Machine Learning-Assisted Sensor Array Based on Poly(amidoamine) (PAMAM) Dendrimers for Diagnosing Alzheimer's Disease

Artificial intelligence based advancements in nanomedicine for brain disorder management: an updated narrative review

Nanomedicines are nanoscale, biocompatible materials that offer promising alternatives to conventional treatment options for brain disorders. The recent technological developments in artificial intelligence (AI), particularly machine learning (ML) and deep learning (DL), are transforming the nanomedicine field by improving disease diagnosis, biomarker identification, prognostic assessment and disease monitoring, targeted drug delivery, and therapeutic intervention as well as contributing to computational and methodological developments. These advancements can be achieved by analysis of large clinical datasets and facilitating the design and optimization of nanomaterials for in vivo testing. Such advancement offers exciting possibilities for the improvement in the management of brain disorders, including brain cancer, Alzheimer's disease, Parkinson's disease, and multiple sclerosis, where early diagnosis, targeted delivery, and effective treatment strategies remain a great challenge. This review article provides an overview of recent advances in AI-based nanomedicine development to accelerate effective and quick diagnosis, biomarker identification, prognosis, drug delivery, methodological advancement and patient-specific therapies for managing brain disorders.

NiO/ZnO Nanocomposites for Multimodal Intelligent MEMS Gas Sensors

Gas sensor arrays designed for pattern recognition face persistent challenges in achieving high sensitivity and selectivity for multiple volatile organic compounds (VOCs), particularly under varying environmental conditions. To address these limitations, we developed multimodal intelligent MEMS gas sensors by precisely tailoring the nanocomposite ratio of NiO and ZnO components. These sensors demonstrate enhanced responses to ethylene glycol (EG) and limonene (LM) at different operating temperatures, demonstrating material-specific selectivity. Additionally, a multitask deep learning model is employed for real-time, quantitative detection of VOCs, accurately predicting their concentration and type. These results showcase the effectiveness of combining material optimization with advanced algorithms for real-world VOCs detection, advancing the field of odor analysis tools.

Machine Learning-Assisted Chemical Tongues Based on Dual-channel Inclusion Complexes for Rapid Identification of Nonsteroidal Anti-inflammatory Drugs in Food

The improper application of nonsteroidal anti-inflammatory drugs (NSAIDs) presents significant health hazards via vector food contamination. A critical limitation of these traditional existing approaches is their inability to concurrently discern and distinguish among diverse NSAIDs, presenting a notable gap in the analytical capabilities within this domain. Herein, a creative dual-channel fluorescence sensor array was developed for the rapid discrimination and determination of NSAIDs, utilizing complexes of cucurbit[8]uril (CB[8]) with three distinct modified poly(ethylenimines) (PEIs) to address this challenge. The array successfully differentiated and identified 19 NSAIDs with 97% accuracy at a concentration of 1 mM. In addition, it also achieved analyses of individual NSAIDs across a range of concentrations, NSAID mixtures, and impurities of aspirin using statistical analysis methods. More importantly, the approach effectively detected NSAIDs in complex matrices, such as milk and urine, demonstrating its potential for real-world applications.

An ensemble model of machine learning regression techniques and color spaces integrated with a color sensor: application to color-changing biochemical assays

Non-destructive color sensors are widely applied for rapid analysis of various biological and healthcare point-of-care applications. However, existing red, green, blue (RGB)-based color sensor systems, relying on the conversion to human-perceptible color spaces like hue, saturation, lightness (HSL), hue, saturation, value (HSV), as well as cyan, magenta, yellow, key (CMYK) and the CIE L*a*b* (CIELAB) exhibit limitations compared to spectroscopic methods. The integration of machine learning (ML) techniques presents an opportunity to enhance data analysis and interpretation, enabling insights discovery, prediction, process automation, and decision-making. In this study, we utilized four different regression models integrated with an RGB sensor for colorimetric analysis. Colorimetric protein concentration assays, such as the bicinchoninic acid (BCA) assay and the Bradford assay, were chosen as model studies to evaluate the performance of the ML-based color sensor. Leveraging regression models, the sensor effectively interprets and processes color data, facilitating precision color detection and analysis. Furthermore, the incorporation of diverse color spaces enhances the sensor's adaptability to various color perception models, promising precise measurement, and analysis capabilities for a range of applications.

