Optical fiber was once relegated to simple light delivery and collection in medical instruments, often for exploratory procedures. Today, this fiber winds its way into all manner of instrumentation in the laboratory and clinic, capturing health data from complex — and often hard-to-reach — biological systems. The data then reveals information ranging from temperature and blood flow to cellular signaling for diagnostics. This integration into the medical workflow has fueled innovation and discovery in neuroscience. Fiber optic technology is the key to many types of modern medical diagnostics, including neuroimaging. Courtesy of iStock.com/kynny. The structural and chemical composition of brain tissue, as well as its optical properties, make study beyond the surface cortex — the outer layer of neural tissue of the cerebrum — challenging for most imaging modalities. But optical fibers can deliver light deep within brain tissue. This capability has proved to be useful in a range of procedures that measure brain signaling in both stationary animals (such as with patch cords) and freely moving animals (such as with implants). Two of the most common deployments of fiber optics in neuroscience are optogenetics and fiber photometry. Optogenetics is the use of a light source, emitted from an optical fiber, to manipulate light-sensitive proteins called opsins that are attached to specific neural cells. This allows control of precisely defined brain activity. Fiber photometry is the use of light sent through an optical fiber to excite fluorescent biosensors, with the resulting fluorescence tracked via a detector to monitor dynamic brain activity. This technique helps to establish patterns related to brain function and behavior. It’s all material The most commonly used material by far in the creation of medical optical fibers is silica glass. Companies including Armadillo SIA specialize in sourcing silica-core fibers and engineering assemblies such as bundles, tapers, and fused ends. These silica fibers offer several advantages over other potential materials, including wide wavelength transmission — especially useful when exploring optimal excitation and emission in the ultraviolet and near-infrared ranges of the spectrum — and chemical inertness (lack of reactivity with other chemicals).. Despite its ubiquity in current designs, silica’s material properties can present a challenge when observing brain activity in time-delayed research. The stiffness of the material can cause inflammation after prolonged contact with tissue. “Silica is stiffer compared to polymers,” said Mario Paredes, business development manager at Armadillo. “Typically, we see silica fibers used in shorter-term studies or combined with coatings and assembly designs that minimize tissue response. We support researchers with geometries that help reduce this limitation.” According to Paredes, the vast majority of optical fiber used in optogenetics (Figure 1) and fiber photometry (Figure 2) is multimode fiber, which offers a larger core diameter that supports multiple light paths over short distances. Single-mode fiber theoretically allows flexibility in probe design and has a smaller core to support a single light path over greater distances (useful for communications). But scientists have found it has limited ability to excite and detect fluorescence. Figure 1. A diagram outlining the implementation of optogenetics, which includes (from left) the use of light-activated opsins, connected with specific neurons, targeted in the brain to produce a result. Courtesy of Doric Lenses. Figure 2. The methodology of fiber photometry includes (from left) identifying a biosensor, detecting activity within the brain, collecting the signal, and recording it over time. Courtesy of Doric Lenses. “While single-mode fibers have niche uses in delivery or interference-based techniques, multimode fibers dominate in photometry and optogenetics because of their ability to both deliver and collect light efficiently,” he said. Paredes added that researchers sometimes request his company to customize the fiber in terms of tapering and bundled assemblies, which allow spatial control of light delivery and collection — often essential in neurological studies. Fiber guides brain research Sead Doric, president of Doric Lenses, said that optical fiber becomes invaluable when studying regions of interest deep within the brain. “One of the primary uses of optic fibers — especially fiber optic cannulas — in neuroscience is to deliver and collect light at these great depths in freely behaving animals, with high spatial and temporal resolution,” he said. “And fiber optic cannulas are implants designed for long-term experiments (hours to weeks) and can be used to address scientific questions such as sleep, memory formation, disease progression, and so on.” The core diameter of fiber used for either optogenetics or fiber photometry typically ranges from 200 to 400 µm. The numerical aperture (NA) used for fiber photometry is high to achieve maximum intensity and a superior signal-to-noise ratio, while in optogenetics, a low NA is used for targeted excitation and a higher NA for stimulating larger brain regions. Other factors include the flexibility of the fiber, which comes into play when animals in an experiment must behave in a somewhat natural way. The equipment used for fiber photometry and designed by Doric Lenses can be used on an individual animal or a group of animals, targeting a single site or multiple sites for excitation, which can then be used to study feeding behavior, sleep cycles, social interactions, or metabolic changes. The core of these systems is the fiber photometry minicube, which is essentially a self-contained unit incorporating components such as the light sources, collimator, dichroic mirrors, filters, and ports. “The first generation was a box containing dichroic mirrors and collimators and all ports with a fiber optic connector,” Doric said. “Light sources and detectors were connected with fibers. In the latest generation, everything is integrated into a single box — light sources, detectors, and