Early-Stage Lung Cancer Diagnosis by Deep Learning-Based Spectroscopic Analysis of Circulating Exosomes

Deciphering the molecular fingerprint of haemoglobin in lung cancer: A new strategy for early diagnosis using two-trace two-dimensional correlation near infrared spectroscopy (2T2D-NIRS) and machine learning techniques

Lung cancer remains one of the deadliest malignancies worldwide, highlighting the need for highly sensitive and minimally invasive early diagnostic methods. Near-infrared spectroscopy (NIRS) offers unique advantages in probing molecular vibrational information from blood, effectively capturing potential structural changes in haemoglobin (Hb) in lung cancer patients. In this study, we address the challenge of detecting subtle Hb features within the broader blood matrix and introduce an innovative two-stage spectral analysis framework. First, continuous wavelet transform (CWT) is employed to enhance spectral resolution and reinforce the key absorption bands of Hb. Subsequently, two-trace two-dimensional correlation spectroscopy (2T2D-COS) is applied to examine the fine vibrational differences-in both synchronous and asynchronous spectra-between lung cancer patients and healthy controls, revealing alterations in Hb secondary structures (e.g., α-helices and β-sheets). Results show that critical Hb-related peaks at 4862 cm-1, 4615 cm-1, and 4432 cm-1 undergo significant changes in lung cancer samples. Furthermore, combining these refined spectral features with machine learning classifiers (e.g., support vector machines) achieves an overall accuracy of 97.50 % and a sensitivity of 100.00 %. This work not only confirms the value of NIRS in detecting protein-level molecular information in blood but also presents a promising, efficient spectroscopic strategy for early lung cancer diagnosis, offering broad biomedical applicability.

Recent advances in bioreceptor-based sensing for extracellular vesicle analysis

Extracellular vesicles (EVs) are nanoscale, membrane-bound structures secreted by various cell types into biofluids. They show great potential as biomarkers for disease diagnostics, owing to their ability to carry molecular cargo that reflects their cellular origin. However, the inherent heterogeneity of EVs in terms of size, composition, and source presents significant challenges for reliable detection and analysis. Recent advances in bioreceptor-based biosensor technologies provide promising solutions by offering high sensitivity and specificity in EV detection and characterization. These technologies address the limitations of conventional methods, such as ultracentrifugation and bulk analysis. Biosensors utilizing antibodies, aptamers, peptides, lectins, and molecularly imprinted polymers enable precise detection of EV subpopulations by targeting specific EV surface markers, including proteins, lipids, and glycans. Additionally, these biosensors support multiplexed and real-time analysis while preserving the structural integrity of EVs. This review highlights the transformative potential of combining modern biosensing tools with bioreceptor technologies to advance EV research and diagnostics, paving the way for innovations in disease diagnostics and therapeutic monitoring.

Wearable SERS devices in health management: Challenges and prospects

Surface-Enhanced Raman Scattering (SERS) is an advanced analytical technique renowned for its heightened sensitivity in detecting molecular vibrations. Its integration into wearable technologies facilitates the monitoring of biofluids, such as sweat and tears, enabling continuous, non-invasive, real-time analysis of human chemical and biomolecular processes. This capability underscores its significant potential for early disease detection and the advancement of personalized medicine. SERS has attracted considerable research attention in the fields of wearable flexible sensing and point-of-care testing (POCT) within medical diagnostics. Nonetheless, the integration of SERS with wearable technology presents several challenges, including device miniaturization, reliable biofluid sampling, user comfort, biocompatibility, and data interpretation. The ongoing advancements in nanotechnology and artificial intelligence are instrumental in addressing these challenges. This review provides a comprehensive analysis of design strategies for wearable SERS sensors and explores their applications within this domain. Finally, it addresses the current challenges in this area and the future prospects of combining SERS wearable sensors with other portable health monitoring systems for POCT medical diagnostics. Wearable SERS is a promising innovation in future healthcare, potentially enhancing individual health outcomes and reducing healthcare costs by fostering preventive health management approaches.

Deep learning enabled open-set bacteria recognition using surface-enhanced Raman spectroscopy

Accurate bacterial identification is vital in medical and healthcare settings. Traditional methods, though reliable, are often time-consuming, underscoring the need for faster, more efficient alternatives. Deep learning-assisted Surface-enhanced Raman spectroscopy (SERS) offers a rapid and sensitive method, demonstrating high accuracy in bacterial identification. However, current deep learning models for bacterial SERS spectra classification typically operate under a closed-set paradigm, limiting their effectiveness when encountering bacterial species outside the training set. In response to this challenge, we propose a transformer-based neural network for open-set bacterial recognition using SERS spectra. Our model utilizes a combination of classification and reconstruction tasks, rejecting unknown species by analyzing reconstruction errors. Experimental results show that the proposed model outperforms traditional open-set recognition approaches, providing superior accuracy in both classifying known species and rejecting unknown ones. This study addresses the limitations of existing closed-set methods, improving the robustness of bacterial identification in real-world scenarios and demonstrating the potential of integrating SERS with transformer models for medical and healthcare applications.

Microfluidics-based label-free SERS profiling of exosomes with machine learning for osteosarcoma diagnosis

Osteosarcoma (OS) calls for early diagnosis to significantly improve patient survival rates. Exosomes hold significant potential as noninvasive biomarkers for the early diagnosis of cancer. Here, we design a microfluidic device to purify and analyze plasma-derived exosomes by label-free surface-enhanced Raman spectroscopy (SERS) profiling for OS diagnosis. Exosomes were isolated, purified, and enriched using a size-dependent microfluidic chip with tangential flow filtration, achieving a high recovery rate of 82 %. The isolated exosomes were then analyzed by label-free SERS using a nanoarray chip with self-assembly monolayers of gold nanoparticles (GNPs). Exosomes originating from different OS cell types were differentiated based on the intrinsic SERS signals. Our approach was further employed to analyze the plasma-derived exosomes from healthy donors and OS patients without the need for specific biomarker labeling. A machine learning-based diagnostic model for OS was constructed, achieving an accuracy of 93 %. The findings indicate that our method is valuable for noninvasive and precise diagnosis of OS and could be generalized to other diseases in the future.

Role of artificial intelligence in early identification and risk evaluation of non-communicable diseases: a bibliometric analysis of global research trends

OBJECTIVE

This study aims to shed light on the transformative potential of artificial intelligence (AI) in the early detection and risk assessment of non-communicable diseases (NCDs).

STUDY DESIGN

Bibliometric analysis.

SETTING

Articles related to AI in early identification and risk evaluation of NCDs from 2000 to 2024 were retrieved from the Scopus database.

METHODS

This comprehensive bibliometric study focuses on a single database, Scopus and employs narrative synthesis for concise yet informative summaries. Microsoft Excel V.365 and VOSviewer software (V.1.6.20) were used to summarise bibliometric features.

RESULTS

The study retrieved 1745 relevant articles, with a notable surge in research activity in recent years. Core journals included Scientific Reports and IEEE Access, and core institutions included the Harvard Medical School and the Ministry of Education of the People's Republic of China, while core countries comprised China, the USA, India, the UK and Saudi Arabia. Citation trends indicated substantial growth and recognition of AI's impact on NCDs management. Frequent author keywords identified key research hotspots, including specific NCDs like Alzheimer's disease and diabetes. Risk assessment studies demonstrated improved predictions for heart failure, cardiovascular risk, breast cancer, diabetes and inflammatory bowel disease.

CONCLUSION

Our findings highlight the increasing role of AI in early detection and risk prediction of NCDs, emphasising its widening research impact and future clinical potential.