Machine learning-assisted pattern recognition and imaging of multiplexed cancer cells via a porphyrin-embedded dendrimer array

Early cancer detection plays a vital role in improving the survival rate of cancer patients, underscoring the importance of developing cancer detection methods. However, it is a great challenge to achieve simple, rapid, and accurate methods for simultaneously discerning various cancers. Herein we developed a 5-element porphyrin-embedded dendrimer-based sensor array, targeting the parallel discrimination of multiple cancers. The porphyrin-embedded dendrimers were modified with various functional groups to generate differentiated interactions with diverse cancer cells, which has been validated by fluorescence responses and laser confocal microscopy imaging. The dual-channel, five-element array, featuring ten signal outputs, achieved 100% accuracy in distinguishing between one human normal cell and six human cancerous cells, as well as in differentiating among mixed cells. Moreover, the screen 6-channel array can accurately distinguish 9 cells from mice and humans in minutes through optimization by multiple machine learning algorithms, including two normal cells and 7 cancerous cells with only 1000 cells, highlighting the significant potential of a porphyrin-embedded dendrimer-based parallel discriminating platform in early cancer diagnosis.

Tuning multispectral fluorescence quantum dot-based identification of short-length amyloid β peptides by applying Cu(II) ions

Currently available methods for detecting amyloid β (Aβ) derivatives are mainly dedicated to determining the long forms Aβ1-42 and Aβ1-40. At the same time, the number of physiologically occurring Aβ analogs is much higher, including those truncated at the N- and C-termini. Their identification using standard methods is challenging due to the structural similarity of various Aβ analogs, but could highly benefit from both biomarkers discovery and pathophysiological studies of Alzheimer's disease. Therefore a "chemical tongue" sensing strategy was employed for the detection of seven Aβ peptide derivatives: Aβ1-16, Aβ4-16, Aβ4-9, Aβ5-16, Aβ5-12, Aβ5-9, Aβ12-16. The proposed sensing system is based on competitive interactions between quantum dots, Cu(II) ions, and Aβ peptides, providing unique fluorescence fingerprints useful for the identification of analytes. After carefully evaluating the Aβ sample preparation protocol, perfect determination of all studied Aβ peptides was achieved using partial least square-discriminant analysis (PLS-DA). The developed PLS-DA models are characterized by excellent accuracy, sensitivity, precision, and specificity of analyte determination, emphasizing the potential of the proposed sensing strategy.

Luminescent Metal-Organic Framework-Based Fluorescent Sensor Array for Screening and Discrimination of Bisphenols

Extensive applications of bisphenols in industrial products have led to their release into aquatic environments, causing a great threat to human health due to their endocrine-disrupting effects, whereas existing methods are difficult to implement the rapid and high-throughput detection of multiple bisphenols. To circumvent this issue, we constructed a sensor array using two luminescent metal-organic frameworks (LMOFs) (Zr-BUT-12 and Ga-MIL-61) for the rapid discrimination of six bisphenol contaminants (BPA, BPS, BPB, BPF, BPAF, and TBBPA). Wherein, Zr-BUT-12 and Ga-MIL-61 exhibited different fluorescence-emission properties and good luminescent stability. Interestingly, bisphenols with different structures had diverse quenching effects on the fluorescence intensity of Zr-BUT-12 and Ga-MIL-61 via the adsorptive interaction, resulting in unique fluorescent fingerprints. Based on pattern recognition methods, different bisphenols were successfully identified, with the limit of detection in the range of 1.59-16.7 ng/mL for six bisphenols. More importantly, the developed sensor array could be effectively utilized for distinguishing different ratios of mixed bisphenols, which was further applied for bisphenol discrimination in real water samples. Consequently, our finding provides a promising strategy for the simultaneous recognition of multiple bisphenols, which encourages the development of a sensor array for the detection of multiple contaminants in environmental monitoring and food safety.

