Fiber optic-based integrated system for in vivo multiscale pharmacokinetic monitoring

This paper presents the development of a fiber-optic-based fluorescence detection system for multi-scale monitoring of drug distribution in living animals. The integrated system utilized dual laser sources at the wavelengths of 488 nm and 650 nm and three photomultiplier channels for multi-color fluorescence detection. The emission spectra of fluorescent substances were tracked using the time-resolved fluorescence spectroscopy module to continuously monitor their blood kinetics. The fiber bundle, consisting of 30,000 optic filaments, was designed for wide-field mesoscopic imaging of the drug’s interactions within organs. The inclusion of a gradient refractive index (GRIN) lens within the setup enabled fluorescence confocal laser scanning microscopy to visualize the drug distribution at the cellular level. The system performance was verified by imaging hepatic and renal tissues in mice using cadmium telluride quantum dots (CdTe QDs) and R3. By acquiring multi-level images and real-time data, our integrated system underscores its potential as a potent tool for drug assessment, specifically within the realms of pharmacokinetic and pharmacodynamic investigations.


Introduction
The kinetics and metabolism of drugs involve drug-target binding with its downstream effects that occur across different scales, including systemic, cellular, and molecular levels.Typically, wholebody investigations of a drug's behavior within animal models, such as bioavailability, tissue distribution profiles, half-life, and clearance rates, can help determine the overall fate of drugs and their efficacy before clinical trials.Blood kinetics and drug concentrations, determined by the drug's formulation and route of administration, are crucial for predicting their pharmacological effects.Cellular and subcellular-level studies provide insights into drug metabolism, including how drugs are internalized by cell, how they undergo intracellular trafficking, and affect specific cellular processes within different types of cells.By integrating knowledge of the dynamic distribution characteristics of drugs in blood vessels, major metabolic organs, and target cells, researchers can gain a comprehensive understanding of pharmacokinetics and metabolism, which is valuable for optimizing dosage regimens, predicting drug-drug interactions, and ensuring drug safety and efficacy in clinical practice.
Conventional end-point methods for obtaining pharmacokinetic (PK) properties of drugs involve the sampling of blood or tissues at various time points.Organs were harvested from animals for ex vivo analysis.Traditional histological techniques, including immunohistochemistry and fluorescence microscopy, can provide further confirmation about drug location and accumulation in specific tissues.Tissue homogenates can be sent for quantitative measurement through spectrophotometry [1], chromatography, and enzyme immunoassay [2].However, this may require a large number of animals, which is ethically and financially costly.Moreover, only discrete data points are obtained, and inter-individual variability can limit continuity and the accuracy of the drug's kinetics.Therefore, in vivo imaging techniques that allow for continuous monitoring are increasingly explored for tracing the distribution of drugs at different scales.
By labelling a drug with a radionuclide or contrast agent, the systemic distribution of drugs within the body and their elimination can be tracked using whole-body imaging techniques, such as positron emission computed tomography (PET), single-photon emission computed tomography (SPECT), computed X-ray tomography (CT), or Magnetic Resonance Imaging (MRI) [3][4][5][6].Fluorescence-based approaches have been employed to analyze drug perfusion and accumulation in specific organs or tissues by recording the dynamic changes in the intensity of fluorescently labelled drugs over time.Due to its high sensitivity, it is particularly beneficial for studying drugs with low bioavailability or those that undergo extensive metabolism.To visualize drug-cell interactions, uptake, and intracellular distribution at cellular and subcellular levels, intravital microscopy or confocal microscopy can be used by offering high resolution.Nevertheless, limited light penetration and background interference from biological samples often restrict the applicability of fluorescence-based PK profiling to drugs mainly amenable to ex vivo analysis [7].
Fiber optic-based fluorescence imaging techniques bridge the gap in in vivo microscopic imaging by providing spatial resolution at the subcellular level.It employs a bundle of optical fibers to transmit excitation light and collect fluorescence signals, which are then processed for image reconstruction.Compared to traditional imaging techniques, the use of fiber optic preserves the integrity of the light path, resulting in improved resolution.In addition, fiber optic probes are highly flexible to be maneuvered into various positions and orientations, allowing for reaching inaccessible areas.This approach also enables real-time monitoring of dynamic drug metabolism processes within living organisms, eliminating the need for tissue harvest and sample preparation [8][9][10][11][12].By creating a two-dimensional (2D) raster scanning pattern and conjugating a pinhole aperture in front of each optical fiber to filter out-of-focus light, sharp and contrast-rich images of focal planes could be obtained at the subcellular level.
Fiber optic-based fluorescence imaging techniques fill the gap in in vivo microscopic imaging by providing spatial resolution at the subcellular level.It employs a bundle of optical fibers to transmit excitation light and collect fluorescence signals, which are then be processed for image reconstruction.Compared to traditional imaging techniques, the use of fiber optic preserves the integrity of the light path, resulting in improved resolution.In addition, fiber optic probes are highly flexible to be maneuvered into various positions and orientations, allowing for reaching inaccessible areas.This non-destructive imaging approach enables real-time monitoring of dynamic processes of drug metabolism within living organisms, eliminating the need for tissue harvest and sample preparation [8][9][10][11][12].By creating a two-dimensional (2D) raster scanning pattern and conjugating a pinhole aperture in front of each optical fiber to filter out the out-of-focus light, sharp and contrast-rich images of focal planes could be obtained at subcellular level.
In this paper, with the aim of visualizing and tracking the drug distribution and metabolism at different levels in living animals, a multi-scale imaging system was developed based on the optical fiber fluorescence detection.Three functional modules, namely, macroscopic monitoring of blood kinetics of drugs, microscopic imaging of metabolic organs, and mesoscopic imaging of drug transport within target cells, were integrated.The system is equipped with laser sources with dual-wavelengths and three distinct channels for fluorescence detection, which is compatible with a wide range of fluorescent labeling substances, allowing for the simultaneous visualization of specific subcellular components.With further development, we believe that the system can be accessible to a broader scenario of in vivo studies on the physiological and pharmacokinetic properties of drugs, contributing to improved drug design and administration protocols.

