A Carbon Nanotube Sensor Array for the Label-Free Discrimination of Live and Dead Cells with Machine Learning.

Developing robust cell recognition strategies is important in biochemical research, but the lack of well-defined target molecules creates a bottleneck in some applications. In this paper, a carbon nanotube sensor array was constructed for the label-free discrimination of live and dead mammalian cells. Three types of carbon nanotube field-effect transistors were fabricated, and different features were extracted from the transfer characteristic curves for model training with linear discriminant analysis (LDA) and support-vector machines (SVM). Live and dead cells were accurately classified in more than 90% of samples in each sensor group using LDA as the algorithm. The recursive feature elimination with cross-validation (RFECV) method was applied to handle the overfitting and optimize the model, and cells could be successfully classified with as few as four features and a higher validation accuracy (up to 97.9%) after model optimization. The RFECV method also revealed the crucial features in the classification, indicating the participation of different sensing mechanisms in the classification. Finally, the optimized LDA model was applied for the prediction of unknown samples with an accuracy of 87.5-93.8%, indicating that live and dead cell samples could be well-recognized with the constructed model.

Bacterial Vaginosis Monitoring with Carbon Nanotube Field-Effect Transistors.

The ability to rapidly and reliably screen for bacterial vaginosis (BV) during pregnancy is of great significance for maternal health and pregnancy outcomes. In this proof-of-concept study, we demonstrated the potential of carbon nanotube field-effect transistors (NTFET) in the rapid diagnostics of BV with the sensing of BV-related factors such as pH and biogenic amines. The fabricated sensors showed good linearity to pH changes with a linear correlation coefficient of 0.99. The pH sensing performance was stable after more than one month of sensor storage. In addition, the sensor was able to classify BV-related biogenic amine-negative/positive samples with machine learning, utilizing different test strategies and algorithms, including linear discriminant analysis (LDA), support vector machine (SVM), and principal component analysis (PCA). The biogenic amine sample status could be well classified using a soft-margin SVM model with a validation accuracy of 87.5%. The accuracy could be further improved using a gold gate electrode for measurement, with accuracy higher than 90% in both LDA and SVM models. We also explored the sensing mechanisms and found that the change in NTFET off current was crucial for classification. The fabricated sensors successfully detect BV-related factors, demonstrating the competitive advantage of NTFET for point-of-care diagnostics of BV.

Machine learning-assisted calibration of Hg2+ sensors based on carbon nanotube field-effect transistors.

Nanomaterial-based electronic sensors have demonstrated ultra-low detection limits, down to parts-per-billion (ppb) or parts-per-trillion (ppt) concentrations. However, these extreme sensitivities also make them susceptible to signal saturation at higher concentrations and restrict their usage primarily to low concentrations. Here, we report machine learning techniques to create a calibration method for carbon nanotube-based field-effect transistor (FET) devices. We started with linear regression, followed by regression splines to capture the non-linearity in the data. Further improvements in model performance were obtained with regression trees. Finally we lowered the model variance and further boosted the model performance by introducing random forest. The resulting performance as measured by R2 was estimated to be 0.8260 using out-of-bag error. The methodology avoids saturation and extends the dynamic range of the nanosensors up to 12 orders of magnitude in analyte concentrations. Further investigations of the sensing mechanism include analysis of feature importance in each of the model we tested. Functionalized nanosensors demonstrate selective detection of Hg2+ ions with detection limits 10-14.36±0.78 M, and maintain calibration to concentrations as high as 1 mM. Application of machine learning techniques to investigate which features in the FET signal maximally correlate with concentration changes provide valuable insight into the carbon nanotube sensing mechanism and assist in the rational design of future nanosensors.

Machine-Learning Identification of the Sensing Descriptors Relevant in Molecular Interactions with Metal Nanoparticle-Decorated Nanotube Field-Effect Transistors.

Carbon nanotube-based field-effect transistors (NTFETs) are ideal sensor devices as they provide rich information regarding carbon nanotube interactions with target analytes and have potential for miniaturization in diverse applications in medical, safety, environmental, and energy sectors. Herein, we investigate chemical detection with cross-sensitive NTFETs sensor arrays comprised of metal nanoparticle-decorated single-walled carbon nanotubes (SWCNTs). By combining analysis of NTFET device characteristics with supervised machine-learning algorithms, we have successfully discriminated among five selected purine compounds, adenine, guanine, xanthine, uric acid, and caffeine. Interactions of purine compounds with metal nanoparticle-decorated SWCNTs were corroborated by density functional theory calculations. Furthermore, by testing a variety of prepared as well as commercial solutions with and without caffeine, our approach accurately discerns the presence of caffeine in 95% of the samples with 48 features using a linear discriminant analysis and in 93.4% of the samples with only 11 features when using a support vector machine analysis. We also performed recursive feature elimination and identified three NTFET parameters, transconductance, threshold voltage, and minimum conductance, as the most crucial features to analyte prediction accuracy.

