IVIS Spectrum - Technology overview ================================== In vivo optical imaging is a versatile, non-invasive, non-destructive modality that enables the **visualization and quantification** **of biological processes in living organisms** through the **detection of emitted light** generated from within the biological system. This approach offers unique advantages for small-animal imaging, including high sensitivity, real-time monitoring, and longitudinal assessment under physiological conditions. In the context of preclinical biomedical research, in vivo optical imaging systems have become widely adopted for whole-body imaging in small animal models such as mice and rats. Operating in the **visible to far-red spectral range (450-900nm)**, these systems provide an effective balance between spatial resolution, tissue penetration, high sensitivity, and minimal physiological disruption. Their ability to conduct non-invasive, high-throughput, and longitudinal imaging makes them indispensable tools for **monitoring** **molecular and physiological events** over time. Two major emission-based imaging modalities dominate this field: **fluorescence imaging** and **bioluminescence imaging**. Important considerations ------------------------ .. image:: ../_static/light_interaction.png :alt: *Visible light interaction with biological matter* :width: 1000px :align: center *Visible light interaction with biological matter* .. raw:: html Non-invasive in vivo optical imaging provides a powerful platform for monitoring biological processes in live animals. However, it faces several **interrelated challenges**, primarily arising from the **interactions between visible-range photons** and biological tissues, which can affect imaging performance, spatial accuracy, and data interpretability: - **Limited tissue penetration** Both excitation and emission photons are subject to **significant absorption and scattering** by biological tissues. This dramatically restricts **photon penetration depth**, particularly for visible-wavelength fluorophores or reporters, thereby limiting the modality’s utility for imaging deep anatomical structures. To mitigate this limitation, far-red fluorescent reporters with high brightness should be prioritized, as far-red signals (650–900 nm) penetrate more deeply into tissues (brain, lung, and liver). .. image:: ../_static/optical_window.png :alt: *The optical window for in vivo optical imaging* :width: 1000px :align: center *The optical window for in vivo optical imaging* .. raw:: html - **Limited spatial resolution** Photon scattering in biological tissues limits the **spatial resolution** of in vivo optical imaging by **reducing localization** **accuracy**, particularly for deep-seated emission sources. The deeper the signal origin, the more pronounced the scattering, resulting in **increased signal blurring** and **decreased anatomical precision**. Bioluminescence imaging avoids autofluorescence but suffers from low photon output and tissue scattering, which contribute to low spatial resolution and reduced signal localization accuracy. - **Autofluorescence and background noise** Endogenous tissue autofluorescence, primarily from skin, fur, and diet, **reduces signal-to-noise ratios** in fluorescence imaging. - **Quantitative limitations** Signal intensity is influenced by reporter brightness, depth, biological variability, and animal positioning, making fluorescence and bioluminescence **data semi-quantitative** unless normalized with appropriate controls. - **Substrate and probe delivery issues** Bioluminescence imaging requires **substrate administration** (D-luciferin), and signal intensity is influenced by delivery efficiency, biodistribution, metabolism, and perfusion. Moreover, because luciferase activity is dependent on intracellular ATP and oxygen, the signal is sensitive to hypoxic or necrotic environments, potentially leading to variability across tissues or in disease states. Similarly, fluorescent probes may suffer from non-specific distribution, poor bioavailability, or off-target accumulation, all of which can compromise signal specificity. Optical signal propagation is further influenced by tissue pigmentation, vascularization, and hydration. - **Restricted multiplexing (bioluminescence)** While fluorescence imaging supports multiplexing through spectral unmixing, enabled by the wide availability of spectrally distinct reporters, bioluminescence imaging is limited by overlapping emission spectra and shared substrate requirements, which constrain the ability to perform simultaneous multi-reporter imaging. .. _fluorescence-imaging: Fluorescence imaging -------------------- Fluorescence imaging relies on the use of **fluorophores** (synthetic dyes, nanoparticles, or genetically encoded fluorescent proteins) which are **molecules capable of emitting light upon the absorption of photons** at specific excitation wavelengths. These fluorophores absorb light at a defined excitation wavelength and subsequently emit light at a longer wavelength due to energy dissipation. .. image:: ../_static/fluorescence.png :alt: *Principle of fluorescence* :width: 1000px :align: center *Principle of fluorescence* .. raw:: html In in vivo imaging systems, excitation light is typically delivered using spectrally filtered light-emitting diodes (LEDs), (AMI HT system), or via broadband white-light sources combined with excitation filters (IVIS Spectrum). These illumination strategies