System design
The system comprises several components: a dual-wavelength laser light source, a timeresolved fluorescence spectroscopy for real-time monitoring of blood kinetics of drugs, a fiber optic fluorescence microscopy for tissue-level imaging, a laser scanning confocal fluorescence microscopy for cell-level imaging, and a system control and processing module equipped with embedded software.The schematic diagram of the system composition is presented in Fig. 1  for the simultaneous visualization of specific subcellular components.With further development, we believe that the system can be accessible to a broader scenario of in vivo studies on the physiological and pharmacokinetic properties of drugs, contributing to improved drug design and administration protocols.

System Design
The system comprises several components: a dual-wavelength laser light source, a timeresolved fluorescence spectroscopy for real-time monitoring of blood kinetics of drugs, a fiber optic fluorescence microscopy for tissue-level imaging, a laser scanning confocal fluorescence microscopy for cell-level imaging, and a system control and processing module equipped with embedded software.The schematic diagram of the system composition is presented in Fig. 1(a), and the optical path diagram is depicted in Fig. 1(b).Fig. 2(e) shows the overview of the hardware components of the integrated fiber optic-based system for multiscale imaging.

Laser source module
The light sources in the system consist of two semiconductor continuous-wave lasers with wavelengths of 488 nm (Laser 1, MBL-SF-488-70 mW, CNI, China) and 650 nm (Laser 2, MRL-III-650L-200 mW, CNI, China).These wavelengths were chosen to excite a variety of widely used fluorescent dyes, such as fluorescein isothiocyanate isomer (FITC), 3,3'dioctadecyloxacarbocyanine perchlorate (DiO), and Cyanine5 (Cy5) series.Light distribution and light intensity adjustment can be achieved through a combination of half-wave plates (HW) and beam splitters (SP) [13].Electrically controlled light-blocking shutters (SH1 and SH2) are equipped in front of both laser diodes for switching the optical paths.As shown in Fig. 1(b), the light from Laser 2 is directed through the reflector (M1) and the dichroic mirror (DM3) before entering the main optical path, converging with the light from Laser 1.

Time-resolved fluorescence spectroscopy module
This module is designed for continuously detecting the emission intensity of different fluorescent drug substances and quantifying their accumulation in blood over time.Monitoring changes in fluorescence intensity of drugs in blood using a high sensitivity spectrometer (Ocean Optics, QE Pro, SNR, 1000:1).The excitation light from the laser source reaches the dichroic mirror (DM1) after being divided by the splitter (SP1).Then it is coupled into the quartz multi-fiber bundle (FB1) through the lens (L2) to excite the sample.Once the sample is excited, the generated fluorescent signal is transmitted through the fiber bundle's port and returned to DM1.The filtered light is collected by a portable QE Pro Fiber Optic Spectrometer (SD). Figure 2(a) depicts the configuration of the multi-fiber bundle probe with a single 400 µm excitation fiber at its center and surrounded by nine 200 µm fibers for collecting emitting fluorescent signals.

Tissue-level fluorescent microscopy imaging module
This module is mainly used for real-time wide-field mesoscopic imaging.Figure 2(b) illustrates the design of the image transmission fiber bundle, which consists of 30000 fiber optic filaments arranged in a regular hexagon with a diameter of 4 µm.Figure 2(f) shows the top-down view of the hardware composition.After passing through the dichroic mirror (DM2) and then entering the lens (L3), the excitation light is coupled into the imaging fiber bundle (FB2) [14].Next, the light beam is adjusted by a series of micro lenses (L4, L5, and L6) to fill the aperture of the micro-objective (OBJ1) and to capture an image with a diameter of 1 mm.To ensure uniform irradiation and high-quality fluorescence imaging, the objective lens is surrounded by an excitation light guide optical fiber composed of 200 fiber optic filaments with a diameter of 30 µm.This prevents the excitation light from interfering with the reflected fluorescence image.The fluorescence signal is collected by the objective lens and directed back to DM2 along the original optical path.It is then reflected to the detection optical path, filtered by a filter (F2), and enters the electron-multiplying charge-coupled device (EMCCD) for mesoscopic imaging.The fiber optic probe is made of stainless steel and has a diameter of 1.1 mm.The module resolution is determined by referencing the 1951 USAF resolution test target (MIL-STD-150A standard) with known line width and spacing.To measure the corresponding field of view size, the fiber optic probe is placed at various distances in the Z-axis direction from the standard scale ruler (Thorlabs), with a spacing of 1 mm between large scale lines and 0.5 mm between small scale lines.

