Use of Machine Learning with Temporal Photoluminescence Signals from CdTe Quantum Dots for Temperature Measurement in Microfluidic Devices

Luminescence Thermometry Beyond the Biological Realm

As the field of luminescence thermometry has matured, practical applications of luminescence thermometry techniques have grown in both frequency and scope. Due to the biocompatibility of most luminescent thermometers, many of these applications fall within the realm of biology. However, luminescence thermometry is increasingly employed beyond the biological realm, with expanding applications in areas such as thermal characterization of microelectronics, catalysis, and plasmonics. Here, we review the motivations, methodologies, and advances linked to nonbiological applications of luminescence thermometry. We begin with a brief overview of luminescence thermometry probes and techniques, focusing on those most commonly used for nonbiological applications. We then address measurement capabilities that are particularly relevant for these applications and provide a detailed survey of results across various application categories. Throughout the review, we highlight measurement challenges and requirements that are distinct from those of biological applications. Finally, we discuss emerging areas and future directions that present opportunities for continued research.

Microfluidic Platforms for Real-Time In Situ Monitoring of Biomarkers for Cellular Processes

Cellular processes are mechanisms carried out at the cellular level that are aimed at guaranteeing the stability of the organism they comprise. The investigation of cellular processes is key to understanding cell fate, understanding pathogenic mechanisms, and developing new therapeutic technologies. Microfluidic platforms are thought to be the most powerful tools among all methodologies for investigating cellular processes because they can integrate almost all types of the existing intracellular and extracellular biomarker-sensing methods and observation approaches for cell behavior, combined with precisely controlled cell culture, manipulation, stimulation, and analysis. Most importantly, microfluidic platforms can realize real-time in situ detection of secreted proteins, exosomes, and other biomarkers produced during cell physiological processes, thereby providing the possibility to draw the whole picture for a cellular process. Owing to their advantages of high throughput, low sample consumption, and precise cell control, microfluidic platforms with real-time in situ monitoring characteristics are widely being used in cell analysis, disease diagnosis, pharmaceutical research, and biological production. This review focuses on the basic concepts, recent progress, and application prospects of microfluidic platforms for real-time in situ monitoring of biomarkers in cellular processes.

Neural Networks Push the Limits of Luminescence Lifetime Nanosensing

Luminescence lifetime-based sensing is ideally suited to monitor biological systems due to its minimal invasiveness and remote working principle. Yet, its applicability is limited in conditions of low signal-to-noise ratio (SNR) induced by, e.g., short exposure times and presence of opaque tissues. Herein this limitation is overcome by applying a U-shaped convolutional neural network (U-NET) to improve luminescence lifetime estimation under conditions of extremely low SNR. Specifically, the prowess of the U-NET is showcased in the context of luminescence lifetime thermometry, achieving more precise thermal readouts using Ag2 S nanothermometers. Compared to traditional analysis methods of decay curve fitting and integration, the U-NET can extract average lifetimes more precisely and consistently regardless of the SNR value. The improvement achieved in the sensing performance using the U-NET is demonstrated with two experiments characterized by extreme measurement conditions: thermal monitoring of free-falling droplets, and monitoring of thermal transients in suspended droplets through an opaque medium. These results broaden the applicability of luminescence lifetime-based sensing in fields including in vivo experimentation and microfluidics, while, hopefully, spurring further research on the implementation of machine learning (ML) in luminescence sensing.