Precision screening with sequential multi-algorithm reclassification technique (SMART): Saving bladders from unnecessary cystectomy

Bladder cancer, when diagnosed at an advanced stage, often necessitates inevitable invasive intervention. Consequently, non-invasive biosensor-based cancer detection and AI-based precision screening are being actively employed. However, the misclassification of cancer patients as normal-referred to as false negatives-remains a significant concern, as it could lead to fatal outcomes in lifespan. Moreover, while ensemble techniques such as soft voting and other methods can improve model accuracy and reduce misclassification, their effectiveness is limited and not applicable to all diagnostic tasks. Here, we developed a double stage cancer screening system that utilizes a sensitive urinary electrical biosensor implemented with an AI model and XAI interpretation tools. This system is designed for screening bladder cancer, well-known for its notable recurrence and high tendency to advance from non-invasive muscle tumors to muscle-invasive tumors. Four urinary biomarkers (CK8, CK18, PD-1, PD-L1) were measured by a field-effect transistor biosensor, and along with gender and age information, patients underwent initial screening by the CatBoost classification model. Patients initially classified as normal were reclassified using local explanations from neural networks offering a different perspective than CatBoost. After the second-stage screening, all of the false negatives from the initial screening could be correctly reclassified as cancer patients. Furthermore, global explanation guides the improvement of the AI model to be trained on an appropriate set of biomarker features to achieve high accuracy.

Microcapillary-Derived Plasmonic-Enhanced Cluster through the Self-Assembly Process for Breast Cancer Diagnosis

Artificial intelligence (AI)-based surface-enhanced Raman scattering (SERS) is a powerful system for cancer diagnosis, leveraging its unique advantages by combining the high sensitivity of the SERS technique with the advanced classification capabilities provided by computing power. While previous studies have yielded significant results through using exosomes, miRNA, and phenotypic biomarkers for detecting breast cancer, these methods frequently entail time-consuming and complex pretreatment steps, demanding highly skilled handling. Here, we present a free-label SERS platform with faster sampling without any pretreat using blood plasma for breast cancer diagnosis. In this study, a cluster structure of gold nanoparticles within a confines space of microcapillary was fabricated to generate close-packing nanoparticles for enhancing electromagnetic field and large number of "hot spot." We demonstrate that our SERS platform can significantly amplify the Raman signal through standard chemical detection of R6G molecules. Consequently, a solution mixed appropriately between blood plasma collected from participants with gold nanoparticles to build the hybrid cluster in the microcapillary for SERS measurement. With the support of a machine learning model, the breast cancer diagnosis has successfully classified between patients and normal participants with a high accuracy of 87.5%.

Emerging biomedical applications of surface-enhanced Raman spectroscopy integrated with artificial intelligence and microfluidic technologies

The integration of surface-enhanced Raman spectroscopy (SERS), artificial intelligence (AI), and microfluidics represent a transformative approach for biomedical applications. By combining the molecular sensitivity of SERS, AI-driven spectral analysis, and the precise sample handling of microfluidics, these novel integrated systems enable ultrasensitive, label-free diagnostics with minimal sample processing. The development of portable, cost-effective platforms could democratize advanced diagnostics for resource-limited settings. However, challenges such as reproducibility, clinical validation, and system integration hinder widespread adoption. This review explores these new integrated platforms, beginning with a discussion of SERS principles, their biomedical applications, and the critical roles of AI and microfluidics in enhancing analytical performance. We evaluate recent advances in the application of these integrated systems, while addressing key challenges such as substrate scalability, biocompatibility, and point-of-care translation, with a focus on nanomaterials, AI models, and lab-on-chip designs. Finally, we outline future directions, including multimodal sensing, sustainable materials, and embedded AI for real-time diagnostics, to bridge the gap between technological innovation and clinical implementation.

Stability Dynamics of Plant-Based Extracellular Vesicles Drug Delivery

Plant-based extracellular vesicles (PBEVs) have been recognized for their wide range of applications in drug delivery however, the extent of their medicinal applicability depends on how well they are preserved and stored. Assessing their physicochemical properties, such as size, particle concentration, shape, and the activity of their cargo, forms the foundation for determining their stability during storage. Moreover, the evaluation of PBEVs is essential to ensure both safety and efficacy, which are critical for advancing their clinical development. Maintaining the biological activity of EVs during storage is a challenging task, similar to the preservation of cells and other cell-derived products like proteins. However, despite limited studies, it is expected that storing drug-loaded EVs may present fewer challenges compared to cell-based therapies, although some limitations are inevitable. This article provides a comprehensive overview of current knowledge on PBEVs preservation and storage methods, particularly focusing on their role as drug carriers. PBEVs hold promise as potential candidates for oral drug administration due to their effective intestinal absorption and ability to withstand both basic and acidic environments. However, maintaining their preservation and stability during storage is critical. Moreover, this review centers on the isolation, characterization, and storage of PBEVs, exploring the potential advantages they offer. Furthermore, it highlights key areas that require further research to overcome existing challenges and enhance the development of effective preservation and storage methods for therapeutic EVs.

Metasurface-enhanced biomedical spectroscopy

Enhancing the sensitivity of biomedical spectroscopy is crucial for advancing medical research and diagnostics. Metasurfaces have emerged as powerful platforms for enhancing the sensitivity of various biomedical spectral detection technologies. This capability arises from their unparalleled ability to improve interactions between light and matter through the localization and enhancement of light fields. In this article, we review representative approaches and recent advances in metasurface-enhanced biomedical spectroscopy. We provide a comprehensive discussion of various biomedical spectral detection technologies enhanced by metasurfaces, including infrared spectroscopy, Raman spectroscopy, fluorescence spectroscopy, and other spectral modalities. We demonstrate the advantages of metasurfaces in improving detection sensitivity, reducing detection limits, and achieving rapid biomolecule detection while discussing the challenges associated with the design, preparation, and stability of metasurfaces in biomedical detection procedures. Finally, we explore future development trends of metasurfaces for enhancing biological detection sensitivity and emphasize their wide-ranging applications.

Weight Differences-Based Multi-level Signal Profiling for Homogeneous and Ultrasensitive Intelligent Bioassays

Current high-sensitivity immunoassay protocols often involve complex signal generation designs or rely on sophisticated signal-loading and readout devices, making it challenging to strike a balance between sensitivity and ease of use. In this study, we propose a homogeneous-based intelligent analysis strategy called Mata, which uses weight analysis to quantify basic immune signals through signal subunits. We perform nanomagnetic labeling of target capture events on micrometer-scale polystyrene subunits, enabling magnetically regulated kinetic signal expression. Signal subunits are classified through the multi-level signal classifier in synergy with the developed signal weight analysis and deep learning recognition models. Subsequently, the basic immune signals are quantified to achieve ultra-high sensitivity. Mata achieves a detection of 0.61 pg/mL in 20 min for interleukin-6 detection, demonstrating sensitivity comparable to conventional digital immunoassays and over 22-fold that of chemiluminescence immunoassay and reducing detection time by more than 70%. The entire process relies on a homogeneous reaction and can be performed using standard bright-field optical imaging. This intelligent analysis strategy balances high sensitivity and convenient operation and has few hardware requirements, presenting a promising high-sensitivity analysis solution with wide accessibility.

Emerging trends in SERS-based veterinary drug detection: multifunctional substrates and intelligent data approaches

Veterinary drug residues in poultry and livestock products present persistent challenges to food safety, necessitating precise and efficient detection methods. Surface-enhanced Raman scattering (SERS) has been identified as a powerful tool for veterinary drug residue analysis due to its high sensitivity and specificity. However, the development of reliable SERS substrates and the interpretation of complex spectral data remain significant obstacles. This review summarizes the development process of SERS substrates, categorizing them into metal-based, rigid, and flexible substrates, and highlighting the emerging trend of multifunctional substrates. The diverse application scenarios and detection requirements for these substrates are also discussed, with a focus on their use in veterinary drug detection. Furthermore, the integration of deep learning techniques into SERS-based detection is explored, including substrate structure design optimization, optical property prediction, spectral preprocessing, and both qualitative and quantitative spectral analyses. Finally, key limitations are briefly outlined, such as challenges in selecting reporter molecules, data imbalance, and computational demands. Future trends and directions for improving SERS-based veterinary drug detection are proposed.