enable selective excitation of fluorophores at their optimal wavelengths, enhancing specificity and minimizing off-target activation. The resulting fluorescent photons are captured by highly sensitive CCD cameras after passing through wavelength-specific emission filters, which selectively isolate the desired signal. This approach, employing narrow-band excitation and emission wavelengths, effectively minimizes background noise originating from tissue autofluorescence and reduces spectral overlap between multiple fluorophores. .. image:: ../_static/epi-illumination.png :alt: *2D in vivo fluorescence imaging: epi-illumination* :width: 1000px :align: center *2D in vivo fluorescence imaging: epi-illumination* .. raw:: html This modality enables the targeted visualization of specific biological structures or molecular events through the use of fluorophores conjugated to targeting moieties such as antibodies, peptides, or receptor ligands. The use of **far-red fluorophores**, typically within the 650–900 nm spectral range, enhances imaging performance by increasing tissue penetration and reducing signal interference from endogenous tissue autofluorescence. Traditionally, in vivo fluorescence imaging employs a **surface-based illumination** strategy, known as an **epi-illumination** **configuration**, in which both excitation and emission light paths originate from the same side of the animal (typically from the top). While this setup is effective for detecting superficial signals, it offers limited sensitivity to fluorophores located deeper within tissue. In contrast, some in vivo optical imaging systems implement a **transillumination-based configuration**, where excitation light is delivered from the side opposite the detector. As the excitation light propagates upward through the tissue, it is absorbed by fluorophores that emit photons detectable at the surface. This alternative approach improves sensitivity to deeper tissue signals and enhances contrast and detection accuracy in 2D fluorescence imaging applications. Advanced techniques for in vivo fluorescence imaging ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ Spectral unmixing """"""""""""""""" Spectral unmixing is an advanced analytical technique used in fluorescence imaging to **separate overlapping fluorescence signals** in vivo. It enables the discrimination of signals originating from multiple fluorophores within the same imaging field, or the isolation of specific reporter signals from endogenous tissue autofluorescence. This capability is particularly critical in **multiplex imaging**, where several fluorescent reporters with distinct (but often partially overlapping) spectral profiles are used simultaneously to monitor different biological processes. However, spectral unmixing is also highly beneficial in single-reporter (singleplex) studies, where it can **significantly improve the signal-to-noise ratio** (SNR) by reducing background fluorescence. It is important to recognize that the fluorescent signal detected by the camera is a **composite of both specific and non-specific fluorescence**. It includes photons emitted by the reporter of interest, as well as contributions from endogenous fluorophores (collagen, NADH, and elastin...), which contribute to tissue autofluorescence. Since the detector integrates all photons indiscriminately, **multi-spectral acquisition** is required to capture a broader spectral signature that enables the separation of these signals. Furthermore, both autofluorescence and exogenous fluorescent reporters exhibit **distinctive, wavelength-dependent excitation and emission spectra**. Although these spectra are characteristic of each fluorophore, they often partially overlap, necessitating the use of **multiple excitation and/or emission filters** to sample distinct spectral bands. This process forms the basis for accurate **spectral deconvolution**. In this workflow, the system acquires a **multi-spectral image dataset**, where each image corresponds to a defined excitation–emission wavelength combination. This generates a **spectral profile** for each pixel, reflecting the combined contributions of all fluorescent and background sources present in the tissue. To resolve these mixed signals, spectral unmixing algorithms are applied. These algorithms mathematically deconvolve the composite fluorescence spectrum using a set of reference spectra, typically derived from **reporter-negative control** animals (for background subtraction) and **single-reporter controls** or well-characterized standards (for signal identification). By fitting the measured spectra to these reference fingerprints, the algorithm estimates the **relative contribution of each fluorophore** at every pixel or region of interest. The result is a series of unmixed fluorescence images, each depicting the spatial distribution of a specific fluorophore, even in regions where spectral overlap occurs. This process enables high-specificity localization and semi-quantitative measurement of fluorescent signals in complex biological tissues, supporting robust in vivo imaging across both multiplex and single-target applications. Fluorescence Imaging Tomography (FLIT) """""""""""""""""""""""""""""""""""""" .. image:: ../_static/trans-illumination.png :alt: *3D in vivo Fluorescence Imaging Tomography: trans-illumination* :width: 1000px :align: center *3D in vivo Fluorescence Imaging Tomography: trans-illumination* .. raw:: html