Using Recurrent Neural Networks to Reconstruct Temperatures from Simulated Fluorescent Data for use in Bio-Microfluidics

Many biological systems have a narrow temperature range of operation, meaning high accuracy and spatial distribution level are needed to study these systems. Most temperature sensors cannot meet both the accuracy and spatial distribution required in the microfluidic systems that are often used to study these systems in isolation. This paper introduces a neural network called the Multi-Directional Fluorescent Temperature Long Short-Term Memory Network (MFTLSTM) that can accurately calculate the temperature at every pixel in a fluorescent image to improve upon the standard fitting practice and other machine learning methods use to relate fluorescent data to temperature. This network takes advantage of the nature of heat diffusion in the image to achieve an accuracy of ±0.0199 K RMSE within the temperature range of 298K to 308 K with simulated data. When applied to experimental data from a 3D printed microfluidic device with a temperature range of 290 K to 380 K, it achieved an accuracy of ±0.0684 K RMSE. These results have the potential to allow high temperature resolution in biological systems than is available in many microfluidic devices.

Using Recurrent Neural Networks to Reconstruct Temperatures from Simulated Fluorescent Data for use in Bio-Microfluidics

Many biological systems have a narrow temperature range of operation, meaning high accuracy and spatial distribution level are needed to study these systems. Most temperature sensors cannot meet both the accuracy and spatial distribution required in the microfluidic systems that are often used to study these systems in isolation. This paper introduces a neural network called the Multi-Directional Fluorescent Temperature Long Short-Term Memory Network (MFTLSTM) that can accurately calculate the temperature at every pixel in a fluorescent image to improve upon the standard fitting practice and other machine learning methods use to relate fluorescent data to temperature. This network takes advantage of the nature of heat diffusion in the image to achieve an accuracy of ±0.0199 K RMSE within the temperature range of 298K to 308 K with simulated data. When applied to experimental data from a 3D printed microfluidic device with a temperature range of 290 K to 380 K, it achieved an accuracy of ±0.0684 K RMSE. These results have the potential to allow high temperature resolution in biological systems than is available in many microfluidic devices.

Spotlight on Luminescence Thermometry: Basics, Challenges, and Cutting-Edge Applications

Luminescence (nano)thermometry is a remote sensing technique that relies on the temperature dependency of the luminescence features (e.g., bandshape, peak energy or intensity, and excited state lifetimes and risetimes) of a phosphor to measure temperature. This technique provides precise thermal readouts with superior spatial resolution in short acquisition times. Although luminescence thermometry is just starting to become a more mature subject, it exhibits enormous potential in several areas, e.g., optoelectronics, photonics, micro- and nanofluidics, and nanomedicine. This work reviews the latest trends in the field, including the establishment of a comprehensive theoretical background and standardized practices. The reliability, repeatability, and reproducibility of the technique are also discussed, along with the use of multiparametric analysis and artificial-intelligence algorithms to enhance thermal readouts. In addition, examples are provided to underscore the challenges that luminescence thermometry faces, alongside the need for a continuous search and design of new materials, experimental techniques, and analysis procedures to improve the competitiveness, accessibility, and popularity of the technology.

Comparison of Performance between Single- and Multiparameter Luminescence Thermometry Methods Based on the Mn5+ Near-Infrared Emission

Herein, we investigate the performance of single- and multiparametric luminescence thermometry founded on the temperature-dependent spectral features of Ca₆BaP₄O₁₇:Mn⁵⁺ near-infrared emission. The material was prepared by a conventional steady-state synthesis, and its photoluminescence emission was measured from 7500 to 10,000 cm^-1 over the 293-373 K temperature range in 5 K increments. The spectra are composed of the emissions from ¹E → ³A₂ and ³T₂ → ³A₂ electronic transitions and Stokes and anti-Stokes vibronic sidebands at 320 cm^-1 and 800 cm^-1 from the maximum of ¹E → ³A₂ emission. Upon temperature increase, the ³T₂ and Stokes bands gained in intensity while the maximum of ¹E emission band is redshifted. We introduced the procedure for the linearization and feature scaling of input variables for linear multiparametric regression. Then, we experimentally determined accuracies and precisions of the luminescence thermometry based on luminescence intensity ratios between emissions from the ¹E and ³T₂ states, between Stokes and anti-Stokes emission sidebands, and at the ¹E energy maximum. The multiparametric luminescence thermometry involving the same spectral features showed similar performance, comparable to the best single-parameter thermometry.