The potential of exosomes in regenerative medicine and in the diagnosis and therapies of neurodegenerative diseases and cancer

Exosomes, nanosized extracellular vesicles released by various cell types, are intensively studied for the diagnosis and treatment of cancer and neurodegenerative diseases, and they also display high usability in regenerative medicine. Emphasizing their diagnostic potential, exosomes serve as carriers of disease-specific biomarkers, enabling non-invasive early detection and personalized medicine. The cargo loading of exosomes with therapeutic agents presents an innovative strategy for targeted drug delivery, minimizing off-target effects and optimizing therapeutic interventions. In regenerative medicine, exosomes play a crucial role in intercellular communication, facilitating tissue regeneration through the transmission of bioactive molecules. While acknowledging existing challenges in standardization and scalability, ongoing research efforts aim to refine methodologies and address regulatory considerations. In summary, this review underscores the transformative potential of exosomes in reshaping the landscape of medical interventions, with a particular emphasis on cancer, neurodegenerative diseases, and regenerative medicine.

Label-Free Exosomal SERS Detection Assisted by Machine Learning for Accurately Discriminating Cell Cycle Stages and Revealing the Molecular Mechanisms during the Mitotic Process

Cell cycle analysis is crucial for disease diagnosis and treatment, especially for investigating cell heterogeneity and regulating cell behaviors. Exosomes are highly appealing as noninvasive biomarkers for monitoring real-time changes in the cell cycle due to their abundant molecular information inherited from their metrocyte cells and reflecting the state of these cells to some extent. However, to our knowledge, the relationship between exosomes and the cell cycle has not been reported. Herein, we successfully monitored the variation of exosomal surface-enhanced Raman spectroscopy (SERS) spectra to discriminate different cell cycle stages (G0/G1, S, and G2/M phases) based on label-free surface-enhanced Raman spectroscopy (SERS) combined with the machine learning method of linear discriminant analysis (LDA). An average accuracy of 85% based on the trained SERS spectra of exosomes from different cell cycle stages confirmed the high reliability of the support vector machine (SVM) algorithm for analyzing dynamic changes in the cell cycle at different time points. Importantly, the related molecular mechanisms among mitotic processes (prometaphase, metaphase, and anaphase/telophase) and unique biomolecular events between cancerous (HeLa) and normal (H8) cells were also revealed by the present label-free SERS detection platform. Based on SERS analysis, the content of phenylalanine (Phe) within HeLa cells increased, and some structures of proteins containing Phe and tryptophan (Trp) residues may be transformed during the mitotic process. Notably, the α-helix and β-sheet of proteins coexisted in HeLa cells; meanwhile, the α-helix of the proteins was more dominant in H8 cells than in HeLa cells. The strategy is effective for discriminating cell cycle stages and elucidating the associated molecular events during the cell mitotic process and will provide potential application value for guiding the cell cycle treatment strategies of cancer in the future.

Accurate Diagnosis of High-Risk Pulmonary Nodules Using a Non-Invasive Epigenetic Biomarker Test

BACKGROUND/OBJECTIVES

Accurate non-invasive tests to improve early detection and diagnosis of lung cancer are urgently needed. However, no regulatory-approved blood tests are available for this purpose. We aimed to improve pulmonary nodule classification to identify malignant nodules in a high-prevalence patient group.

METHODS

This study involved 806 participants with undiagnosed nodules larger than 5 mm, focusing on assessing nucleosome levels and histone modifications (H3.1 and H3K27Me3) in circulating blood. Nodules were classified as malignant or benign. For model development, the data were randomly divided into training (n = 483) and validation (n = 121) datasets. The model's performance was then evaluated using a separate testing dataset (n = 202).

RESULTS

Among the patients, 755 (93.7%) had a tissue diagnosis. The overall malignancy rate was 80.4%. For all datasets, the areas under curves were as follows: training, 0.74; validation, 0.86; and test, 0.79 (accuracy range: 0.80-0.88). Sensitivity showed consistent results across all datasets (0.91, 0.95, and 0.93, respectively), whereas specificity ranged from 0.37 to 0.64. For smaller nodules (5-10 mm), the model recorded accuracy values of 0.76, 0.88, and 0.85. The sensitivity values of 0.91, 1.00, and 0.94 further highlight the robust diagnostic capability of the model. The performance of the model across the reporting and data system (RADS) categories demonstrated consistent accuracy.

CONCLUSIONS

Our epigenetic biomarker panel detected non-small-cell lung cancer early in a high-risk patient group with high sensitivity and accuracy. The epigenetic biomarker model was particularly effective in identifying high-risk lung nodules, including small, part-solid, and non-solid nodules, and provided further evidence for validation.

Nanomedicine Innovations for Lung Cancer Diagnosis and Therapy

Lung cancer remains a challenge within the realm of oncology. Characterized by late-stage diagnosis and resistance to conventional treatments, the currently available therapeutic strategies encompass surgery, radiotherapy, chemotherapy, immunotherapy, and biological therapy; however, overall patient survival remains suboptimal. Nanotechnology has ushered in a new era by offering innovative nanomaterials with the potential to precisely target cancer cells while sparing healthy tissues. It holds the potential to reshape the landscape of cancer management, offering hope for patients and clinicians. The assessment of these nanotechnologies follows a rigorous evaluation process similar to that applied to chemical drugs, which includes considerations of their pharmacokinetics, pharmacodynamics, toxicology, and clinical effectiveness. However, because of the characteristics of nanoparticles, standard toxicological tests require modifications to accommodate their unique characteristics. Effective therapeutic strategies demand a profound understanding of the disease and consideration of clinical outcomes, physicochemical attributes of nanomaterials, nanobiointeractions, nanotoxicity, and regulatory compliance to ensure patient safety. This review explores the promise of nanomedicine in lung cancer treatment by capitalizing on its unique physicochemical properties. We address the multifaceted challenges of lung cancer and its tumor microenvironment and provide an overview of recent developments in nanoplatforms for early diagnosis and treatment that can enhance patient outcomes and overall quality of life.

Role of AI in empowering and redefining the oncology care landscape: perspective from a developing nation

Early diagnosis and accurate prognosis play a pivotal role in the clinical management of cancer and in preventing cancer-related mortalities. The burgeoning population of Asia in general and South Asian countries like India in particular pose significant challenges to the healthcare system. Regrettably, the demand for healthcare services in India far exceeds the available resources, resulting in overcrowded hospitals, prolonged wait times, and inadequate facilities. The scarcity of trained manpower in rural settings, lack of awareness and low penetrance of screening programs further compounded the problem. Artificial Intelligence (AI), driven by advancements in machine learning, deep learning, and natural language processing, can profoundly transform the underlying shortcomings in the healthcare industry, more for populous nations like India. With about 1.4 million cancer cases reported annually and 0.9 million deaths, India has a significant cancer burden that surpassed several nations. Further, India's diverse and large ethnic population is a data goldmine for healthcare research. Under these circumstances, AI-assisted technology, coupled with digital health solutions, could support effective oncology care and reduce the economic burden of GDP loss in terms of years of potential productive life lost (YPPLL) due to India's stupendous cancer burden. This review explores different aspects of cancer management, such as prevention, diagnosis, precision treatment, prognosis, and drug discovery, where AI has demonstrated promising clinical results. By harnessing the capabilities of AI in oncology research, healthcare professionals can enhance their ability to diagnose cancers at earlier stages, leading to more effective treatments and improved patient outcomes. With continued research and development, AI and digital health can play a transformative role in mitigating the challenges posed by the growing population and advancing the fight against cancer in India. Moreover, AI-driven technologies can assist in tailoring personalized treatment plans, optimizing therapeutic strategies, and supporting oncologists in making well-informed decisions. However, it is essential to ensure responsible implementation and address potential ethical and privacy concerns associated with using AI in healthcare.