Identification of Microorganism in Infected Wounds by Positively Charged Selective Sensor Array and Deep Learning Algorithm

Microorganism are ubiquitous and intimately connected with human health and disease management. The accurate and fast identification of pathogenic microorganisms is especially important for diagnosing infections. Herein, three tetraphenylethylene derivatives (S-TDs: TBN, TPN, and TPI) featuring different cationic groups, charge numbers, emission wavelengths, and hydrophobicities were successfully synthesized. Benefiting from distinct cell wall binding properties, S-TDs were collectively utilized to create a sensor array capable of imaging various microorganisms through their characteristic fluorescent signatures. Furthermore, the interaction mechanism between S-TDs and different microorganisms was explored by calculating the binding energy between S-TDs and cell membrane/wall constituents, including phospholipid bilayer and peptidoglycan. Using a combination of the fluorescence sensor array and a deep learning model of residual network (ResNet), readily differentiation of Gram-negative bacteria (G-), Gram-positive bacteria (G+), fungi, and their mixtures was achieved. Specifically, by extensive training of two ResNet models with large quantities of images data from 14 kinds of microorganism stained with S-TDs, identification of microorganism was achieved at high-level accuracy: over 92.8% for both Gram species and antibiotic-resistant species, with 90.35% accuracy for the detection of mixed microorganism in infected wound. This novel method provides a rapid and accurate method for microbial classification, potentially aiding in the diagnosis and treatment of infectious diseases.

Peptide Dendrimer-Based Antibacterial Agents: Synthesis and Applications

Pathogenic bacteria cause the deaths of millions of people every year. With the development of antibiotics, hundreds and thousands of people's lives have been saved. Nevertheless, bacteria can develop resistance to antibiotics, rendering them insensitive to antibiotics over time. Peptides containing specific amino acids can be used as antibacterial agents; however, they can be easily degraded by proteases in vivo. To address these issues, branched peptide dendrimers are now being considered as good antibacterial agents due to their high efficacy, resistance to protease degradation, and low cytotoxicity. The ease with which peptide dendrimers can be synthesized and modified makes them accessible for use in various biological and nonbiological fields. That is, peptide dendrimers hold a promising future as antibacterial agents with prolonged efficacy without bacterial resistance development. Their in vivo stability and multivalence allow them to effectively target multi-drug-resistant strains and prevent biofilm formation. Thus, it is interesting to have an overview of the development and applications of peptide dendrimers in antibacterial research, including the possibility of employing machine learning approaches for the design of AMPs and dendrimers. This review summarizes the synthesis and applications of peptide dendrimers as antibacterial agents. The challenges and perspectives of using peptide dendrimers as the antibacterial agents are also discussed.

Intelligent Identification of Cerebrospinal Fluid for the Diagnosis of Parkinson's Disease

Cerebrospinal fluid (CSF) biomarkers are more sensitive than the Movement Disorder Society (MDS) criteria for detecting prodromal Parkinson's disease (PD). Early detection of PD provides the best chance for successful implementation of disease-modifying treatments, making it crucial to effectively identify CSF extracted from PD patients or normal individuals. In this study, an intelligent sensor array was built by using three metal-organic frameworks (MOFs) that exhibited varying catalytic kinetics after reacting with potential protein markers. Machine learning algorithms were used to process fingerprint response patterns, allowing for qualitative and quantitative assessment of the proteins. The results were robust and capable of discriminating between PD and non-PD patients via CSF detection. The k-nearest neighbor regression algorithm was used to predict MDS scores with a minimum mean square error of 38.88. The intelligent MOF sensor array is expected to promote the detection of CSF biomarkers due to its ability to identify multiple targets and could be used in conjunction with MDS criteria and other techniques to diagnose PD more sensitively and selectively.

Temperature-modulated sensing characteristics of ultrafine Au nanoparticle-loaded porous ZnO nanobelts for identification and determination of BTEX

The identification and determination of benzene, toluene, ethylbenzene, and xylene (BTEX) has always been a formidable challenge for chemiresistive metal oxide sensors owing to their structural similarity and low reactivity, as well as the intrinsic cross sensitivity of metal oxides. In this paper, a temperature-modulated sensing strategy is proposed for the identification and determination of BTEX using a high-performance chemiresistive sensor. Ultrafine Au nanoparticle-loaded porous ZnO nanobelts as sensing materials were synthesized through an exchange reaction followed by thermal oxidation, which exhibited high response toward BTEX. Under dynamic modulation of working temperature, the distinguishable characteristic curves were demonstrated for each BTEX compound. By employing the linear discrimination and convolutional neural network analyses, highly effective BTEX identification was achieved among all investigated volatile organic compounds, which is difficult to realize for single chemiresistive sensors at constant working temperatures. Furthermore, quantitative analysis of BTEX concentrations was accomplished by establishing the relationship between concentration and response at specific points on their response curves. This developed strategy is expected to pave a new way for constructing highly sensitive gas sensors for the identification and analysis of hazardous gases, thereby enhancing their applicability in environmental monitoring.