Three-dimensional (3D) fluorescence imaging tomography (FLIT) is a mesoscopic in vivo imaging technique that enables the **volumetric localization and semi-quantitative assessment of fluorescent probes** within live small animal models. This method is based on **transillumination scanning**, where excitation light is delivered from multiple positions beneath the subject, while fluorescence emission is detected from above using a sensitive CCD camera. During acquisition, a **series of 2D excitation and fluorescence images **are captured from **multiple illumination positions** underneath the imaging platform. While the excitation source is systematically repositioned, the detection geometry remains fixed. These **multiple excitation–emission pairings** allow the system to sample the optical properties across the full volume of interest. Prior to data collection, a **structured light scan** is performed to acquire the animal’s **surface topography**, which provides a 3D anatomical reference for reconstruction. To estimate fluorophore distribution, FLIT integrates the **transmitted excitation light images**, the **fluorescence emission images** together with the **3D topographic data**. These inputs are processed using a **diffusion-based photon propagation model**, which simulates light transport through scattering and absorbing tissues. The algorithm estimates both local excitation fluence (the photon density per unit area at each location) and emission propagation to the detector. The **inverse problem** is then solved to reconstruct the **3D spatial origin of the fluorescent signal**, resulting in a **semi-quantitative fluorescence** **map of fluorophore concentration and depth**. Importantly, FLIT assumes the tissue is homogeneous, treating the imaging volume as a uniformly scattering and absorbing medium. This simplification enables computationally efficient reconstructions but does not account for tissue heterogeneity (such as differences between organs like lung, liver, and muscle). As a result, while FLIT provides meaningful volumetric imaging, it lacks the anatomical precision of more advanced modeling approaches used in dedicated fluorescence molecular tomography (FMT) systems, which incorporate point-source laser scanning and finite-element or Monte Carlo light modeling for greater spatial accuracy and quantitative reliability. Despite these limitations, FLIT offers a robust, high-throughput, and non-invasive solution for 3D in vivo fluorescence imaging, making it particularly well-suited for longitudinal studies of tumor progression, biodistribution, and molecular tracking in preclinical models. Fluorescence tomography with transillumination scanning significantly enhances sensitivity to deep-tissue signals and enables non-invasive, quantitative imaging of biological processes over time. Despite its limited throughput and longer acquisition time, this method is particularly well-suited for longitudinal studies involving orthotopic tumor progression in deep tissues, drug biodistribution, and cellular or molecular tracking in preclinical research settings. Advantages of fluorescence imaging ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ - **Multiplexing capability** Fluorescence imaging supports the simultaneous detection of multiple fluorophores through **spectral separation and unmixing** **algorithms**. This enables the tracking of several molecular targets or cell populations within the same animal. - **Great flexibility in probe design and availability** A wide variety of fluorescent probes is available, differing in excitation/emission wavelengths and reporter brightness. This allows for tailored selection based on imaging depth, tissue penetration, and target specificity. An excellent resource for spectral properties and compatibility is FPbase, https://www.fpbase.org/ an open-access, community-curated database. - **Strong signal intensity and high throughput for longitudinal imaging** Fluorescent probes generate strong and stable signals under optimized excitation, allowing for faster image acquisition compared to bioluminescence. This supports higher throughput and makes fluorescence imaging well-suited for longitudinal studies. - **High spatial resolution** Fluorescence imaging generally provides higher spatial resolution in 2D planar imaging, particularly for superficial signals, due to focused excitation and stronger photon yield. - **Short imaging time and streamlined workflow** Fluorescence imaging allows for rapid acquisition without the need for substrate injection or distribution delays. This enables short setup and imaging times, facilitates high-throughput studies, and reduces variability across imaging sessions, enhancing reproducibility. - **Compatibility with multimodal and cross-scale imaging** Fluorescent reporters can be used across multiple imaging platforms, including intravital microscopy and postmortem techniques such as IHC, confocal microscopy and light-sheet microscopy, providing continuity from mesoscopic to microscopic resolution. Limitations of fluorescence imaging ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ - **High background signal** Fluorescence imaging is affected by endogenous tissue autofluorescence—particularly from skin, fur, and food, which drastically reduces signal-to-noise ratio, especially in the visible spectral range. - **Limited depth penetration** Excitation and emission light in fluorescence imaging are both subject to