Tetrahydrocannabinol Detection Using Semiconductor-Enriched Single-Walled Carbon Nanotube Chemiresistors.

Semiconductor-enriched single-walled carbon nanotubes (s-SWCNTs) have potential for application as a chemiresistor for the detection of breath compounds, including tetrahydrocannabinol (THC), the main psychoactive compound found in the marijuana plant. Herein we show that chemiresistor devices fabricated from s-SWCNT ink using dielectrophoresis can be incorporated into a hand-held breathalyzer with sensitivity toward THC generated from a bubbler containing analytical standard in ethanol and a heated sample evaporator that releases compounds from steel wool. The steel wool was used to capture THC from exhaled marijuana smoke. The generation of the THC from the bubbler and heated breath sample chamber was confirmed using ultraviolet-visible absorption spectroscopy and mass spectrometry, respectively. Enhanced selectivity toward THC over more volatile breath components such as CO2, water, ethanol, methanol, and acetone was achieved by delaying the sensor reading to allow for the desorption of these compounds from the chemiresistor surface. Additionally, machine learning algorithms were utilized to improve the selective detection of THC with better accuracy at increasing quantities of THC delivered to the chemiresistor.

Nanoelectronic Discrimination of Nonmalignant and Malignant Cells Using Nanotube Field-Effect Transistors.

Detection of malignant cells in tissue is a difficult hurdle in medical diagnostics and screening. Carbon nanotubes are extremely sensitive to their local environments, and nanotube-based field-effect transistors (NTFETs) provide a plethora of information regarding the mechanism of interaction with target analytes. Herein, we use a series of functionalized metal nanoparticle-decorated NTFET devices forming an array with multiple nonselective sensor units as the electronic "tongue", sensing all five basic tastes. By extraction of selected NTFET characteristics and using linear discriminant analysis, we have successfully detected and discriminated between malignant and nonmalignant tissues and cells. We also studied the sensing mechanism and what NTFET characteristics are responsible for the variation of response between cell types, allowing for the design of future studies such as detection of malignant cells in a biopsy or the effects of malignant cells on healthy tissue.

16S rRNA gene sequencing is a non-culture method of defining the specific bacterial etiology of ventilator-associated pneumonia.

Ventilator-associated pneumonia (VAP) is an acquired respiratory tract infection following tracheal intubation. The most common hospital-acquired infection among patients with acute respiratory failure, VAP is associated with a mortality rate of 20-30%. The standard bacterial culture method for identifying the etiology of VAP is not specific, timely, or accurate in identifying the bacterial pathogens. This study used 16S rRNA gene metagenomic sequencing to identify and quantify the pathogenic bacteria in lower respiratory tract and oropharyngeal samples of 55 VAP patients. Sequencing of the 16S rRNA gene has served as a valuable tool in bacterial identification, particularly when other biochemical, molecular, or phenotypic identification techniques fail. In this study, 16S rRNA gene sequencing was performed in parallel with the standard bacterial culture method to identify and quantify bacteria present in the collected patient samples. Sequence analysis showed the colonization of multidrug-resistant strains in VAP secretions. Further, this method identified Prevotella, Proteus, Aquabacter, and Sphingomonas bacterial genera that were not detected by the standard bacterial culture method. Seven categories of bacteria, Streptococcus, Neisseria, Corynebacterium, Acinetobacter, Staphylococcus, Pseudomonas and Klebsiella, were detectable by both 16S rRNA gene sequencing and standard bacterial culture methods. Further, 16S rRNA gene sequencing had a significantly higher sensitivity in detecting Streptococcus and Pseudomonas when compared to standard bacterial culture. Together, these data present 16S rRNA gene sequencing as a novel VAP diagnosis tool that will further enable pathogen-specific treatment of VAP.

A colorimetric method for highly sensitive and accurate detection of iodide by finding the critical color in a color change process using silver triangular nanoplates.

In this contribution, we demonstrated a novel colorimetric method for highly sensitive and accurate detection of iodide using citrate-stabilized silver triangular nanoplates (silver TNPs). Very lower concentration of iodide can induce an appreciable color change of silver TNPs solution from blue to yellow by fusing of silver TNPs to nanoparticles, as confirmed by UV-vis absorption spectroscopy and transmission electron microscopy (TEM). The principle of this colorimetric assay is not an ordinary colorimetry, but a new colorimetric strategy by finding the critical color in a color change process. With this strategy, 0.1 µM of iodide can be recognized within 30 min by naked-eyes observation, and lower concentration of iodide down to 8.8 nM can be detected using a spectrophotometer. Furthermore, this high sensitive colorimetric assay has good accuracy, stability and reproducibility comparing with other ordinary colorimetry. We believe this new colorimetric method will open up a fresh insight of simple, rapid and reliable detection of iodide and can find its future application in the biochemical analysis or clinical diagnosis.