Advances in targeted tracking and detection of soluble amyloid-β aggregates as a biomarker of Alzheimer's disease

Misfolding and aggregation of amyloid-β (Aβ) peptides are key hallmarks of Alzheimer's disease (AD). With accumulating evidence suggesting that different Aβ species have varied neurotoxicity and implications in AD development, the discovery of affinity ligands and analytical approaches to selective distinguish, detect, and monitor Aβ becomes an active research area. Remarkable advances have been achieved, which not only promote our understanding of the biophysical chemistry of the protein aggregation during neurodegeneration, but also provide promising tools for early detection of the disease. In view of this, we summarize the recent progress in selective and sensitive approaches for tracking and detection of Aβ species. Specific attentions are given to soluble Aβ oligomers, due to their crucial roles in AD development and occurrence at early stages. The design principle, performance of targeting units, and their cooperative effects with signal reporters for Aβ analysis are discussed. The applications of the novel targeting probes and sensing systems for dynamic monitoring oligomerization, measuring Aβ in biosamples and in vivo imaging in brain are summarized. Finally, the perspective and challenges are discussed regarding the future development of Aβ-targeting analytical tools to explore the unknown field to contribute to the early diagnosis and treatment of AD.

Machine learning-assisted fluorescence sensor array for qualitative and quantitative analysis of pyrethroid pesticides

The simultaneous detection of multiple residues of pyrethroid pesticides (PPs) on vegetables and fruits is still challenging using traditional nanosensing methods due to the high structural similarity of PPs. In this work, sensor arrays composed of three nanocomposite complexes (rhodamine B-CD@Au, rhodamine 6G-CD@Au, and coumarin 6-CD@Au) were constructed to discriminate between structurally similar PPs. Four PPs, deltamethrin, fenvalerate, cyfluthrin, and fenpropathrin, were successfully discriminated. The ability of these sensor units was derived from the different affinity between receptor/analyte and receptor/dye, as well as the non-linear relationship between fluorescence signal and analyte concentration. Upon multivariate pattern recognition analysis, the array performed high-throughput identification of four PPs in unknown samples with 100% classification accuracy. In addition, good accuracy of predicting concentration using the "stepwise prediction" strategy combined with the machine learning method was achieved.

Quantum dots-based "chemical tongue" for the discrimination of short-length Aβ peptides

A "chemical tongue" is proposed based on thiomalic acid-capped quantum dots (QDs) with signal enrichment provided by excitation-emission matrix (EEM) fluorescence spectroscopy for the determination of close structural analogs-short-length amyloid β (Aβ) peptides related to Alzheimer's disease. Excellent discrimination is obtained by principal component analysis (PCA) for seven derivatives: Aβ1-16, Aβ4-16, Aβ4-9, Aβ5-16, Aβ5-12, Aβ5-9, Aβ12-16. Detection of Aβ4-16, Aβ4-16, and Aβ5-9 in binary and ternary mixtures performed by QDs-based chemical tongue using partial least squares-discriminant analysis (PLS-DA) provided perfect 100% accuracy for the two studied peptides (Aβ4-16 and Aβ4-16), while for the third one (Aβ5-9) it was slightly lower (97.9%). Successful detection of Aβ4-16 at 1 pmol/mL (1.6 ng/mL) suggests that the detection limit of the proposed method for short-length Aβ peptides can span nanomolar concentrations. This result is highly promising for the development of simple and efficient methods for sequence recognition in short-length peptides and better understanding of mechanisms at the QD-analyte interface.

Machine-learning assisted multicolor platform for multiplex detection of antibiotics in environmental water samples