Cell-level laser scanning confocal fluorescence microscopy imaging module
As depicted in Fig. 1(b), after the two excitation lights, with wavelengths of 488 nm and 650 nm, are reflected by the mirror (M2), they pass through dichroic mirrors and enter a spatial filtering system consisting of objective lens 1 (Obj1), a pinhole (40 µm, Opm, China), and objective lens 2 (Obj2).The light beams then reach a 2D scanning galvanometer (GM) for beam scanning.The scanning beam is collimated by a 4f optical system, which includes lens groups (L8 and L9), and then enters the objective lens (OBJ3).The objective lens couples the scanning beam into the imaging fiber bundle (FB3, containing ≥30000 fibers, each with a core diameter of 1.1 µm).Each fiber serves as both a point light source and a confocal pinhole to filter out stray light around the focal spot to enhance image resolution.Once the scanning beam passes through the fiber bundle, it is coupled into a 4.65× micro-objective (NA = 0.8; Diameter: 3 mm; Length: 7.5 mm; Working distance: 80 µm) and a gradient refractive index lens (GRIN Lens) with a variable refractive index for further spot compression [15][16][17].The GRIN Lens enables the scanning excitation of the sample, as shown in Fig. 2(c) and Fig. 2(d).The pinhole is conjugated to the focal point of the Obj3 and the micro-objective, ensuring that only the signal from the focal point of the micro-objective is collected.The generated fluorescence signal is reflected back to the detection light path by the dichroic mirrors (DM4, DM5, and DM6).It then passes through filters (F3, F4, and F5) to block non-signal light and enters lenses (L10, L11, and L13).After being focused by the lenses, the signal is coupled into multimode fibers (DMF1, DMF2, and DMF3).The fluorescence signal, containing high-resolution information, is introduced into the photomultipliers (PMT1, PMT2, and PMT3, H7420-40, Hamamatsu, Japan) through the fibers and converted into electrical signal.Finally, by coordinating the laser scanning and data acquisition timing, the fluorescence signals at various points on the focal plane are collected, and a two-dimensional fluorescence image is obtained through image reconstruction.The top-down view of the hardware composition of the laser scanning confocal fluorescence microscopy imaging module is shown in Fig. 2(g).The entire optical imaging setup was mounted on a specialized vibration damping platform to minimize extraneous vibrations that could perturb the mode coupling process.The fiber probes themselves were securely held in place using optical mounts, and fine positional adjustments of the optical fibers were made possible through the use of a precision three-axis displacement stage.standard) with known line width and spacing.To measure the corresponding field of view size, the fiber optic probe is placed at various distances in the Z-axis direction from the standard scale ruler (Thorlabs), with a spacing of 1 mm between large scale lines and 0.5 mm between small scale lines.

Cell-level laser scanning confocal fluorescence microscopy imaging module
As depicted in Fig. 1(b), after the two excitation lights, with wavelengths of 488 nm and 650 nm, are reflected by the mirror (M2), they pass through dichroic mirrors and enter a spatial filtering system consisting of objective lens 1 (Obj1), a pinhole (40 µm, Opm, China), and objective lens 2 (Obj2).The light beams then reach a 2D scanning galvanometer (GM) for beam scanning.The scanning beam is collimated by a 4f optical system, which includes lens groups (L8 and L9), and then enters the objective lens (OBJ3).The objective lens couples the scanning beam into the imaging fiber bundle (FB3, containing ≥30000 fibers, each with a core diameter of 1.1 µm).Each fiber serves as both a point light source and a confocal pinhole to filter out stray light around the focal spot to enhance image resolution.Once the scanning beam passes through the fiber bundle, it is coupled into a 4.65× micro-objective (NA = 0.8; Diameter: 3 mm;