ChatExosome: An Artificial Intelligence (AI) Agent Based on Deep Learning of Exosomes Spectroscopy for Hepatocellular Carcinoma (HCC) Diagnosis

Large language models (LLMs) hold significant promise in the field of medical diagnosis. There are still many challenges in the direct diagnosis of hepatocellular carcinoma (HCC). α-Fetoprotein (AFP) is a commonly used tumor marker for liver cancer. However, relying on AFP can result in missed diagnoses of HCC. We developed an artificial intelligence (AI) agent centered on LLMs, named ChatExosome, which created an interactive and convenient system for clinical spectroscopic analysis and diagnosis. ChatExosome consists of two main components: the first is the deep learning of the Raman fingerprinting of exosomes derived from HCC. Based on a patch-based 1D self-attention mechanism and downsampling, the feature fusion transformer (FFT) was designed to process the Raman spectra of exosomes. It achieved accuracies of 95.8% for cell-derived exosomes and 94.1% for 165 clinical samples, respectively. The second component is the interactive chat agent based on LLM. The retrieval-augmented generation (RAG) method was utilized to enhance the knowledge related to exosomes. Overall, LLM serves as the core of this interactive system, which is capable of identifying users' intentions and invoking the appropriate plugins to process the Raman data of exosomes. This is the first AI agent focusing on exosome spectroscopy and diagnosis, enhancing the interpretability of classification results, enabling physicians to leverage cutting-edge medical research and artificial intelligence techniques to optimize medical decision-making processes, and it shows great potential in intelligent diagnosis.

Reaction Pathway Differentiation Enabled Fingerprinting Signal for Single Nucleotide Variant Detection

Accurate identification of single-nucleotide variants (SNVs) is paramount for disease diagnosis. Despite the facile design of DNA hybridization probes, their limited specificity poses challenges in clinical applications. Here, a differential reaction pathway probe (DRPP) based on a dynamic DNA reaction network is presented. DRPP leverages differences in reaction intermediate concentrations between SNV and WT groups, directing them into distinct reaction pathways. This generates a strong pulse-like signal for SNV and a weak unidirectional increase signal for wild-type (WT). Through the application of machine learning to fluorescence kinetic data analysis, the classification of SNV and WT signals is automated with an accuracy of 99.6%, significantly exceeding the 80.7% accuracy of conventional methods. Additionally, sensitivity for variant allele frequency (VAF) is enhanced down to 0.1%, representing a ten-fold improvement over conventional approaches. DRPP accurately identified D614G and N501Y SNVs in the S gene of SARS-CoV-2 variants in patient swab samples with accuracy over 99% (n = 82). It determined the VAF of ovarian cancer-related mutations KRAS-G12R, NRAS-G12C, and BRAF-V600E in both tissue and blood samples (n = 77), discriminating cancer patients and healthy individuals with significant difference (p < 0.001). The potential integration of DRPP into clinical diagnostics, along with rapid amplification techniques, holds promise for early disease diagnostics and personalized diagnostics.

Deep learning-assisted surface-enhanced Raman spectroscopy detection of intracellular reactive oxygen species

Realizing the intelligent analysis of the intracellular reactive oxygen species (ROS) is beneficial to quick diagnosis of diseases. Herein, surface-enhanced Raman spectroscopy (SERS) technology was combined with deep learning to establish a smart detection method of intracellular ROS based on neural network to improve the SERS analysis ability. Taking the simultaneous detection of peroxynitrite (ONOO-) and hypochlorite (ClO-) as the templates, 4-mercaptophenylboric acid (4-MPBA) and 2-mercapto-4-methoxyphenol (2-MP) molecules were modified on the AuNPs to prepare AuNP/4-MPBA/2-MP nanoprobes. The SERS spectra of AuNP/4-MPBA/2-MP nanoprobes before and after the specific response of ONOO- and ClO- were collected to construct a database, and the neural network model for extraction (ENN) and one-dimensional convolutional neural network model (1D-CNN) for quantification were built. The cosine similarity values of ENN model for ONOO- and ClO- correlation spectra were 0.997 and 0.995, respectively. In addition, the qualitative and quantitative results of the models were basically consistent with the experimental results. Moreover, the models can accurately extract the SERS response spectral information of ONOO- and ClO- and realize their preliminary prediction of concentration in living cells, which has great potential in the high-throughput smart processing and accurate analysis of large-scale complicated SERS data from biological system.

Diagnosis of Lung Cancer Using Endobronchial Ultrasonography Image Based on Multi-Scale Image and Multi-Feature Fusion Framework

Lung cancer is the leading cause of cancer-related deaths globally and ranks among the most common cancer types. Given its low overall five-year survival rate, early diagnosis and timely treatment are essential to improving patient outcomes. In recent years, advances in computer technology have enabled artificial intelligence to make groundbreaking progress in imaging-based lung cancer diagnosis. The primary aim of this study is to develop a computer-aided diagnosis (CAD) system for lung cancer using endobronchial ultrasonography (EBUS) images and deep learning algorithms to facilitate early detection and improve patient survival rates. We propose M3-Net, which is a multi-branch framework that integrates multiple features through an attention-based mechanism, enhancing diagnostic performance by providing more comprehensive information for lung cancer assessment. The framework was validated on a dataset of 95 patient cases, including 13 benign and 82 malignant cases. The dataset comprises 1140 EBUS images, with 540 images used for training, and 300 images each for the validation and test sets. The evaluation yielded the following results: accuracy of 0.76, F1-score of 0.75, AUC of 0.83, PPV of 0.80, NPV of 0.75, sensitivity of 0.72, and specificity of 0.80. These findings indicate that the proposed attention-based multi-feature fusion framework holds significant potential in assisting with lung cancer diagnosis.

Nanomaterials and clinical SERS technology: broad applications in disease diagnosis

The critical need for rapid cancer diagnosis and related illnesses is growing alongside the current healthcare challenges, unfavorable prognosis, and constraints in diagnostic timing. As a result, emphasis on surface-enhanced Raman spectroscopy (SERS) diagnostic methods, including both label-free and labelled approaches, holds significant promise in fields such as analytical chemistry, biomedical science, and physics, due to the user-friendly nature of SERS. Over time, the SERS detection sensitivity and specificity with nanostructured materials for SERS applications (NMs-SERS) in different media have been remarkable. An investigation into electronic dynamics and interactions has revealed a seemingly fair result regarding the complementary effects of electromagnetic (EM) and chemical enhancements (CM), underscoring the operational principles of SERS. Nevertheless, the focus on translational SERS applications, especially beyond preliminary proof-of-concept research, remains limited. This review focuses on the advancements made in clinical SERS diagnostics and the essential role of NMs-SERS, ranging from plasmonic to non-plasmonic materials and other related advancements. Furthermore, it outlines the significant achievements of biomedical SERS in tumor diagnosis, particularly in identifying circulating tumor cells (CTCs), alongside a clear focus on NMs-SERS characteristics such as surface charge, shape, size, detection sensitivity, specificity, signal reproducibility, and recyclability. Finally, it underscores the use of microfluidic chips within the labelled SERS strategy for isolating CTCs, the concept of Ramanomics, and the integration of artificial intelligence (AI) to strengthen SERS data analysis. We hope that this review will help guide and expedite the potential for precise SERS diagnosis of key chronic diseases.

AI in SERS sensing moving from discriminative to generative

This perspective discusses the present and future role of artificial intelligence (AI) and machine learning (ML) in surface-enhanced Raman scattering (SERS) sensing. Our goal is to guide the reader through current applications, mainly focused on discriminative approaches aimed at developing new and improved SERS diagnostic capabilities, towards the future role of AI in SERS sensing, with the use of generative approaches to design new materials and biomaterials.

Identification of SERS Fingerprints for Diagnosis and Staging of Liver Injury

Acute liver injury is a common hepatic condition that, without timely diagnosis and treatment, can easily progress to life-threatening liver failure. Identifying key biomolecular changes in the pathogenesis of liver injury is critical for the precise diagnosis of hepatic injury. In this study, we identified novel molecular fingerprints associated with hepatic injury using artificial intelligence-assisted Surface-enhanced Raman spectroscopy (SERS) technology. By intravenous injection of PEG-modified Au nanoparticles, which efficiently accumulate in the liver, these nanoparticles act as in vivo computed tomography contrast agents, specifically delineating the three-dimensional structure of normal and injured liver tissues. Additionally, as a Raman-enhanced substrate, gold nanoparticles allow for the analysis of SERS signals in liver tissues enriched with these nanoparticles. Leveraging artificial intelligence technology, we achieved a classification accuracy of 96.45% for different degrees of liver injury. Molecular spectral analysis revealed that the ratio of the 1437-1000 cm-1 signal could serve as an SERS fingerprint correlated with hepatic injury. The integrated information system provides a valuable tool for the precise diagnosis of liver diseases.

