 1,780 cases, non-AF: 8,866 cases). The quantitative evaluations demonstrate that the CAT outperforms the conventional deep learning techniques on both single- and 12-lead ECG signals. Moreover, the CAT trained on single-lead ECG is comparable to that of a 12-lead analysis, while conventional methods degraded significantly in performance. Consequently, the CAT is applicable to various single-channel signals such as airway pressure, photoplethysmogram, and blood pressure.

A multi-functional picolinohydrazide-based chemosensor for colorimetric detection of iron and dual responsive detection of hypochlorite.

A novel effective chemosensor HPHN, (E)-6-hydroxy-N'-((2-hydroxynaphthalen-1-yl)methylene) picolinohydrazide, was synthesized. HPHN sensed Fe3+/2+ with the changes of color from yellow to orange without obvious inhibition from other cations. In addition, HPHN could detect ClO- by both the color change from yellow to colorless and the fluorescence quenching. The binding modes of HPHN with Fe3+/2+ and ClO- were determined to be 1:1 with Job plot and ESI-mass analysis. HPHN displayed low detection limits of 0.29 µM for Fe3+ and 0.77 µM for Fe2+. For ClO-, the detection limit was 6.20 µM by colorimetric method and 3.99 µM by fluorescent one, respectively. Moreover, HPHN can be employed to quantify Fe3+ and ClO- in environmental samples and apply to cell imaging for ClO-.

A visible chemosensor based on carbohydrazide for Fe(ii), Co(ii) and Cu(ii) in aqueous solution.

A colorimetric sensor with pyridyl and carbohydrazide components has been synthesized and visibly turns blue in the presence of Fe(ii). The colorless sensor also changes color when exposed to Co(ii) and Cu(ii), but its color becomes yellow. The sensor shows no visible response to other metal ions such Ca2+, Cr3+, Mn2+, Fe3+, Ni2+, Zn2+, Cd2+, Ag+, Hg2+, and Pb2+. The binding ratio of the sensor to Fe(ii), Co(ii), and Cu(ii) is 2 sensors to 1 metal ion. The binding constants of the sensor are: Fe(ii): 1.0 × 109 M-2, Co(ii): 2 × 109 M-2, and Cu(ii): 3.0 × 109 M-2. The sensor works well at neutral pH and micromolar concentrations of Fe(ii), Co(ii), and Cu(ii) can be detected in water samples. The sensor's color response to Cu(ii) is uniquely attenuated by glutathione.

A fluorescent and colorimetric Schiff base chemosensor for the detection of Zn2+ and Cu2+: Application in live cell imaging and colorimetric test kit.

A novel Schiff base chemosensor HMID, ((E)­1­((2­hydroxy­3­methoxybenzylidene)amino)imidazolidine­2,4­dione), have been designed and synthesized. Sensor HMID showed a selectivity to Zn2+ through fluorescence enhancement in aqueous solution. Its detection limit was analyzed as 11.9 µM. Importantly, compound HMID could be applied to image Zn2+ in live cells. Detection mechanism of Zn2+ by HMID was suggested to be an effect of chelation-enhanced fluorescence (CHEF) by DFT calculations. Moreover, HMID could detect Cu2+ with a change of color from colorless to pink. The selective detection mechanism of Cu2+ by HMID was demonstrated to be the promotion of intramolecular charge transfer band by DFT calculations. Additionally, HMID could be employed as a naked-eye colorimetric kit for Cu2+. Therefore, HMID has the ability as a 'single sensor for dual targets'.

A dual target chemosensor for the fluorometric detection of In3+ and colorimetric detection of Fe3.

A new dual target chemosensor 1, 1,1'­((1E,1'E)­((thiobis(ethane­2,1­diyl))bis(azanylylidene))bis(methanylylidene))bis(naphthalen­2­ol), was prepared by the reaction of a hydroxy-naphthaldehyde and a thiobis(ethylamine). Sensor 1 detected In3+ with turn-on fluorescence and Fe3+ via the change of color from colorless to pale violet. The sensing behaviors of 1 toward In3+ and Fe3+ were studied through photophysical experiments, ESI-mass, NMR titration, and theoretical calculations. In particular, 1 can discriminate In3+ from Al3+ and Fe3+ from Fe2+. Limits of detection for the analysis of In3+ and Fe3+ ions turned out to be 5.89 µM and 0.30 µM, respectively. In addition, sensor 1 functioned practically as a naked-eye test strip for Fe3+ and could be recycled by using EDTA for In3+ and DFO for Fe3+.