Microsystem Advances through Integration with Artificial Intelligence

Microfluidics is a rapidly growing discipline that involves studying and manipulating fluids at reduced length scale and volume, typically on the scale of micro- or nanoliters. Under the reduced length scale and larger surface-to-volume ratio, advantages of low reagent consumption, faster reaction kinetics, and more compact systems are evident in microfluidics. However, miniaturization of microfluidic chips and systems introduces challenges of stricter tolerances in designing and controlling them for interdisciplinary applications. Recent advances in artificial intelligence (AI) have brought innovation to microfluidics from design, simulation, automation, and optimization to bioanalysis and data analytics. In microfluidics, the Navier-Stokes equations, which are partial differential equations describing viscous fluid motion that in complete form are known to not have a general analytical solution, can be simplified and have fair performance through numerical approximation due to low inertia and laminar flow. Approximation using neural networks trained by rules of physical knowledge introduces a new possibility to predict the physicochemical nature. The combination of microfluidics and automation can produce large amounts of data, where features and patterns that are difficult to discern by a human can be extracted by machine learning. Therefore, integration with AI introduces the potential to revolutionize the microfluidic workflow by enabling the precision control and automation of data analysis. Deployment of smart microfluidics may be tremendously beneficial in various applications in the future, including high-throughput drug discovery, rapid point-of-care-testing (POCT), and personalized medicine. In this review, we summarize key microfluidic advances integrated with AI and discuss the outlook and possibilities of combining AI and microfluidics.

Machine Learning Enabled Image Analysis of Time-Temperature Sensing Colloidal Arrays

Smart, responsive materials are required in various advanced applications ranging from anti-counterfeiting to autonomous sensing. Colloidal crystals are a versatile material class for optically based sensing applications owing to their photonic stopband. A careful combination of materials synthesis and colloidal mesostructure rendered such systems helpful in responding to stimuli such as gases, humidity, or temperature. Here, an approach is demonstrated to simultaneously and independently measure the time and temperature solely based on the inherent material properties of complex colloidal crystal mixtures. An array of colloidal crystals, each featuring unique film formation kinetics, is fabricated. Combined with machine learning-enabled image analysis, the colloidal crystal arrays can autonomously record isothermal heating events - readout proceeds by acquiring photographs of the applied sensor using a standard smartphone camera. The concept shows how the progressing use of machine learning in materials science has the potential to allow non-classical forms of data acquisition and evaluation. This can provide novel insights into multiparameter systems and simplify applications of novel materials.

Implementing neural network approach to create carbon-based optical nanosensor of heavy metal ions in liquid media

The present study is devoted to the creation of multifunctional optical carbon dots-based nanosensor to simultaneously measure concentrations of metal ions (Cu²⁺, Ni²⁺, Cr³⁺) and NO₃^- anion in liquid media. Such nanosensor operates on the basis of its fluorescence (FL) change under the influence of ions in the medium. However, the absence of analytical model, describing CD FL mechanism, the superposition of various luminescence quenching mechanisms during the interaction of carbon dots (CD) with cations, hampers the usage of classical approaches to solve this inverse multiparametric spectroscopic problem. To solve it neural networks were used that analyzed complex fluorescence signal from CD aqueous suspensions comprising Cu²⁺, Ni²⁺, Cr³⁺, NO₃^- ions in the concentration range from 0 to 4.95 mM. The following neural network architectures ensured optical spectroscopy inverse problem solution: multilayer perceptrons, 1D and 2D convolutional neural networks. The developed sensor enables simultaneous determination of the concentrations of heavy metal ions Cu²⁺, Ni²⁺, Cr³⁺ with a root mean squared error of 0.28 mM, 0.79 mM, 0.24 mM respectively. Based on the data given in the literature we can assert that the accuracy of the studied nanosensor satisfies the needs of monitoring the composition of waste and technological water. The developed nanosensor has a unique multimodality: with the simplicity of the synthesis protocol the sensor enables simultaneous determination of three heavy metal ions concentrations, while analogues are being developed mainly to measure the concentration of one (in rare cases two) heavy metal ions.