Antibiotic (AB) resistance is one of daunting challenges of our time, attributed to overuse of ABs and usage of AB-contaminated food resources. Due to their detrimental impact on human health, development of visual detection methods for multiplex sensing of ABs is a top priority. In present study, a colorimetric sensor array consisting of two types of gold nanoparticles (AuNPs) were designed for identification and determination of ABs. Design principle of the probe was based on aggregation of AuNPs in the presence of ABs at different buffer conditions. The utilization of machine learning algorithms in this design enables classification and quantification of ABs in various samples. The response profile of the array was analyzed using linear discriminant analysis algorithm for classification of ABs. This colorimetric sensor array is capable of accurate distinguishing between individual ABs and their combinations. Partial least squares regression was also applied for quantitation purposes. The obtained analytical figures of merit demonstrated the potential applicability of the developed sensor array in multiplex detection of ABs. The response profiles of the array were linearly correlated to the concentrations of ABs in a wide range of concentration with limit of detections of 0.05, 0.03, 0.04, 0.01, 0.06, 0.05 and 0.04 μg.mL-1 for azithromycin, amoxicillin, ciprofloxacin, clindamycin, cefixime, doxycycline and metronidazole respectively. The practical applicability of this method was further investigated by analysis of mixture samples of ABs and determination of ABs in river and underground water with successful verification.

Electrochromic Sensor Augmented with Machine Learning for Enzyme-Free Analysis of Antioxidants

Our study presents an electrochromic sensor that operates without the need for enzymes or multiple oxidant reagents. This sensor is augmented with machine learning algorithms, enabling the identification, classification, and prediction of six different antioxidants with high accuracy. We utilized polyaniline (PANI), Prussian blue (PB), and copper-Prussian blue analogues (Cu-PBA) at their respective oxidation states as electrochromic materials (ECMs). By designing three readout channels with these materials, we were able to achieve visual detection of antioxidants without relying on traditional "lock and key" specific interactions. Our sensing approach is based on the direct electrochemical reactions between oxidized electrochromic materials (ECMsox) as electron acceptors and various antioxidants, which act as electron donors. This interaction generates unique fingerprint patterns by switching the ECMsox to reduced electrochromic materials (ECMsred), causing their colors to change. Through the application of density functional theory (DFT), we demonstrated the molecular-level basis for the distinct multicolor patterns. Additionally, machine learning algorithms were employed to correlate the optical patterns with RGB data, enabling complex data analysis and the prediction of unknown samples. To demonstrate the practical applications of our design, we successfully used the EC sensor to diagnose antioxidants in serum samples, indicating its potential for the on-site monitoring of antioxidant-related diseases. This advancement holds promise for various applications, including the real-time monitoring of antioxidant levels in biological samples, the early diagnosis of antioxidant-related diseases, and personalized medicine. Furthermore, the success of our electrochromic sensor design highlights the potential for exploring similar strategies in the development of sensors for diverse analytes, showcasing the versatility and adaptability of this approach.

Machine Learning-Assisted Nanoenzyme/Bioenzyme Dual-Coupled Array for Rapid Detection of Amyloids

Array-based sensing methods offer significant advantages in the simultaneous detection of multiple amyloid biomarkers and thus have great potential for diagnosing early-stage Alzheimer's disease. Yet, detecting low concentrations of amyloids remains exceptionally challenging. Here, we have developed a fluorescent sensor array based on the dual coupling of a nanoenzyme (AuNPs) and bioenzyme (horseradish peroxidase) to detect amyloids. Various ss-DNAs were bound to the nanoenzyme for regulating enzymatic activity and recognizing amyloids. A simplified sensor array was generated from a screening model via machine learning algorithms and achieved signal amplification through a two-step enzymatic reaction. As a result, our sensing system could discriminate the aggregation species and aggregation kinetics at 200 nM with 100% accuracy. Moreover, AD model mice and healthy mice were distinguished with 100% accuracy through the sensor array, providing a powerful sensing platform for diagnosing AD.

A Multichannel Fluorescent Tongue for Amyloid-β Aggregates Detection

Attention has been paid to the early diagnosis of Alzheimer's disease, due to the maximum benefit acquired from the early-stage intervention and treatment. However, the sensing techniques primarily depended upon for neuroimaging and immunological assays for the detection of AD biomarkers are expensive, time-consuming and instrument dependent. Here, we developed a multichannel fluorescent tongue consisting of four fluorescent dyes and GO through electrostatic and π-π interaction. The array distinguished multiple aggregation states of 1 µM Aβ40/Aβ42 with 100% prediction accuracy via 10-channel signal outputs, illustrating the rationality of the array design. Screening vital sensor elements for the simplified sensor array and the optimization of sensing system was achieved by machine learning algorithms. Moreover, our sensing tongue was able to detect the aggregation states of Aβ40/Aβ42 in serum, demonstrating the great potential of multichannel array in diagnosing the Alzheimer's diseases.