System control and timing
The integrated system incorporates three fluorescence monitoring and imaging modules for multiple scales, which utilize a shared excitation light path.In current system setup, Channel 1 exhibits a coupling efficiency of 80%, and Channel 2 has a lower coupling efficiency of 50%, both within an acceptable range for the proof-of-concept stage.The coupling in Channel 3 is accomplished through an objective lens into a bundle of fibers, which involve a longer optical path and, consequently, the efficiency is intrinsically more challenging to optimize, the coupling efficiency is approximately 20%.The light path must deliver specific excitation wavelengths tailored to the distinct requirements of each module.Consequently, synchronizing optical switches and data acquisition across these modules demands an orchestrated operation between sophisticated software interfaces and precision control hardware.
The core function of the control system is to temporally coordinate the initialization of different modules to facilitate data acquisition across varying resolutions.This synchronization is achieved with a multi-channel high-speed data acquisition card (PCI 6110, National Instruments, United States).Figure 3(a) depicts the coordinated timing scheme utilized by the three channels, namely, fluorescence spectrum detection, mesoscopic imaging, and microscopic imaging.These functions operate in concert during a sampling cycle through time-division multiplexing.Keeping the positional stability of the cell-level confocal fluorescence imaging module is critical, as minor probe drifts can impact on the imaging integrity.Therefore, a series of three PMTs works with dichroic mirrors and filters to concurrently capture three distinct fluorescence signals.During the in vivo experiments, we are aware of the potential for ambient light to interfere with our imaging data and have taken measures to conduct experiments in a controlled, dark environment to minimize this.Furthermore, a background calibration algorithm was designed to optimize interference.
As demonstrated in Fig. 3(b), we have designed a temporal sequence to manage each function of the integrated system.The two laser sources at wavelengths of 488 nm and 650 nm are connected to a 220 V power supply, maintaining them in a default "on" state.The corresponding laser shutters, controlled separately by their proprietary controllers, are managed by the master computer system, which governs their open or closed status.The "open" state is indicated by rectangles in Fig. 3(b).The fluorescence spectrometer interfaces with the master computer via USB, which orchestrates its initiation and sampling duration.During each sampling period, the EMCCD dictates the acquisition period, determining both laser shutter activation duration and temporal window for filter wheel transitions.The GM and PMTs interface with the master computer through the data acquisition card, with the GM initiating the scans.This synchronization allows the PMTs to concurrently capture fluorescence intensity data from the targeted cells at the GM's position.Following the completion of a sampling cycle, the system idles until the onset of the subsequent cycle, as determined by the pre-set sampling frequency.For the in vivo monitoring, an ICR mouse weighing approximately 20 g was secured on a stereotactic apparatus and the anesthesia machine's tracheal mask was placed over the mouse's mouth and nose for continuous administration of isoflurane.An incision was made along the mouse's neck to fully expose the jugular vein using sterilized surgical tools.The optical fiber probe of our integrated system was then positioned over the exposed vessel to facilitate the real-time monitoring of dynamic fluorescence intensity fluctuations within the bloodstream.Each mouse (n = 15) received an intravenous injection of 750 µg of fluorescent Cadmium Telluride quantum dots (CdTe QDs, excitation/emission: 488 nm/640 nm) dissolved in 200 µl of saline via the tail vein.

Sample preparation for in vivo tissue-level fluorescence imaging
The secondary module of our multiscale imaging system is the fiber optic fluorescence microscopy unit, tailored to monitor drug distribution within biological specimens at the tissue scale.The optical resolution achievable via this microscopy technique is governed by the numerical aperture of the optical fibers and the specification of system parameters pertinent to the specific configuration [18].For resolution assessment, we employed the 1951 USAF Resolution Test Target, which conforms to the MIL-STD-150A standard.Subsequently, we proceeded to image two fluorescent dyes, namely CdTe QDs and a fluorescent reagent R3, to validate the capabilities of our fiber optic system with dual-wavelength excitation.CdTe QDs are known for their robust fluorescence intensity and stability, exhibiting peak absorption wavelengths in the 480-490 nm range and maximum emission wavelengths from 630-650 nm [19].R3, a fluorescent agent developed in collaboration with our research group, is utilized for the labeling of peptides, proteins, and nanoscale drug delivery systems [20].It generates a potent fluorescence signal with peak absorption and emission wavelengths at 650 nm and 730 nm, respectively.CdTe QDs and R3 were injected into mice via the tail vein at concentrations of 1.5 mg/mL and 0.5 mg/mL respectively.These doses were chosen based on preliminary studies that established the optimum concentration for visualization without inducing toxicity.The 488 nm laser was adopted to induce fluorescence in CdTe QDs, and the 650 nm laser was utilized to excite R3.
H&E staining is usually used to evaluate histological morphology and perform pathological analysis based on it [21].Therefore, we designed a comparison between in vivo captured images and in vitro tissue slices to ensure that the images captured by the system are accurate and can be used for scientific research and experimental analysis.To prepare H&E-stained sections, firstly, select mice of the same experimental breed, separate liver tissue blocks from euthanized them, soak them in a 4% paraformaldehyde solution for 24 hours, dehydrate them with xylene, embed them in paraffin, and finally stain the sections to obtain H&E-stained sections of the tissue.

