Molecularly Engineered Room-Temperature Phosphorescence for Biomedical Application: From the Visible toward Second Near-Infrared Window

Chemical Reviews, Journal Year: 2023, Volume and Issue: 123(24), P. 13966 - 14037

Published: Nov. 22, 2023

Phosphorescence, characterized by luminescent lifetimes significantly longer than that of biological autofluorescence under ambient environment, is great value for biomedical applications. Academic evidence fluorescence imaging indicates virtually all metrics (sensitivity, resolution, and penetration depths) are improved when progressing into wavelength regions, especially the recently reported second near-infrared (NIR-II, 1000–1700 nm) window. Although emission probes does matter, it not clear whether guideline "the wavelength, better effect" still suitable developing phosphorescent probes. For tissue-specific bioimaging, long-lived probes, even if they emit visible phosphorescence, enable accurate visualization large deep tissues. studies dealing with bioimaging tiny architectures or dynamic physiopathological activities, prerequisite rigorous planning long-wavelength being aware cooperative contribution long wavelengths improving spatiotemporal depth, sensitivity bioimaging. In this Review, emerging molecular engineering methods room-temperature phosphorescence discussed through lens photophysical mechanisms. We highlight roles from to NIR-II windows toward bioapplications. To appreciate such advances, challenges prospects in rapidly growing described.

Language: Английский

---

Machine Learning Enhances the Performance of Bioreceptor-Free Biosensors

Sensors, Journal Year: 2021, Volume and Issue: 21(16), P. 5519 - 5519

Published: Aug. 17, 2021

Since their inception, biosensors have frequently employed simple regression models to calculate analyte composition based on the biosensor’s signal magnitude. Traditionally, bioreceptors provide excellent sensitivity and specificity biosensor. Increasingly, however, bioreceptor-free been developed for a wide range of applications. Without bioreceptor, maintaining strong low limit detection become major challenge. Machine learning (ML) has introduced improve performance these biosensors, effectively replacing bioreceptor with modeling gain specificity. Here, we present how ML used enhance biosensors. Particularly, discuss imaging, Enose Etongue, surface-enhanced Raman spectroscopy (SERS) Notably, principal component analysis (PCA) combined support vector machine (SVM) various artificial neural network (ANN) algorithms shown outstanding in variety tasks. We anticipate that will continue especially prospects sharing trained cloud computing mobile computation. To facilitate this, biosensing community would benefit from increased contributions open-access data repositories biosensor data.

Language: Английский

---

Deep Learning in Biomedical Optics

Lasers in Surgery and Medicine, Journal Year: 2021, Volume and Issue: 53(6), P. 748 - 775

Published: May 20, 2021

This article reviews deep learning applications in biomedical optics with a particular emphasis on image formation. The review is organized by imaging domains within and includes microscopy, fluorescence lifetime imaging, vivo widefield endoscopy, optical coherence tomography, photoacoustic diffuse functional brain imaging. For each of these domains, we summarize how has been applied highlight methods which can enable new capabilities for medicine. Challenges opportunities to improve translation adoption are also summarized. Lasers Surg. Med. © 2021 Wiley Periodicals LLC.

Language: Английский

---

Recent innovations in fluorescence lifetime imaging microscopy for biology and medicine

Journal of Biomedical Optics, Journal Year: 2021, Volume and Issue: 26(07)

Published: July 10, 2021

Significance: Fluorescence lifetime imaging microscopy (FLIM) measures the decay rate of fluorophores, thus providing insights into molecular interactions. FLIM is a powerful technique that widely used in biology and medicine. Aim: This perspective highlights some major advances instrumentation, analysis, biological clinical applications we have found impactful over last year. Approach: Innovations instrumentation resulted faster acquisition speeds, rapid large fields view, integration with complementary modalities such as single-molecule or light-sheet microscopy. There were significant developments analysis machine learning approaches to enhance processing fit-free techniques analyze images without priori knowledge, open-source resources. The advantages limitations these recent are summarized. Finally, year include label-free biology, ophthalmology, intraoperative imaging, new fluorescent probes, lifetime-based Förster resonance energy transfer measurements. Conclusions: A number high-quality publications signifies growing interest ensures continued technological improvements expanding biomedical research.

Language: Английский

---

Deep learning in macroscopic diffuse optical imaging

Journal of Biomedical Optics, Journal Year: 2022, Volume and Issue: 27(02)

Published: Feb. 25, 2022

Biomedical optics system design, image formation, and analysis have primarily been guided by classical physical modeling signal processing methodologies. Recently, however, deep learning (DL) has become a major paradigm in computational demonstrated utility numerous scientific domains various forms of data analysis.