electronics.” For researchers in neurology seeking multiple sets of experimental data from brain activity, Doric Lenses offers its Neuroscience Studio and data analysis software, danse. The solution enables end users to oversee multiple components and avenues for analysis. It can control light sources, a bundled imaging system, and a behavior camera, as well as manage signal and behavior analysis. “The ability to synchronize multiple devices — fiber photometry systems, optogenetic sources, behavior measures — and record all the data in the same file is a great advantage,” Doric said. “It also allows the scientist to have a single interface with all their measures, instead of jumping between different software and sometimes even different computers.” Brain biomarker detection In today’s research, several initiatives are combining the capabilities of fiber-based imaging and sensing with an AI component. In one such pursuit, a team in the Department of Chemical Engineering at Imperial College London is devising a system to continuously monitor multiple brain indicators from cerebrospinal fluid (CSF) over time. In a study published last year, the team noted that — particularly in acute medical scenarios — the ability of emergency workers and clinicians to monitor multiple brain biomarkers is limited based on existing technologies used in hospitals. In response, the team built a system that integrated seven silica multimode fibers bundled together, with the distal ends coated in polymers containing fluorescent films within a catheter tip. The fluorescence is excited with a multiwavelength laser matched with a spectrometer via an optical coupler. They validated the setup using lamb brain models and human CSF samples. The design aimed to monitor six CSF biomarkers: temperature, dissolved oxygen, glucose, pH, sodium, and calcium concentrations. Each of these factors fluctuates following traumatic brain injury, stroke, and other emergencies, when conditions such as hypoxia (lack of oxygen in tissue), excitotoxicity (nerve cell damage), and other complications can rise to life-threatening levels1. The size and design of the system could provide crucial data when patients are at their most vulnerable in an intensive care unit, said Ali K. Yetisen, director of the Centre for Biochemical Sensors in the Department of Chemical Engineering at Imperial College London. According to Yetisen, the idea for the study came from conversations the team had with neurosurgeons from within the Imperial College Healthcare National Health Service Trust. “The miniaturization of the fiber optic system and method to coat each fiber with a fluorescent sensor for real-time measurements is what makes this system different,” he said. “There have been some multiplexed optical fiber sensors in the literature. But this is the first time the optical fiber system contained these six specific sensors and was used in the context of traumatic brain injury monitoring.” To expedite analysis, computational algorithms were developed and applied to process the spectral data collected by the spectrometer. Machine learning helped not only to report information from multiple biomarkers simultaneously but also to remove any overlap in the emission spectra that could prevent an accurate reading. The postprocessing software also performed spectrum denoising, baseline correction, and labeling, among other steps1. “The AI system was trained to detect erroneous data, distinguish fluorescence emission peaks, and provide real-time measurements,” Yetisen said. The team plans to test its system in a variety of different research and clinical environments in the future. Real-time behavior observation Researchers at Stanford University focused on the development and use of a fiber optic apparatus to optically track the rapid changes in the transmembrane potential of neurons expressing fluorescent genetically encoded voltage indicators. This technique reveals rapid changes in membrane potential in user-targeted neurons. While scientists have long recognized these voltage indicators as a means to study neural activity, traditional methods — such as EEG (electroencephalogram) — captured only isolated isolated locations of activity. These methods also failed to deliver the sensitivity necessary to capture changes across broader populations of neural cells. The Stanford team used two technologies in its transmembrane electrical measurements performed optically (TEMPO) system to observe mice. Fiber photometry tracked neural oscillations in mice engaged in a series of activities, and imaging via a mesoscope captured neural waves as they moved through regions of the brain while the mice were stationary2. “With the fiber optic, we were taking measurements at the tip, while we used the mesoscope to get images at the micron scale of neural wave oscillations while the animals’ heads were fixed under a microscope,” said Mark Schnitzer, the Anne T. and Robert M. Bass Professor in the School of Humanities and Sciences at Stanford and co-author of the study. Researchers implanted 400 µm-core borosilicate glass doped glass fiberoptics from Doric Lenses, which have minimized autofluorescence, through a cannula into the animals. An amplitude-modulated, solid-state laser light source excited the fluorescent indicators, with a Faraday isolator placed after each light source to prevent back-reflection. An optical diffuser was used to vary the path lengths of the light to prevent mode hopping in the fiber; this step was crucial to eliminate illumination noise in the fiber optic induced by mouse motion. The mesoscope captured a wide field of view across the cortex of restrained mice. A high-NA objective lens and two scientific CMOS (sCMOS) cameras enabled the mesoscope’s operation. The researchers identified two types of high-frequency beta waves, which are indicative of sensory processing and movement, and a lower-frequency theta wave traveling in the reverse direction, which may show how signals affect learning. Further exploration in this area could lead to insights into the specific ways brain waves propagate in patients with neurodegenerative