Sample preparation for cell-level confocal fluorescence microscopy imaging
Considering the short working distance of the micro lens in the fiber optic probe (only 80 µm), we measured the resolution of the laser scanning confocal fluorescence microscopy module in our system by imaging fluorescent beads with a diameter of 500 nm [22][23][24].These calibration beads have an excitation peak of approximately 488 nm and an emission peak of approximately 515 nm.A monolayer of diluted fluorescent beads was prepared on a microscope slide and left to air-dry.The microscope slide was then fixed onto a Thorlabs MAX313D/M stage, which offers both coarse and fine adjustment range dials.The coarse adjustment has a scale of 10 µm, while the fine adjustment dial has a scale of 1 µm.By moving the microscope slide in the X and Y planes controlled by the three-axis stage, we documented images of the fluorescent beads before and after movement using the fiber optic probe.The pixel coordinate difference between the two images was obtained using ImageJ.By correlating the pixel shifts observed in the images with the actual known displacement of the fluorescent beads, we calculated the physical distance represented by each pixel, from which we defined the system's resolution accordingly.
We then proceeded in vivo cellular imaging using CdTe QDs and R3 as fluorescent markers.The fiber probe was inserted into the mice through a minimally invasive abdominal incision and positioned proximal to the liver and kidney tissues at the requisite working distance.The fluorescent dyes were administered via the tail vein and were excited by 488 nm and 650 nm lasers.By fine-tuning the three-axis stage, fluorescent images were captured for each PMT, depicting various staining outcomes from the identical region.
To affirm the fidelity of images captured in vivo using our integrated system, we aligned them with ex vivo cryo-sections and histological sections stained with hematoxylin and eosin (H&E).Before cryo-sectioning, the liver and kidney tissues were harvested and fixed in 4% paraformaldehyde for 2 h.The tissues were rinsed with phosphate-buffered saline (PBS), immersed in 30% sucrose solution for cryoprotection, and embedded in optimal cutting temperature (OCT) compound.Observing and storing slices through the optical microscope.H&E staining is a widely accepted technique, enabling the elucidation of morphological characteristics and the facilitation of pathological diagnosis [21].Briefly, liver tissue blocks were excised and fixed in a 4% paraformaldehyde solution for 24 h, dehydrated using xylene, embedded in paraffin, and ultimately sectioned and stained for comparative analysis [25].

Plasma kinetics of CdTe QDs by in vivo fluorescence monitoring
The dynamic behaviors of drugs in the bloodstream, including aspects such as absorption, distribution, metabolism, and excretion (ADME), are crucial for understanding how drugs are processed in the body and thus are vital for proper drug development [26][27][28].Conventional plasma kinetic studies involve a combination of blood sampling at different time points and analyzing blood drug concentrations in a discrete manner.Continuous monitoring techniques are more favorable for more precise determination of PK parameters over time, particularly in situations where drug levels fluctuate rapidly or exhibit short half-lives [29].The first function of our integrated system is to detect the dynamic changes in plasma drug concentrations in real time by adopting fiber optic-based fluorescence intensity measurement.CdTe QDs hold promise as drug carriers and therapeutics while possessing strong fluorescence suitable for in vivo imaging and tracking.We thus tested the system's performance by tracking the time-resolved changes of CdTe QDs concentration in blood of living animals.
Fig. 4(a) shows the representative emission spectra of the prepared CdTe QDs solution under excitation of a 488 nm wavelength laser.Figure 4(b) shows the spectral pattern of CdTe quantum dots detected by the system.The fluorescence intensity of CdTe QDs reaches the maximum value at the wavelength between 640-650 nm.During the in vivo study, fluorescence spectra of CdTe QDs were acquired by the fiber-based fluorescence spectroscopy in our system, starting from the time of CdTe QDs injection into the tail vein, with data collected continuously every three seconds for one hour.The system detection sensitivity can reach 10 ng/ml.The light intensity value in the 640-650 nm wavelength range was extracted from the emitted light spectra and averaged, and the mean plasma concentration-time profile of CdTe QDs is presented in Fig. 4(c).
As shown in Fig. 4(c), the plasma concentration of CdTe QDs post-administration exhibits an immediate increase, culminating in a peak within minutes.This is followed by a progressive decline, with a steeper decrease occurring within the first 20 min.The fluorescence intensity changes of QDs in plasma over time concurred with what is expected in a typical pharmacokinetic profile, revealing an initial rapid rise in concentration following intravenous administration (absorption and distribution phases) followed by a gradual declining concentration over time (metabolic and elimination phases).The transient presence of CdTe QDs in plasma, as detected by our fiber optic system, is consistent with the findings from previous in vivo assessments using intermittent blood sampling and inductively coupled plasma-mass spectrometry (ICP-MS) [30].Our system enables uninterrupted tracking of plasma kinetics, thereby eliminating the need CdTe QDs were acquired by the fiber-based fluorescence spectroscopy in our system, starting from the time of CdTe QDs injection into the tail vein, with data collected continuously every three seconds for one hour.The system detection sensitivity can reach 10ng/ml.The light intensity value in the 640-650 nm wavelength range was extracted from the emitted light spectra and averaged, and the mean plasma concentration-time profile of CdTe QDs is presented in Fig. 4(c).
As shown in Fig. 4(c), the plasma concentration of CdTe QDs post-administration exhibits an immediate increase, culminating in a peak within minutes.This is followed by a progressive decline, with a steeper decrease occurring within the first 20 min.The fluorescence intensity changes of QDs in plasma over time concurred with what is expected in a typical pharmacokinetic profile, revealing an initial rapid rise in concentration following intravenous administration (absorption and distribution phases) followed by a gradual declining concentration over time (metabolic and elimination phases).The transient presence of CdTe QDs in plasma, as detected by our fiber optic system, is consistent with the findings from previous in vivo assessments using intermittent blood sampling and inductively coupled plasmamass spectrometry (ICP-MS) [31].Our system enables uninterrupted tracking of plasma  for periodic sampling interventions or end-point procedures.By analyzing the data points on the curve, we can deduce pharmacokinetic parameters for the drug using a non-compartmental model within the physiological context.For instance, the area under the curve (AUC) provides an aggregate measure of the systemic exposure to the drug, the peak concentration (Cmax) and the time to reach peak concentration (Tmax) offer insights into the drug's absorption rate, and the half-life (t1/2) can reveal how long the drug persists within the systemic circulation.