Surface-Enhanced Raman Scattering (SERS) for exosome detection

BACKGROUND

Exosomes, nanoscale extracellular vesicles secreted by various cells, are abundantly present in biological fluids. They have been identified as carriers of specific molecules, suggesting their potential role in early disease detection. However, their clinical application is hindered by several challenges, including the need for large sample volumes for enrichment, limitations of traditional detection methods, and the complexity involved in phenotype analysis and separation.

OBJECTIVE

This review aims to explore the application of Surface-Enhanced Raman Scattering (SERS) technology in exosome detection. SERS, known for its unique photonic properties and high sensitivity, offers a promising solution for detecting exosomes without the need for large sample volumes or extensive phenotypic analysis. This review focuses on the real-time and non-invasive assessment capabilities of SERS in exosome detection, providing insights into its potential for early disease diagnosis.

CONCLUSION

The review concludes by emphasizing the potential of SERS-based exosome detection in advancing early disease diagnosis. By overcoming existing challenges, SERS technology offers a promising approach for the development of sensitive and specific diagnostic assays, contributing to better patient outcomes and personalized medicine.

Raman spectroscopy in extracellular vesicles analysis: Techniques, applications and advancements

Raman spectroscopy provides a robust approach for detailed analysis of the chemical and molecular profiles of extracellular vesicles (EVs). Recent advancements in Raman techniques have significantly enhanced the sensitivity and accuracy of EV characterization, enabling precise detection and profiling of molecular components within EV samples. This review introduces and compares various Raman-based techniques for EV characterization. These include Raman spectroscopy (RS), which provides fundamental molecular information; Raman trapping analysis (RTA), which combines optical trapping with Raman scattering for the manipulation and analysis of individual EVs; surface-enhanced Raman spectroscopy (SERS), which enhances the Raman signal through the use of metallic nanostructures, significantly improving sensitivity; and microfluidic SERS, which integrates SERS with microfluidic platforms to allow high-throughput, label-free analysis of EVs in biological fluids. In addition to comparing various Raman techniques, this review provides a comprehensive analysis that includes comparisons of machine learning methods, EV isolation techniques, and characterization strategies. By integrating these approaches, the review presents a holistic perspective on Raman-based EV analysis, covering profiling, purity, heterogeneity and size analysis as well as imaging. The combined assessment of Raman technologies with advanced computational and experimental methodologies supports the development of more robust diagnostic and therapeutic applications involving EVs.

Advanced Design for Stimuli-Reversible Chromic Wearables With Customizable Functionalities

Smart wearable devices with dynamically reversible color displays are crucial for the next generation of smart textiles, and promising for bio-robots, adaptive camouflage, and visual health monitoring. The rapid advancement of technology brings out different categories that feature fundamentally different color-reversing mechanisms, including thermochromic, mechanochromic, electrochromic, and photochromic smart wearables. Although some reviews have showcased relevant developments from unique perspectives, reviews focusing on the advanced design of flexible chromic wearable devices within each category have not been reported. In this review, the development history and recent progress in smart chromic wearables across each category are systematically examined. The design strategies for each chromic wearable device are outlined with a focus on functional materials, synthesis processes, and advanced applications. Furthermore, integrated devices based on dual-stimuli and multi-stimuli responsive chromics with customizable functionalities are summarized. Finally, challenges and perspectives on the future development of smart chromic wearables are proposed. Such a systematic summary will serve as a valuable insight for researchers in this field.

Label-Free SERS Analysis of Biological and Physical Information Heterogeneity of Nanoscale Extracellular Vesicle by Matching Specific Sizes of Enhanced Particles

The heterogeneity of extracellular vesicles (EVs) surface information represents different functions, which is neglected in previous studies. In this study, a label-free SERS analysis approach is demonstrated to study fundamental EV biological and physical information heterogeneity by matching specific sizes of nano-enhanced particles. This strategy reveals informative, comprehensive, and high-quality SERS spectra of the overall exosome surface, and effectively circumvents the key information loss caused by the spatial resistance of NPs binding to the 293 exosomes' concave structure. The classification of normal and cancerous cell-derived exosomes by PCA method, the accuracy is improved from 91.2% to 95.1% by optimizing sizes of nano-enhanced particles. In addition, stem cell-derived EVs of diverse sizes and morphologies similarly show acuity of spectrum variation to NPs size, which is conductive to qualitative studies. This new strategy will offer a widened in-depth understanding of the surface information, size, and morphology of EVs, which can be applied to the study of biological functions.

Emerging Wearable Acoustic Sensing Technologies

Sound signals not only serve as the primary communication medium but also find application in fields such as medical diagnosis and fault detection. With public healthcare resources increasingly under pressure, and challenges faced by disabled individuals on a daily basis, solutions that facilitate low-cost private healthcare hold considerable promise. Acoustic methods have been widely studied because of their lower technical complexity compared to other medical solutions, as well as the high safety threshold of the human body to acoustic energy. Furthermore, with the recent development of artificial intelligence technology applied to speech recognition, speech recognition devices, and systems capable of assisting disabled individuals in interacting with scenes are constantly being updated. This review meticulously summarizes the sensing mechanisms, materials, structural design, and multidisciplinary applications of wearable acoustic devices applied to human health and human-computer interaction. Further, the advantages and disadvantages of the different approaches used in flexible acoustic devices in various fields are examined. Finally, the current challenges and a roadmap for future research are analyzed based on existing research progress to achieve more comprehensive and personalized healthcare.

Metallomic Classification of Pulmonary Nodules Using Blood by Deep-Learning-Boosted Synchrotron Radiation X-ray Fluorescence

Ambient air pollution is an important contributor to increasing cases of lung cancer, which is a malignant cancer with the highest mortality among all cancers. It primarily manifests in the form of pulmonary nodules, but not all will develop into lung cancer. Therefore, it is highly desired to distinguish between benign and malignant pulmonary nodules for the early prevention and treatment of lung cancer. Currently, histopathological examination is the gold standard for classifying pulmonary nodules, which is invasive, time-consuming, and labor-intensive. This study proposes a metallomics approach through synchrotron radiation X-ray fluorescence (SRXRF) with a simplified one-dimensional convolutional neural network (1DCNN) to distinguish pulmonary nodules by using serum samples. SRXRF spectra of serum samples were obtained and preliminarily analyzed using principal component analysis (PCA). Subsequently, machine learning algorithms (MLs) and 1DCNN were applied to develop classification models. Both MLs and 1DCNN based on full-channel spectra could distinguish patients with benign and malignant pulmonary nodules, but the highest accuracy rate of 96.7% was achieved when using 1DCNN. In addition, it was found that characteristic elements in serum from patients with malignant nodules were different from those in benign nodules, which can serve as the fingerprint metallome profile. The simplified model based on characteristic elements resulted in good performance of sensitivity and F1-score > 91.30%, G-mean, MCC and Kappa > 85.59%, and accuracy = 94.34%. In summary, metallomic classification of benign and malignant pulmonary nodules using serum samples can be achieved through 1DCNN-boosted SRXRF, which is easy to handle and much less invasive compared to histopathological examination.