Language: Английский

---

Single cell metabolic imaging of tumor and immune cells in vivo in melanoma bearing mice

Frontiers in Oncology, Journal Year: 2023, Volume and Issue: 13

Published: March 20, 2023

Introduction Metabolic reprogramming of cancer and immune cells occurs during tumorigenesis has a significant impact on progression. Unfortunately, current techniques to measure tumor cell metabolism require sample destruction and/or isolations that remove the spatial context. Two-photon fluorescence lifetime imaging microscopy (FLIM) autofluorescent metabolic coenzymes nicotinamide adenine dinucleotide (phosphate) (NAD(P)H) flavin (FAD) provides in vivo images at single level. Methods Here, we report an immunocompetent mCherry reporter mouse model for express CD4 either differentiation or CD8 their mature state perform within syngeneic B78 melanoma model. We also algorithm segmentation mCherry-expressing images. Results found tumors exhibited decreased FAD mean increased proportion bound compared spleens. Tumor infiltrating size from These changes are consistent with shift towards activation proliferation protein-bound same tumor. Single heterogeneity was observed both . Discussion This approach can be used monitor study promising treatments native

Language: Английский

---

FLIMJ: An open-source ImageJ toolkit for fluorescence lifetime image data analysis

PLoS ONE, Journal Year: 2020, Volume and Issue: 15(12), P. e0238327 - e0238327

Published: Dec. 30, 2020

In the field of fluorescence microscopy, there is continued demand for dynamic technologies that can exploit complete information from every pixel an image. One imaging technique with proven ability yielding additional Fluorescence Lifetime Imaging Microscopy (FLIM). FLIM allows measurement how long a fluorophore stays in excited energy state, and this affected by changes its chemical microenvironment, such as proximity to other fluorophores, pH, hydrophobic regions. This provide about microenvironment has made powerful tool cellular studies ranging metabolic measuring distances between proteins. The increased use necessitated development computational tools integrating analysis image data processing. To address need, we have created FLIMJ, ImageJ plugin toolkit easy extensible workflows data. Built on FLIMLib decay curve fitting library Ops framework, FLIMJ offers routines seamless integration many components, be extended create complex workflows. Building also enables FLIMJ's used Jupyter notebooks integrate naturally science-friendly programming in, e.g., Python Groovy. We show extensibility two scenarios: lifetime-based segmentation colocalization. validate comparing them against industry standards.

Language: Английский

---

Phasor-based image segmentation: machine learning clustering techniques

Biomedical Optics Express, Journal Year: 2021, Volume and Issue: 12(6), P. 3410 - 3410

Published: April 23, 2021

The phasor approach is a well-established method for data visualization and image analysis in spectral lifetime fluorescence microscopy. Nevertheless, it typically applied user-dependent manner by manually selecting regions of interest on the space to find distinct images. In this paper we present our work using machine learning clustering techniques establish an unsupervised automatic that can be used identifying populations fluorescent species imaging. We demonstrate both synthetic data, created sampling photon arrival times plotting distributions plot, real live cells samples, staining cellular organelles with selection commercial probes.

Language: Английский

---

Review of Fluorescence Lifetime Imaging Microscopy (FLIM) Data Analysis Using Machine Learning

Journal of Experimental and Theoretical Analyses, Journal Year: 2023, Volume and Issue: 1(1), P. 44 - 63

Published: Sept. 21, 2023

Fluorescence lifetime imaging microscopy (FLIM) has emerged as a promising tool for all scientific studies in recent years. However, the utilization of FLIM data requires complex modeling techniques, such curve-fitting procedures. These conventional procedures are not only computationally intensive but also time-consuming. To address this limitation, machine learning (ML), particularly deep (DL), can be employed. This review aims to focus on ML and DL methods analysis. Subsequently, strategies evaluating discussed, consisting preprocessing, modeling, inverse modeling. Additionally, advantages reviewed deliberated alongside future implications. Furthermore, several freely available software packages analyzing highlighted.

Language: Английский

---

Coupling a recurrent neural network to SPAD TCSPC systems for real-time fluorescence lifetime imaging

Scientific Reports, Journal Year: 2024, Volume and Issue: 14(1)

Published: Feb. 8, 2024

Abstract Fluorescence lifetime imaging (FLI) has been receiving increased attention in recent years as a powerful diagnostic technique biological and medical research. However, existing FLI systems often suffer from tradeoff between processing speed, accuracy, robustness. Inspired by the concept of Edge Artificial Intelligence (Edge AI), we propose robust approach that enables fast with no degradation accuracy. This couples recurrent neural network (RNN), which is trained to estimate fluorescence directly raw timestamps without building histograms, SPAD TCSPC systems, thereby drastically reducing transfer data volumes hardware resource utilization, enabling real-time acquisition. We train two variants RNN on synthetic dataset compare results those obtained using center-of-mass method (CMM) least squares fitting (LS fitting). Results demonstrate variants, gated unit (GRU) long short-term memory (LSTM), are comparable CMM LS terms while outperforming them presence background noise large margin. To explore ultimate limits approach, derive Cramer-Rao lower bound measurement, showing yields estimations near-optimal precision. operation, build microscope based an system comprising 32 $$\times $$ × sensor named Piccolo. Four quantized GRU cores, capable up 4 million photons per second, deployed Xilinx Kintex-7 FPGA controls Powered GRU, setup can retrieve images at 10 frames second. The proposed promising ideally suited for biomedical applications, including imaging, diagnostics, fluorescence-assisted surgery, etc.

Language: Английский

---