Resolution determination of microscopic fluorescent imaging
Firstly, the 1951 USAF resolution test target (Fig. 5(a)) was precisely positioned onto a uniformly illuminated background that provides a consistent white light output.The filter turret was set to a bypass (or open) position, thereby omitting any filtration during the operation.Next, the optical axis of the optic probe was meticulously aligned with the designated resolution marker on the test chart.The probe-to-chart distance was adjusted to achieve optimal focal alignment.Once the target was brought into sharp focus, high-resolution image was acquired using the EMCCD.Figure 5(b) illustrates the resultant micrograph, showing a homogeneous intensity distribution across the end-face of the fiber bundle.The white arrow highlights the minimum resolvable pattern by the imaging system.By referring to the corresponding resolution chart, this particular element equates to a resolutive capacity of 50.8-line pairs per millimeter (lp/mm).Therefore, the spatial resolution of the imaging system is inferred to be approximately 9.8 µm.

Resolution determination of confocal fluorescence microscopic imaging
The fluorescence microsphere imaging results of the target cell imaging system are shown in Fig. 5(c).The fiber optic fibers in the imaging fiber bundle are arranged in a hexagonal shape, and then a straight line is drawn along the direction of the fiber optic bundle arrangement, which is the yellow line in Fig. 5(d), and the grayscale value change curve of the pixel points on the line is drawn, as shown in Fig. 5(e),the distance between adjacent peaks in the curve represents the distance between the centers of adjacent fiber cores, with an average interval of 7.46 pixels.Since each pixel represents 0.147 µm, we assume the diameter of each fiber (d) to be 1.1 µm.For the hexagon arrangement, the sampling interval of pixels in the image bundle at 0°, 60°, and 120°d irections is ( √ 3)/2 d, while it is 1/2 d at 30°, 90°, and 150°directions.The optical imaging resolution was calculated as N = 1/( √ 3d) and N = 1/d, r, respectively.Consequently, a resolution of 525 lp/mm was calculated at 0°, 60°, and 120°directions, with an optical imaging resolution of 0.95µm, and the resolution at 30°, 90°, and 150°directions was 909 lp/mm, indicating an optical imaging resolution of 0.55 µm.
Once the target was brought into sharp focus, high-resolution image was acquired using the EMCCD.Therefore, the spatial resolution of the imaging system is inferred to be approximately 9.8 μm.After comparing the staining of the two dyes, we observed slight differences in the image qualities, which may further prove the high resolution of the microscopic imaging.To be more specific, R3, as a lipophilic dye, is more prone to aggregation and metabolism through the liver.It thus demonstrated a more vibrant and sharp staining with enhanced clarity and contrast.Conversely, we use water-soluble CdTe Dots, which has lower expressive power in liver tissue distribution than R3 dye, and the clarity of the image is inferior to the former.

Confocal fluorescence microscopy imaging for in vivo monitoring
Visualizing the cellular uptake and dynamics of the drugs at the cellular level is crucial for identifying potential drug targets, understanding their mechanisms of action and potential efficacy.The confocal fluorescence microscopy imaging function of our system facilitates cellular-level observation of drug distribution.Fig. 6(d) demonstrates liver cells labeled with CdTe QDs taken in situ.The vibrant hues within these images correspond to the structural configurations of liver cells as revealed by the fluorescence of CdTe QDs.The arrangement and morphology of the cells share similarity with images obtained ex vivo, which was derived from cryo-sections (Fig. 6e) and H&E-stained sections (Fig. 6f).Fig. 6(g-i) present in situ images taken from liver cells stained with R3, as well as ex vivo images taken from cryo-sections and H&Estained sections.The live-cell imaging by both dyes showcased a substantial degree of concordance with respect to regional distribution and morphological characteristics when