Demonstration of Neural Networks to Reconstruct Temperatures from Simulated Fluorescent Data Toward Use in Bio-microfluidics

Biological systems often have a narrow temperature range of operation, which require highly accurate spatially resolved temperature measurements, often near ±0.1 K. However, many temperature sensors cannot meet both accuracy and spatial distribution requirements, often because their accuracy is limited by data fitting and temperature reconstruction models. Machine learning algorithms have the potential to meet this need, but their usage in generating spatial distributions of temperature is severely lacking in the literature. This work presents the first instance of using neural networks to process fluorescent images to map the spatial distribution of temperature. Three standard network architectures were investigated using non-spatially resolved fluorescent thermometry (simply-connected feed-forward network) or during image or pixel identification (U-net and convolutional neural network, CNN). Simulated fluorescent images based on experimental data were generated based on known temperature distributions where Gaussian white noise with a standard deviation of ±0.1 K was added. The poor results from these standard networks motivated the creation of what is termed a moving CNN, with an RMSE error of ±0.23 K, where the elements of the matrix represent the neighboring pixels. Finally, the performance of this MCNN is investigated when trained and applied to three distinctive temperature distributions characteristic within microfluidic devices, where the fluorescent image is simulated at either three or five different wavelengths. The results demonstrate that having a minimum of 10 ^3.5 data points per temperature and the broadest range of temperatures during training provides temperature predictions nearest to the true temperatures of the images, with a minimum RMSE of ±0.15 K. When compared to traditional curve fitting techniques, this work demonstrates that greater accuracy when spatially mapping temperature from fluorescent images can be achieved when using convolutional neural networks.

Less is more: dimensionality reduction as a general strategy for more precise luminescence thermometry

Thermal resolution (also referred to as temperature uncertainty) establishes the minimum discernible temperature change sensed by luminescent thermometers and is a key figure of merit to rank them. Much has been done to minimize its value via probe optimization and correction of readout artifacts, but little effort was put into a better exploitation of calibration datasets. In this context, this work aims at providing a new perspective on the definition of luminescence-based thermometric parameters using dimensionality reduction techniques that emerged in the last years. The application of linear (Principal Component Analysis) and non-linear (t-distributed Stochastic Neighbor Embedding) transformations to the calibration datasets obtained from rare-earth nanoparticles and semiconductor nanocrystals resulted in an improvement in thermal resolution compared to the more classical intensity-based and ratiometric approaches. This, in turn, enabled precise monitoring of temperature changes smaller than 0.1 °C. The methods here presented allow choosing superior thermometric parameters compared to the more classical ones, pushing the performance of luminescent thermometers close to the experimentally achievable limits.

Exploiting machine learning for bestowing intelligence to microfluidics

Intelligent microfluidics is an emerging cross-discipline research area formed by combining microfluidics with machine learning. It uses the advantages of microfluidics, such as high throughput and controllability, and the powerful data processing capabilities of machine learning, resulting in improved systems in biotechnology and chemistry. Compared to traditional microfluidics using manual analysis methods, intelligent microfluidics needs less human intervention, and results in a more user-friendly experience with faster processing. There is a paucity of literature reviewing this burgeoning and highly promising cross-discipline. Therefore, we herein comprehensively and systematically summarize several aspects of microfluidic applications enabled by machine learning. We list the types of microfluidics used in intelligent microfluidic applications over the last five years, as well as the machine learning algorithms and the hardware used for training. We also present the most recent advances in key technologies, developments, challenges, and the emerging opportunities created by intelligent microfluidics.