diseases or other conditions. “We looked at one cell type in our experiment, but there are thousands of cell types we could be looking at in the future,” said Simon Haziza, a research scientist in the Biology Department at Stanford and author of the paper. Advancements in design As applications in neuroscience have advanced, so too has the incorporation of fiber optics. Charles Golub, medical marketing manager at Lightera, said that today’s fibers (Figure 3) are integral to guidewires, cardiac ablation catheters, neurovascular microcatheters, biopsy needles, spectroscopy probes, robotic surgical tools, and laparoscopic devices. Figure 3. Optical fiber is now incorporated into many medical procedures, including neurological examination and treatment. Courtesy of Lightera. Endoscopic probes have evolved over the last decade to include CMOS chip-on-tip cameras as well as coherent fiber bundles. Endoscopic procedures may now involve staining of tumors and polyps, so optical fiber can transmit light at the wavelengths required to image these growths. “CMOS-based endoscopes enable greater articulation in the scope; however, this comes with many drawbacks, including lag and poor resolution,” Golub said. “Light-delivery fibers are common; however, the role of fiber has expanded to vascular imaging (OCT), shape sensing, and power delivery (often laser), rather than carrying images.” Lightera produces fibers that support laser ablation, a process in which lasers are delivered through a fiber to target specific areas for treatment — such as in the brain. This capability allows for targeted therapy on cancerous or damaged areas without harming the tissue around it. Golub said, however, that pulsed lasers put strain on optical fiber due to high peak power, nonlinear effects, and thermal load. Therefore, Lightera offers solutions ranging from multicore fibers and mode-stripping coatings to double- and triple-clad fibers, including those in its HCXtreme series of fibers. “HCXtreme was originally developed for urology, where there is a tight corner near the kidney,” Golub said. “In neuroscience research, simpler multimode patch fibers are more common. Fluoropolymer buffer acts as a second cladding, enabling less light to escape from the core. The more light that escapes in a tight bend, the more inaccuracies could occur from the power transmitting through the fiber.” Lightera’s optical fiber also features fiber Bragg gratings at multiple points. These microstructures generate an interference pattern in the fiber core and can be used to monitor factors such as changes to temperature and strain. This capability enhances optical fiber’s ability to provide guidance during surgery. According to Golub, the industry is widening its scope and exploring the potential to expand applications of optical fiber in neurology. These use cases, he said, are just beginning to be realized. “Novel shape-sensing fibers, for example, offer a high level of resolution without the need for angiograms or unnecessary radiation from x-ray imaging traditionally required for catheter placement,” Golub said. “By incorporating shape-sensing fiber technology, practitioners can achieve highly accurate catheter positioning and confidently navigate complex anatomies while reducing radiation exposure.” References 1. Y. Zhang et al. (2024). Fully automated and AI-assisted optical fiber sensing system for multiplexed and continuous brain monitoring. ACS Sens, Vol. 9, No. 12, pp. 6605-6620. 2. S. Haziza et al. (2025). Imaging high-frequency voltage dynamics in multiple neuron classes of behaving mammals. Cell, Vol. 188, No. 16, pp. 4401-4423. The FOTEMP fiber optic probe and signal conditioner system is used for temperature measurements. Courtesy of Micronor Systems. Cost remains a barrier in smaller-scale environments, but industry and R&D sectors have deployed fiber sensors across a wide range of applications and settings. High-voltage and other hazardous applications are particularly well established, including those in aerospace, power and energy, the life sciences, MRI, microwaves, battery testing, and fuel distribution. The components in these sensing systems include the sensor itself, glass wire, an interrogator or controller, and a light source. One of the most basic uses of these sensors is the monitoring of temperature and strain, said Dennis Horwitz, president and technical sales manager at Micronor Sensors. Micronor offers two types of sensors in this category: the FOTEMP series, which uses a 200-µm multimode fiber to illuminate a gallium arsenide crystal, and a fiber Bragg grating (FBG)-based 850-nm single-mode fiber with microstructures inscribed into the fiber core by an ultraviolet laser. “These fibers fit in different size probes, depending on the size of the sample you’re measuring and the degree of rigidity you need,” Horwitz said. “In an FBG, you could be testing at multiple points, not just at the tip of the probe. If you’re testing an electric vehicle battery, for example, you can monitor battery pack bulging and temperature at up to 30 points in a single sensor.” Recently, a fiber optic thermometer called FOTEMP-PLUS from Micronor was used to determine the temperature at which brain tissue could be frozen in 2-methylbutane while still retaining its RNA integrity1. This process is vital for MRI-based histopathological study, but an adequate temperature threshold for preservation has not been firmly established. Additionally, significant developments have occurred in recent years in radiofrequency ablation (RFA) and in MRI-guided robots. RFAs generate a localized high-temperature field to precisely remove a cancerous tumor while nearby healthy tissue remains undamaged. Proper temperature and strain are maintained when guided by the data from FBG sensors. MRI-guided biopsy and surgical robots also benefit from the magnetic immunity of fiber optic sensors. Special MRI-compatible rotary and linear encoders are used for position feedback and surgical guidance. Reference 1. G. Nair et al. (2024). A method to image brain tissue frozen at autopsy. NeuroImage, Vol. 296, No. 120680.