Raman Spectral Feature Enhancement Framework for Complex Multiclassification Tasks

Raman spectroscopy enables label-free clinical diagnosis in a single step. However, identifying an individual carrying a specific disease from people with a multi-disease background is challenging. To address this, we developed a Raman spectral implicit feature augmentation with a Raman Intersection, Union, and Subtraction augmentation strategy (RIUS). RIUS expands the data set without requiring additional labeled data by leveraging set operations at the feature level, significantly enhancing model performance across various applications. On a challenging 30-class bacterial classification task, RIUS demonstrated a substantial improvement, increasing the accuracy of ResNet by 2.1% and that of SE-ResNet by 1.4%, achieving accuracies of 85.7% and 87.1%, respectively, on the Bacteria-ID-4 Data set, where RIUS improved ResNet and SE-ResNet accuracies by 13.6% and 14.5%, respectively, with only ten samples per category. When the sample size was reduced, accuracy gains increased to 31.7% and 38.3%, demonstrating the method's robustness across different sample volumes. Compared to basic augmentation, our method exhibited superior performance across various sample volumes and demonstrated exceptional adaptability to different levels of complexity. RIUS exhibited superior performance, particularly in complex settings. Moreover, cluster analysis validated the effectiveness of the implicit feature augmentation module and the consistency between theoretical design and experimental results. We further validated our approach using clinical serum samples from 70 breast cancer patients and 70 controls, achieving an AUC of 0.94 and a sensitivity of 92.9%. Our approach enhances the potential for precisely identifying diseases in complex settings and offers plug-and-play enhancement for existing classification models.

Circulating Extracellular Vesicles as Promising Biomarkers for Precession Diagnostics: A Perspective on Lung Cancer

Extracellular vesicles (EVs) have emerged as promising biomarkers in liquid biopsy, owing to their ubiquitous presence in bodily fluids and their ability to carry disease-related cargo. Recognizing their significance in disease diagnosis and treatment, substantial efforts have been dedicated to developing efficient methods for EV isolation, detection, and analysis. EVs, heterogeneous membrane-encapsulated vesicles secreted by all cells, contain bioactive substances capable of modulating recipient cell biology upon internalization, including proteins, lipids, DNA, and various RNAs. Their prevalence across bodily fluids has positioned them as pivotal mediators in physiological and pathological processes, notably in cancer, where they hold potential as straightforward tumor biomarkers. This review offers a comprehensive examination of advanced nanotechnology-based techniques for detecting lung cancer through EV analysis. It begins by providing a brief overview of exosomes and their role in lung cancer progression. Furthermore, this review explores the evolving landscape of EV isolation and cargo analysis, highlighting the importance of characterizing specific biomolecular signatures within EVs for improved diagnostic accuracy in lung cancer patients. Innovative strategies for enhancing the sensitivity and specificity of EV isolation and detection, including the integration of microfluidic platforms and multiplexed biosensing technologies are summarized. The discussion then extends to key challenges associated with EV-based liquid biopsies, such as the standardization of isolation and detection protocols and the establishment of robust analytical platforms for clinical translation. This review highlights the transformative impact of EV-based liquid biopsy in lung cancer diagnosis, heralding a new era of personalized medicine and improved patient care.

Advances in exosome plasmonic sensing: Device integration strategies and AI-aided diagnosis

Exosomes, as next-generation biomarkers, has great potential in tracking cancer progression. They face many detection limitations in cancer diagnosis. Plasmonic biosensors have attracted considerable attention at the forefront of exosome detection, due to their label-free, real-time, and high-sensitivity features. Their advantages in multiplex immunoassays of minimal liquid samples establish the leading position in various diagnostic studies. This review delineates the application principles of plasmonic sensing technologies, highlighting the importance of exosomes-based spectrum and image signals in disease diagnostics. It also introduces advancements in miniaturizing plasmonic biosensing platforms of exosomes, which can facilitate point-of-care testing for future healthcare. Nowadays, inspired by the surge of artificial intelligence (AI) for science and technology, more and more AI algorithms are being adopted to process the exosome spectrum and image data from plasmonic detection. Using representative algorithms of machine learning has become a mainstream trend in plasmonic biosensing research for exosome liquid biopsy. Typically, these algorithms process complex exosome datasets efficiently and establish powerful predictive models for precise diagnosis. This review further discusses critical strategies of AI algorithm selection in exosome-based diagnosis. Particularly, we categorize the AI algorithms into the interpretable and uninterpretable groups for exosome plasmonic detection applications. The interpretable AI enhances the transparency and reliability of diagnosis by elucidating the decision-making process, while the uninterpretable AI provides high diagnostic accuracy with robust data processing by a "black-box" working mode. We believe that AI will continue to promote significant progress of exosome plasmonic detection and mobile healthcare in the near future.

Rapid detection of perfluorooctanoic acid by surface enhanced Raman spectroscopy and deep learning

Perfluorooctanoic acid (PFOA) has received increasing concerns in recent years due to its wide distribution and potential toxicity. Existing detection techniques of PFOA require complex pre-treatment, therefore often taking several hours. Here, we developed a rapid PFOA detection mode to detect approximate concentrations of PFOA (ranging from 10-15 to 10-3 mol/L) in deionized water, and detecting one sample takes only 20 min. The detection mode was achieved using a deep learning model trained by a large surface enhanced Raman spectra dataset, based on the agglomeration of PFOA with crystal violet. In addition, transfer learning approach was used to fine tune the model, the fine-tuned model was generalizable across water samples with different impurities and environments to determine whether meet the safety standards of PFOA, the accuracy was 96.25 % and 94.67 % for tap water and lake water samples, respectively. The mechanism and specificity of the detection mode were further confirmed by molecular dynamics simulation. Our work provides a promising solution for PFOA detection, especially in the context of the increasingly widespread application of PFOA.

Engineering Strategies of Plant-Derived Exosome-Like Nanovesicles: Current Knowledge and Future Perspectives

Plant-derived exosome-like nanovesicles (PELNs) from edible plants, isolated by ultracentrifugation, size exclusion chromatography or other methods, were proved to contain a variety of biologically active and therapeutically specific components. Recently, investigations in the field of PELN-based biomedicine have been conducted, which positioned those nanovesicles as promising tools for prevention and treatment of several diseases, with their natural origin potentially offering superior biocompatibility and bioavailability. However, the inadequate targeting and limited therapeutic effects constrain the utility and clinical translation of PELNs. Thus, strategies aiming at bridging the gap by engineering natural PELNs have been of great interest. Those approaches include membrane hybridization, physical and chemical surface functionalization and encapsulation of therapeutic payloads. Herein, we provide a comprehensive overview of the biogenesis and composition, isolation and purification methods and characterization of PELNs, as well as their therapeutic functions. Current knowledge on the construction strategies and biomedical application of engineered PELNs were reviewed. Additionally, future directions and perspectives in this field were discussed in order to further enrich and expand the prospects for the application of engineered PELNs.

Integrated SERS-Microfluidic Sensor Based on Nano-Micro Hierarchical Cactus-like Array Substrates for the Early Diagnosis of Prostate Cancer

The detection and analysis of cancer cell exosomes with high sensitivity and precision are pivotal for the early diagnosis and treatment strategies of prostate cancer. To this end, a microfluidic chip, equipped with a cactus-like array substrate (CAS) based on surface-enhanced Raman spectroscopy (SERS) was designed and fabricated for the detection of exosome concentrations in Lymph Node Carcinoma of the Prostate (LNCaP). Double layers of polystyrene (PS) microspheres were self-assembled onto a polyethylene terephthalate (PET) film to form an ordered cactus-like nanoarray for detection and analysis. By combining EpCAM aptamer-labeled SERS nanoprobes and a CD63 aptamer-labeled CAS, a 'sandwich' structure was formed and applied to the microfluidic chips, further enhancing the Raman scattering signal of Raman reporter molecules. The results indicate that the integrated microfluidic sensor exhibits a good linear response within the detection concentration range of 105 particles μL-1 to 1 particle μL-1. The detection limit of exosomes in cancer cells can reach 1 particle μL-1. Therefore, we believed that the CAS integrated microfluidic sensor offers a superior solution for the early diagnosis and therapeutic intervention of prostate cancer.