Confocal fluorescence microscopy imaging for in vivo monitoring
Visualizing the cellular uptake and dynamics of the drugs at the cellular level is crucial for identifying potential drug targets, understanding their mechanisms of action and potential efficacy.The confocal fluorescence microscopy imaging function of our system facilitates cellular-level observation of drug distribution.Figure 6(d) demonstrates liver cells labeled with CdTe QDs taken in situ.The vibrant hues within these images correspond to the structural configurations of liver cells as revealed by the fluorescence of CdTe QDs.The arrangement and morphology of the cells share similarity with images obtained ex vivo, which was derived from cryo-sections (Fig. 6e) and H&E-stained sections (Fig. 6(f)).Figure 6(g-i) present in situ images taken from liver cells stained with R3, as well as ex vivo images taken from cryo-sections and H&E-stained sections.The live-cell imaging by both dyes showcased a substantial degree of concordance with respect to regional distribution and morphological characteristics when comparing with ex vivo observation of the cryosections and H&E-stained sections of the same tissue.Upon comparative analysis of the resultant images of both dyes, subtle differences in the imaging resolution were detected.Specifically, R3, being a lipophilic probe, is more inclined to accumulate within the liver, resulting in more pronounced and sharply defined images characterized by augmented clarity and contrast.In contrast, the hydrophilic CdTe QDs demonstrated a less robust representation of tissue distribution and inferior image clarity when compared to R3.This outcome is reflective of the discrete solubility and tissue interaction properties of the two fluorescent compounds.Similar to what observed in the liver tissue, the distinctive patterns achieved with R3 staining are also discernible in renal tissues.Figure 6(j-l) illustrate renal cells treated with R3 stain, along with their cryosections and H&E-stained images.The visualized morphology of the renal tubules resembles a series of tubes, typically spanning lengths of approximately 30-50 mm, representing a core component of the nephron.

5.
The multiscale system represents an advanced integration of macroscopic monitoring, mesoscopic imaging, and microscopic observation, allowing for the in vivo detection of multiple fluorescent dyes.This system enables detailed visualization of drug distribution and transportation within the body and permits real-time tracking of dynamic processes at targeted sites of interest.The adoption of this system in the streamline of pharmacokinetic and pharmacodynamic analysis marks a substantial advancement in comprehending the complex interactions between drugs and biological systems.
(a), and the optical path diagram is depicted in Fig. 1(b).

Figure 2 (
e) shows the overview of the hardware components of the integrated fiber optic-based system for multiscale imaging.

Fig. 1 .
Fig. 1.Schematic diagram of hardware system components.(a)System schematic diagram.(b) Overall optical path diagram of the system.

Fig. 1 .
Fig. 1.Schematic diagram of hardware system components.(a)System schematic diagram.(b) Overall optical path diagram of the system.

Fig. 2 .
Fig. 2. Fiber optic design and physical drawings.(a)Configuration of the multi-fiber bundle probe for the timeresolved fluorescence spectroscopy module.(b) Design of the image transmission fiber bundle for tissue-level fluorescent microscopy imaging.(c) Configuration and (d) schematic representation of the GRIN Lens for celllevel laser scanning confocal fluorescence microscopy imaging.(e) An overview of the integrated fiber optic-based system for multiscale imaging.Top-down view of the hardware composition of (f) the tissue-level fluorescent microscopy imaging module and (g) the laser scanning confocal fluorescence microscopy imaging module.

Fig. 2 .
Fig. 2. Fiber optic design and physical drawings.(a)Configuration of the multi-fiber bundle probe for the time-resolved fluorescence spectroscopy module.(b) Design of the image transmission fiber bundle for tissue-level fluorescent microscopy imaging.(c) Configuration and (d) schematic representation of the GRIN Lens for cell-level laser scanning confocal fluorescence microscopy imaging.(e) An overview of the integrated fiber optic-based system for multiscale imaging.Top-down view of the hardware composition of (f) the tissue-level fluorescent microscopy imaging module and (g) the laser scanning confocal fluorescence microscopy imaging module.
of three PMTs works with dichroic mirrors and filters to concurrently capture three distinct fluorescence signals.During the in vivo experiments, we are aware of the potential for ambient light to interfere with our imaging data and have taken measures to conduct experiments in a controlled, dark environment to minimize this.Furthermore, a background calibration algorithm was designed to optimize interference.As demonstrated in Fig.3(b), we have designed a temporal sequence to manage each function of the integrated system.The two laser sources at wavelengths of 488 nm and 650 nm are connected to a 220 V power supply, maintaining them in a default "on" state.The corresponding laser shutters, controlled separately by their proprietary controllers, are managed by the master computer system, which governs their open or closed status.The "open" state is indicated by rectangles in Fig.3(b).The fluorescence spectrometer interfaces with the master computer via USB, which orchestrates its initiation and sampling duration.During each sampling period, the EMCCD dictates the acquisition period, determining both laser shutter activation duration and temporal window for filter wheel transitions.The GM and PMTs interface with the master computer through the data acquisition card, with the GM initiating the scans.This synchronization allows the PMTs to concurrently capture fluorescence intensity data from the targeted cells at the GM's position.Following the completion of a sampling cycle, the system idles until the onset of the subsequent cycle, as determined by the pre-set sampling frequency.