strong scattering and absorption, limiting signal detection from deeper anatomical regions compared to bioluminescence. - **Low specificity** The use of an external excitation light source can result in off-target activation and non-specific fluorescence, as well as increased background due to autofluorescence, reducing the specificity of the detected signal. - **Low sensitivity** Due to background interference and light attenuation, fluorescence imaging is generally less sensitive than bioluminescence for detecting low-abundance targets or weak signals. - **Complex quantification** Fluorescence intensity is influenced by excitation light distribution, tissue optical properties, and probe concentration, making quantitative analysis more challenging. Accurate quantification requires careful normalization, background subtraction, and appropriate controls to isolate specific reporter signals. Bioluminescence imaging ----------------------- Bioluminescence imaging is a highly sensitive, non-invasive imaging modality that relies on the **detection of light produced** **by enzymatic reactions** within living organisms. In this process, genetically encoded enzymes known as **luciferases** catalyze the oxidation of small-molecule substrates (D-luciferin, coelenterazine) in the presence of ATP and oxygen, leading to the emission of visible photons. .. image:: ../_static/bioluminescence.png :alt: *Principle of bioluminescence* :width: 1000px :align: center *Principle of bioluminescence* .. raw:: html In in vivo bioluminescence imaging systems, the substrate (D-luciferin) is systemically administered, typically via intraperitoneal or intravenous injection, prior to image acquisition. Once distributed to target tissues, the substrate is locally oxidized by luciferase-expressing cells, producing visible photons through an enzymatic reaction that requires ATP and oxygen. This light then diffuses through surrounding tissues and is captured by a cooled, high-sensitivity CCD camera positioned above the animal. Because bioluminescence produces inherently low photon output, cooled CCD detectors are essential to reduce electronic noise and enable detection of weak signals. Unlike fluorescence imaging, bioluminescence imaging does not require external excitation or spectral separation; as a result, optical filters are typically not required, and total photon emission can be collected directly across the full spectrum. .. image:: ../_static/2D-bioluminescence.png :alt: *2D in vivo bioluminescence imaging* :width: 1000px :align: center *2D in vivo bioluminescence imaging* .. raw:: html Bioluminescence imaging systems generally operate in a planar 2D acquisition mode, in which signal is integrated over the surface of the animal. The resulting images reflect the spatial distribution and magnitude of reporter gene expression or cell localization. Advanced techniques for in vivo bioluminescence imaging ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ Diffuse Light Imaging Tomography (DLIT) """"""""""""""""""""""""""""""""""""""" .. image:: ../_static/3D-bioluminescence.png :alt: *3D in vivo Diffuse Light Imaging Tomography* :width: 1000px :align: center *3D in vivo Diffuse Light Imaging Tomography* .. raw:: html Diffuse Light Imaging Tomography (DLIT) is a three-dimensional (3D) in vivo imaging technique designed to estimate the **volumetric distribution of bioluminescent sources** in small animal models. DLIT expands upon conventional 2D bioluminescence imaging by integrating **surface topography mapping** and **multi-spectral photon detection**, enabling the **tomographic** **reconstruction of light sources** within scattering tissues. In DLIT, the bioluminescent signal is acquired at **multiple emission wavelengths**, typically using a set of broadband spectral filters. Because the emission spectrum of the bioluminescent reporter is known, multi-spectral acquisition allows for **modeling how light of different wavelengths is scattered and absorbed by tissue**, which provides information critical for **estimating both the depth and location of the signal**. Before acquisition, a structured light scan is used to generate a **3D surface map of the animal**, which serves as the anatomical frame for photon propagation modeling. The reconstruction algorithm applies a **diffusion-based photon propagation model** to simulate how light travels from internal sources through scattering and absorbing tissue. By solving the inverse problem using surface photon data, the system estimates the 3D spatial origin and intensity of the bioluminescent signal. However, because bioluminescence is isotropically emitted (in all directions), the precision of depth localization is inherently lower than in FLIT, where directionally controlled excitation improves reconstruction accuracy. As a result, this approach enables depth localization of luciferase-expressing cells or regions, semi-quantitative three-dimensional mapping of signal intensity, and improved discrimination of overlapping signals in adjacent anatomical compartments. Although DLIT retains the high sensitivity and specificity of planar bioluminescence imaging, it adds valuable spatial context, enhancing biological interpretation—particularly in applications such as orthotopic tumor models, inflammation, and tissue-specific gene expression studies. Advantages of bioluminescence imaging ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ - **Exceptional sensitivity and low