Liquid Biopsy-Based Detection and Response Prediction for Depression

Proactively predicting antidepressant treatment response before medication failures is crucial, as it reduces unsuccessful attempts and facilitates the development of personalized therapeutic strategies, ultimately enhancing treatment efficacy. The current decision-making process, which heavily depends on subjective indicators, underscores the need for an objective, indicator-based approach. This study developed a method for detecting depression and predicting treatment response through deep learning-based spectroscopic analysis of extracellular vesicles (EVs) from plasma. EVs were isolated from the plasma of both nondepressed and depressed groups, followed by Raman signal acquisition, which was used for AI algorithm development. The algorithm successfully distinguished depression patients from healthy individuals and those with panic disorder, achieving an AUC accuracy of 0.95. This demonstrates the model's capability to selectively diagnose depression within a nondepressed group, including those with other mental health disorders. Furthermore, the algorithm identified depression-diagnosed patients likely to respond to antidepressants, classifying responders and nonresponders with an AUC accuracy of 0.91. To establish a diagnostic foundation, the algorithm applied explainable AI (XAI), enabling personalized medicine for companion diagnostics and highlighting its potential for the development of liquid biopsy-based mental disorder diagnosis.

Electroactivated SERS Nanoplatform for Rapid and Sensitive Detection and Identification of Tumor-Derived Exosome miRNA

Exosomal miRNA expression derived from tumor cells provides a valuable and promising noninvasive modality for the early diagnosis and assessment of the efficacy of cancer treatment. However, accurate detection and identification of miRNA within exosomes have been challenging due to its low abundance and the complexity and tedious extraction with large sample volumes in the separation process. Here, we developed an electrically activated nanoplatform for rapid and sensitive detection and identification of exosome miRNA, through triggering miRNA release by opening exosomes that were captured on the electrode surface using a slightly applied electric field (50 mV), and simultaneously detected them with surface-enhanced Raman spectroscopy (SERS) in situ. The method possessed superior specificity and sensitivity for exosomal miRNA detection, with a low detection concentration of 0.5 nM. The SERS sensor chips also showed a superior sensing performance of exosomal miRNA in complex body fluids such as urine and blood. We found that exosomal miRNA contents derived from tumor cells were significantly higher than those in normal cells, and importantly, the concentrations of exosomes secreted from three different cell lines were distinctly augmented after mild electrical stimulation (ES) treatment. Furthermore, the miRNA expression within exosomes was upregulated after the ES treatment of cells. The developed approach and SERS detection platform for exosomal miRNA are promising for noninvasive and precise screening, classification, and monitoring of cancer.

Machine learning-assisted label-free colorectal cancer diagnosis using plasmonic needle-endoscopy system

Early and accurate detection of colorectal cancer (CRC) is critical for improving patient outcomes. Existing diagnostic techniques are often invasive and carry risks of complications. Herein, we introduce a plasmonic gold nanopolyhedron (AuNH)-coated needle-based surface-enhanced Raman scattering (SERS) sensor, integrated with endoscopy, for direct mucus sampling and label-free detection of CRC. The thin and flexible stainless-steel needle is coated with polymerized dopamine, which serves as an adhesive layer and simultaneously initiates the nucleation of gold nanoparticle (AuNP) seeds on the needle surface. The AuNP seeds are further grown through a surface-directed reduction using Au ions-hydroxylamine hydrochloride solution, resulting in the formation of dense AuNHs. The formation mechanism of AuNHs and the layered structure of the plasmonic needle-based SERS (PNS) sensor are thoroughly analyzed. Furthermore, a strong field enhancement of the PNS sensor is observed, amplified around the edges of the polyhedral shapes and at nanogap sites between AuNHs. The feasibility of the PNS sensor combined with endoscopy system is further investigated using mouse models for direct colonic mucus sampling and verifying noninvasive label-free classification of CRC from normal controls. A logistic regression-based machine learning method is employed and successfully differentiates CRC and normal mice, achieving 100% sensitivity, 93.33% specificity, and 96.67% accuracy. Moreover, Raman profiling of metabolites and their correlations with Raman signals of mucus samples are analyzed using the Pearson correlation coefficient, offering insights for identifying potential cancer biomarkers. The developed PNS-assisted endoscopy technology is expected to advance the early screening and diagnosis approach of CRC in the future.

Identification of Molecular Profile of Cell Membrane via Magnetic Plasmonic Nanoprobe

Cell membranes are primarily composed of lipids, membrane proteins, and carbohydrates, and the related studies of membrane components and structures at different stages of disease development, especially membrane proteins, are of great significance. Here, we investigate the chemical signature profiles of cell membranes as biomarkers for cancer cells via label-free surface-enhanced Raman scattering (SERS). A magnetic plasmonic nanoprobe was proposed by decorating magnetic beads with silver nanoparticles, allowing self-driven cell membrane-targeted accumulation within 5 min. SERS profiles of three types of breast cells were achieved under the plasmon enhancement effect of these nanoprobes. Membrane fingerprint spectra from breast cell lines were further classified with the convolutional neural network model, which perfectly distinguished between two breast cancer subtypes. We further tested various clinical samples using this method and obtained fingerprint spectra from primary cells and frozen slices. This study presents a practical, user-friendly approach for label-free and in situ analysis of cell membranes, which can work for early tumor screening and treatment assessment for tumors reliant on the Molecular profiles of cell membranes. Additionally, this method can be applied universally to explore cell membrane components of other cells, thus assisting Human Cell Atlas.

Recent Advancements in Nanophotonics for Optofluidics

Optofluidics is dedicated to achieving integrated control of particle and fluid motion, particularly on the micrometer scale, by utilizing light to direct fluid flow and particle motion. The field has seen significant growth recently, driven by the concerted efforts of researchers across various scientific disciplines, notably for its successful applications in biomedical science. In this review, we explore a range of optofluidic architectures developed over the past decade, with a primary focus on mechanisms for precise control of micro and nanoscale biological objects and their applications in sensing. Regarding nanoparticle manipulation, we delve into mechanisms based on optical nanotweezers using nanolocalized light fields and light-based hybrid effects with dramatically improved performance and capabilities. In the context of sensing, we emphasize those works that used optofluidics to aggregate molecules or particles to promote sensing and detection. Additionally, we highlight emerging research directions, encompassing both fundamental principles and practical applications in the field.

Interpretable Machine Learning-Aided Optical Deciphering of Serum Exosomes for Early Detection, Staging, and Subtyping of Lung Cancer

Lung cancer (LC) is the leading cause of cancer-related mortality worldwide, underscoring an urgent need for strategies that enable early detection and phenotypic classification. Here, we conducted a label-free surface-enhanced Raman spectroscopic (SERS) analysis of serum exosomes from 643 participants to elucidate the biochemical deregulation associated with LC progression and the unique phenotypes of different LC subtypes. Iodide-modified silver nanofilms were prepared to rapidly acquire SERS spectra with a high signal-to-noise ratio using 0.5 μL of patient exosomes. We performed interpretable and automated machine learning (ML) analysis of differential SERS features of serum exosomes to build LC diagnostic models, which achieved accuracies of 100% and 81% for stage I lung adenocarcinoma and its preneoplasia, respectively. In addition, the ML-derived exosomal SERS models effectively recognized different LC subtypes and disease stages to guide precision treatment. Our findings demonstrate that spectral fingerprinting of circulating exosomes holds promise for decoding the clinical status of LC, thus aiding in improving the clinical management of patients.

A cavity induced mode hybridization plasmonic sensor for portable detection of exosomes

Exosomes have been considered as promising biomarkers for cancer diagnosis due to their abundant information from originating cells. However, sensitive and reliable detection of exosomes is still facing technically challenges due to the lack of a sensing platform with high sensitivity and reproducibility. To address the challenges, here we propose a portable surface plasmon resonance (SPR) sensing of exosomes with a three-layer Au mirror/SiO2 spacer/Au nanohole sensor, fabricated by an economical polystyrene nanosphere self-assembly method. The SiO2 spacer can act as an optical cavity and induce mode hybridization, leading to excellent optimization of both sensitivity and full width at half maximum compared with normal single layer Au nanohole sensors. When modified with CD63 or EpCAM aptamers, a detection of limit (LOD) of as low as 600 particles/μL was achieved. The sensors showed good capability to distinguish between non-tumor derived L02 exosomes and tumor derived HepG2 exosomes. Additionally, high reproducibility was also achieved in detection of artificial serum samples with RSD as low as 2%, making it feasible for clinical applications. This mode hybridization plasmonic sensor provides an effective approach to optimize the detection sensitivity of exosomes, pushing SPR sensing one step further towards cancer diagnosis.