3. Experimental section 3 . 1 .
Animal surgery and in vivo monitoing of pharmacokinetics in blood ICR mice weighing 20 g-25 g, Bale/c nude mice weighing 18 g-22 g, and SD rats weighing 180 g-200 g were purchased from Nanjing Qin Longshan Laboratory Animal Technology Company.The animals were provided with a standard laboratory diet and maintained under controlled environmental conditions, with a 12-h light/dark cycle.All animal studies were conducted in compliance with the Animal Management Rules of the Ministry of Health of the People's Republic of China (document No. 55, 2001) and the guidelines for the Care and Use of Laboratory Animals of the China Pharmaceutical University.

Fig. 4 .
Fig. 4. CdTe emits light spectra and fluorescence intensity metabolic curves.(a) Emit light spectrum.(b) spectrogram of Cdte quantum dots detected by the system.(c) Fluorescence intensity metabolic curve.

Fig. 4 .
Fig. 4. CdTe emits light spectra and fluorescence intensity metabolic curves.(a) Emit light spectrum.(b) spectrogram of Cdte quantum dots detected by the system.(c) Fluorescence intensity metabolic curve.

Fig. 5 (
b) illustrates the resultant micrograph, showing a homogeneous intensity distribution across the end-face of the fiber bundle.The white arrow highlights the minimum resolvable pattern by the imaging system.By referring to the corresponding resolution chart, this particular element equates to a resolutive capacity of 50.8-line pairs per millimeter (lp/mm).

Figure 5 .
Figure 5. System timing diagram.(a) 1951 USAF Resolution Test Target Physical Image.(b) Resolution target imaging results of tissue imaging system.(c) Resultant image of the fluorescent beads using the confocal fluorescence microscopic imaging.(d) Enlarged view showing the fiber bundle configuration.(e) The corresponding grayscale intensity fluctuation profile along the yellow line.

Fig. 5 .
Fig. 5. System timing diagram.(a) 1951 USAF Resolution Test Target Physical Image.(b) Resolution target imaging results of tissue imaging system.(c) Resultant image of the fluorescent beads using the confocal fluorescence microscopic imaging.(d) Enlarged view showing the fiber bundle configuration.(e) The corresponding grayscale intensity fluctuation profile along the yellow line.

4. 4 .
In vivo fluorescence microscopy for tissue imagingLiver and kidney are two principal organs involved in drug accumulation and metabolism, which are central to the investigation of pharmacodynamics.We verified the in vivo imaging efficacy of the fiber-optic fluorescence microscopy module in hepatic and renal tissues by utilizing CdTe QDs and R3 as fluorescent probes.These fluorescent dyes can be chemically modified or conjugated to drugs for tracing the drug distribution.For improved visualization, we assigned artificial pseudo-colors to grayscale images, based on the respective wavelengths of fluorescence signals.As shown in Fig.6, Fig.6(a) and Fig.6(b) present hepatic tissues stained with CdTe QDs and R3, respectively.Representative H&E-stained tissue section is shown in Fig.6(c).By comparing with Fig.6(a) and Fig.6(c), we can discern the localization of CdTe QDs within the hepatic lobular cell regions, with notable dark zones in the central venous areas.As R3 probe yields distinct high-contrast fluorescence signals, Fig.6(b) also illustrates typical hepatic architecture, marked by hexagonal liver lobules and prominently defined liver plates.Observing Fig.6(b), the liver is composed of prismatic liver lobules, and we can see distinct liver plates represented by very substantial liver tissue.Some dark areas in the figure are blood in the portal area, and R3 dye has a very clear imaging of liver tissue.

Fig. 6 .
Fig. 6.Microimaging.(a) Liver tissue image stained with CdTe Dots.(b) R3 stained liver tissue image.(c) Representative image of H&E-stained liver tissue sections.(d) in vivo and (e) in cryo-sections using the fiber-optic confocal fluorescence microscopy.(f) H&E-stained sections of liver tissue post CdTe QDs injection.Representative image showing R3 distribution in liver cells captured.(g) in vivo and (h) in cryo-sections using the fiber-optic confocal fluorescence microscopy.(i) H&E-stained sections of liver tissue post R3 injection.Representative image showing R3 distribution in renal cells captured (j) in vivo and (k) in cryo-sections using the fiber-optic confocal fluorescence microscopy.(l) H&E-stained sections of renal tissue post R3 injection.

Fig. 6 .
Fig. 6.Microimaging.(a) Liver tissue image stained with CdTe Dots.(b) R3 stained liver tissue image.(c) Representative image of H&E-stained liver tissue sections.(d) in vivo and (e) in cryo-sections using the fiber-optic confocal fluorescence microscopy.(f) H&E-stained sections of liver tissue post CdTe QDs injection.Representative image showing R3 distribution in liver cells captured.(g) in vivo and (h) in cryo-sections using the fiber-optic confocal fluorescence microscopy.(i) H&E-stained sections of liver tissue post R3 injection.Representative image showing R3 distribution in renal cells captured (j) in vivo and (k) in cryo-sections using the fiber-optic confocal fluorescence microscopy.(l) H&E-stained sections of renal tissue post R3 injection.