background** Bioluminescence imaging generates light through enzyme-catalyzed reactions (luciferase and substrate), producing signal only where the reporter is expressed. Since mammalian tissues do not express endogenous luciferase-type enzymes, there is no intrinsic background signal. Combined with the absence of external excitation light, and therefore no tissue autofluorescence, this results in an exceptionally high signal-to-noise ratio, enabling detection of even low-abundance targets. - **High specificity** Signal is confined to regions where both luciferase expression and substrate presence overlap, providing strong spatial specificity for monitoring gene expression, cell viability, or promoter activity. - **Quantitative potential** Under controlled conditions (consistent substrate delivery and oxygenation), bioluminescence signal intensity can reflect cell number, reporter gene expression, or tumor burden in a highly linear and quantitative manner. - **No autofluorescence or photobleaching** The absence of excitation light eliminates tissue autofluorescence, photobleaching, and phototoxicity, allowing for repeated longitudinal imaging with minimal biological disruption. - **Enhanced deep-tissue sensitivity** Many luciferase systems emit light in the red to near-infrared range, which penetrates tissue more effectively than shorter-wavelength fluorescence. Combined with low background, this allows for better detection of signals from deep anatomical sites. - **Total photon detection** Because bioluminescence does not require excitation, all emitted photons can be collected without spectral filtering, maximizing detection sensitivity and simplifying optical configuration. - **Simplified signal analysis** Bioluminescence imaging produces background-free signals without the need for excitation light, autofluorescence correction, or spectral unmixing. This allows for cleaner data interpretation and more straightforward quantification, especially in single-reporter studies. - **Well suited for systemic imaging** Bioluminescence is ideal for tracking systemically distributed processes, such as metastasis, immune cell trafficking, infection, or gene expression across the entire organism. Limitations of bioluminescence imaging ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ - **Low spatial resolution** Bioluminescent photons are emitted isotropically and often originate from deeper tissues. As they scatter through biological structures, the resulting signal becomes diffuse and blurred, reducing spatial resolution, particularly in 2D imaging. - **Dependence on substrate delivery and biodistribution** Bioluminescence imaging requires systemic administration of a substrate, and signal intensity depends on substrate availability, biodistribution, and tissue uptake. These factors can vary not only between animals but also with pathological conditions (tumors, necrosis, or inflammation) which may alter vascularization, perfusion, and metabolic clearance. This leads to both inter-individual and intra-individual variability, especially as disease states evolve over time. - **Requirement for pilot studies to define imaging conditions** A pilot study is typically required to characterize substrate kinetics, including the time to signal plateau, signal duration, and the optimal imaging window. This ensures consistent quantification across time points and subjects, and helps account for variability in substrate uptake and signal dynamics. - **ATP and oxygen dependence** Luciferase activity requires intracellular ATP and oxygen, making signal strength sensitive to cell viability and metabolic state. These factors can vary significantly across disease states, such as in tumors, ischemic tissues, or inflammatory sites, potentially leading to reduced or inconsistent signal unrelated to reporter expression. - **Longer experimentation and acquisition time** Bioluminescence imaging typically requires longer exposure times (seconds to minutes) due to lower photon yield, increasing the risk of motion artifacts and limiting high-speed imaging. Additionally, there is a delay between substrate injection and signal acquisition, as time is needed for the substrate to circulate, distribute, and reach the target tissue, further extending total imaging duration. - **Limited multiplexing** Bioluminescent reporters have broad and overlapping emission spectra, and many share the same substrate. This makes simultaneous multi-reporter imaging technically challenging and often requires sequential imaging or spectral separation strategies. - **Substrate cost and handling requirements** Substrates such as D-luciferin or coelenterazine are relatively expensive, require careful preparation and storage, and may degrade or vary between batches, adding to logistical complexity and experimental variability. - **Expensive hardware requirements** Due to the extremely low photon output of bioluminescent signals, imaging systems must use high-sensitivity, cooled CCD cameras to detect signal reliably. This increases equipment cost compared to basic fluorescent systems. Comparison of in vivo fluorescence and bioluminescence imaging -------------------------------------------------------------- .. image:: ../_static/comparison-fluorescence-bioluminescence.png :alt: *Comparison of in vivo fluorescence and bioluminescence imaging* :width: 1000px :align: center *Comparison of in vivo fluorescence and bioluminescence imaging* .. raw:: html