SERS-Based Droplet Microfluidic Platform for Sensitive and High-Throughput Detection of Cancer Exosomes

Exosomes, nanosized extracellular vesicles containing biomolecular cargo, are increasingly recognized as promising noninvasive biomarkers for cancer diagnosis, particularly for their role in carrying tumor-specific molecular information. Traditional methods for exosome detection face challenges such as complexity, time consumption, and the need for sophisticated equipment. This study addresses these challenges by introducing a novel droplet microfluidic platform integrated with a surface-enhanced Raman spectroscopy (SERS)-based aptasensor for the rapid and sensitive detection of HER2-positive exosomes from breast cancer cells. Our approach utilized an on-chip salt-induced gold nanoparticles (GNPs) aggregation process in the presence of HER2 aptamers and HER2-positive exosomes, enhancing the hot spot-based SERS signal amplification. This platform achieved a limit of detection of 4.5 log10 particles/mL with a sample-to-result time of 5 min per sample. Moreover, this platform has been successfully applied for HER2 status testing in clinical samples to distinguish HER2-positive breast cancer patients from HER2-negative breast cancer patients. High sensitivity, specificity, and the potential for high-throughput screening of specific tumor exosomes make this SERS-based droplet system a potential liquid biopsy technology for early cancer diagnosis.

Role of Machine Learning Assisted Biosensors in Point-of-Care-Testing For Clinical Decisions

Point-of-Care-Testing (PoCT) has emerged as an essential component of modern healthcare, providing rapid, low-cost, and simple diagnostic options. The integration of Machine Learning (ML) into biosensors has ushered in a new era of innovation in the field of PoCT. This article investigates the numerous uses and transformational possibilities of ML in improving biosensors for PoCT. ML algorithms, which are capable of processing and interpreting complicated biological data, have transformed the accuracy, sensitivity, and speed of diagnostic procedures in a variety of healthcare contexts. This review explores the multifaceted applications of ML models, including classification and regression, displaying how they contribute to improving the diagnostic capabilities of biosensors. The roles of ML-assisted electrochemical sensors, lab-on-a-chip sensors, electrochemiluminescence/chemiluminescence sensors, colorimetric sensors, and wearable sensors in diagnosis are explained in detail. Given the increasingly important role of ML in biosensors for PoCT, this study serves as a valuable reference for researchers, clinicians, and policymakers interested in understanding the emerging landscape of ML in point-of-care diagnostics.

Integrating, Validating, and Expanding Information Space in Single-Molecule Surface-Enhanced Raman Spectroscopy for Biomolecules

Single-molecule surface-enhanced Raman spectroscopy (SM-SERS) is an ultrahigh-resolution spectroscopic method for directly obtaining the complex vibrational mode information on individual molecules. SM-SERS offers a wide range of submolecular information on the hidden heterogeneity in its functional groups and varying structures, dynamics of conformational changes, binding and reaction kinetics, and interactions with the neighboring molecule and environment. Despite the richness in information on individual molecules and potential of SM-SERS in various detection targets, including large and complex biomolecules, several issues and practical considerations remain to be addressed, such as the requirement of long integration time, challenges in forming reliable and controllable interfaces between nanostructures and biomolecules, difficulty in determining hotspot size and shape, and most importantly, insufficient signal reproducibility and stability. Moreover, utilizing and interpreting SERS spectra is challenging, mainly because of the complexity and dynamic nature of molecular fingerprint Raman spectra, and this leads to fragmentary analysis and incomplete understanding of the spectra. In this Perspective, we discuss the current challenges and future opportunities of SM-SERS in views of system approaches by integrating molecules of interest, Raman dyes, plasmonic nanostructures, and artificial intelligence, particularly for detecting and analyzing biomolecules to realize the validation and expansion of information space in SM-SERS.

Emerging research trends in artificial intelligence for cancer diagnostic systems: A comprehensive review

This review article offers a comprehensive analysis of current developments in the application of machine learning for cancer diagnostic systems. The effectiveness of machine learning approaches has become evident in improving the accuracy and speed of cancer detection, addressing the complexities of large and intricate medical datasets. This review aims to evaluate modern machine learning techniques employed in cancer diagnostics, covering various algorithms, including supervised and unsupervised learning, as well as deep learning and federated learning methodologies. Data acquisition and preprocessing methods for different types of data, such as imaging, genomics, and clinical records, are discussed. The paper also examines feature extraction and selection techniques specific to cancer diagnosis. Model training, evaluation metrics, and performance comparison methods are explored. Additionally, the review provides insights into the applications of machine learning in various cancer types and discusses challenges related to dataset limitations, model interpretability, multi-omics integration, and ethical considerations. The emerging field of explainable artificial intelligence (XAI) in cancer diagnosis is highlighted, emphasizing specific XAI techniques proposed to improve cancer diagnostics. These techniques include interactive visualization of model decisions and feature importance analysis tailored for enhanced clinical interpretation, aiming to enhance both diagnostic accuracy and transparency in medical decision-making. The paper concludes by outlining future directions, including personalized medicine, federated learning, deep learning advancements, and ethical considerations. This review aims to guide researchers, clinicians, and policymakers in the development of efficient and interpretable machine learning-based cancer diagnostic systems.

Artificial intelligence-based plasma exosome label-free SERS profiling strategy for early lung cancer detection

As a lung cancer biomarker, exosomes were utilized for in vitro diagnosis to overcome the lack of sensitivity of conventional imaging and the potential harm caused by tissue biopsy. However, given the inherent heterogeneity of exosomes, the challenge of accurately and reliably recognizing subtle differences in the composition of exosomes from clinical samples remains significant. Herein, we report an artificial intelligence-assisted surface-enhanced Raman spectroscopy (SERS) strategy for label-free profiling of plasma exosomes for accurate diagnosis of early-stage lung cancer. Specifically, we build a deep learning model using exosome spectral data from lung cancer cell lines and normal cell lines. Then, we extracted the features of cellular exosomes by training a convolutional neural network (CNN) model on the spectral data of cellular exosomes and used them as inputs to a support vector machine (SVM) model. Eventually, the spectral features of plasma exosomes were combined to effectively distinguish adenocarcinoma in situ (AIS) from healthy controls (HC). Notably, the approach demonstrated significant performance in distinguishing AIS from HC samples, with an area under the curve (AUC) of 0.84, sensitivity of 83.3%, and specificity of 83.3%. Together, the results demonstrate the utility of exosomes as a biomarker for the early diagnosis of lung cancer and provide a new approach to prescreening techniques for lung cancer.

Recent advances to address challenges in extracellular vesicle-based applications for lung cancer

Lung cancer, highly prevalent and the leading cause of cancer-related death globally, persists as a significant challenge due to the lack of definitive tumor markers for early diagnosis and personalized therapeutic interventions. Recently, extracellular vesicles (EVs), functioning as natural carriers for intercellular communication, have received increasing attention due to their ability to traverse biological barriers and deliver diverse biological cargoes, including cytosolic proteins, cell surface proteins, microRNA, lncRNA, circRNA, DNA, and lipids. EVs are increasingly recognized as a valuable resource for non-invasive liquid biopsy, as well as drug delivery platforms, and anticancer vaccines for precision medicine in lung cancer. Herein, given the diagnostic and therapeutic potential of tumor-associated EVs for lung cancer, we discuss this topic from a translational standpoint. We delve into the specific roles that EVs play in lung cancer carcinogenesis and offer a particular perspective on how advanced engineering technologies can overcome the current challenges and expedite and/or enhance the translation of EVs from laboratory research to clinical settings.