Search tips
Search criteria 


Logo of nihpaAbout Author manuscriptsSubmit a manuscriptHHS Public Access; Author Manuscript; Accepted for publication in peer reviewed journal;
J Neurosci Methods. Author manuscript; available in PMC 2007 June 30.
Published in final edited form as:
PMCID: PMC1769312

A feasibility study of optical coherence tomography for guiding deep brain probes


Deep brain simulation (DBS) is effective for the treatment of various diseases including Parkinson's disease and essential tremor. However, anatomical targeting combined with microelectrode mapping of the region requires significant surgical time. Also, the fine-tipped microelectrode imposes a risk of hemorrhage in the event that the trajectory intersects subcortical vessels. To reduce the operation time and the risk of hemorrhage, we propose to use optical coherence tomography (OCT) to guide the insertion of the DBS probe. We conducted in vitro experiments in the rat brain to study the feasibility of this application. The result shows that OCT is able to differentiate structures in the rat brain. White matter tends to have higher peak reflectivity and steeper attenuation rate compared to gray matter. This structural information may help guide DBS probe advance and electrical measurements.

Keywords: Optical coherence domain reflectometry, Optical coherence tomography, Rat brain, Deep brain stimulation, Tissue classification

1. Introduction

Advances in neurosurgical treatments are aimed at providing a better quality life for patients with neurological and movement disorders such as Parkinson's disease and essential tremor. Treatments include ablation (e.g., thalamotomy and pallidotomy), transplantation of tissue or cells, and deep brain stimulation (DBS) (Benabid et al., 1991; Limousin et al., 1998). All these neurosurgical treatments share the need for localization and target verification and the avoidance of the subcortical vasculature. Due to the difficulties and lengthy surgical procedures, neurosurgical treatments are generally considered as the second treatment option following pharmacological therapy.

In the surgical procedure to implant a DBS lead, a physiological mapping process involving repeated penetrations of metal microelectrodes into deep-brain matter is often used to refine anatomical or imaging-based stereotactic targeting techniques. The therapeutic radius of the DBS lead is approximately 2–3 mm, and the targets are typically in close proximity to other structures. Stimulation of these nearby structures can lead to undesirable side effects such as motor activations, paresthesias, and visual phenomena. Thus, a high degree of precision is essential. Currently, microelectrode mapping is considered the gold standard for deep-brain target localization. Targets typically are identified by the neuronal frequency and pattern of resting activity, the response to somatosensory stimuli and the effects of electrical stimulation through the electrode tip.

Such refinement, however, comes at a cost. Physiological mapping with microelectrodes is highly demanding and time-consuming. The patients must be kept alert during what often is a long surgical procedure. The neurosurgical team must be highly knowledgeable and regulate patient responses. With physiologic mapping, it is often necessary to reinsert the microelectrode along several tracks to derive a meaningful map of the subcortical structures and thus optimize targeting. Each track can require up to 2 h of surgical time and increase the risk of complications such as intracerebral hemorrhage caused by the fine-tipped microelectrode encountering a deep-seated vessel.

A possible alternative or complement to current physiological mapping techniques is optical coherence domain reflectometry (OCDR). It will increase speed, safety, and accuracy of DBS surgery. The probe of OCDR can be as thin as the diameter of the optical fiber, typically 0.125 mm; thus, it can be replaced or combined with the microelectrode probe. OCDR is analogous to other range-finding techniques that use radio (RADAR) and sound (SONAR). They emit a wave toward the target object and detect the reflections. In OCDR, the depths of sample reflections are measured indirectly by interference with a reference reflection. The delays of reflected light provide information on the distance of the object from the device. OCDR scans up to 2 mm ahead of the probe and can easily produce real-time capability. OCDR can also detect blood vessels and thus, reduce the potential of hemorrhage.

Huang et al. incorporated transverse scanning and developed OCDR into the cross-sectional imaging modality optical coherence tomography (OCT), which is analogous to B-mode ultrasonography (Huang et al., 1991). OCT is a novel and noninvasive imaging technique providing microscopic sectioning of biological tissues (Rollins et al., 1998). OCT has widely been used in the eye, and the potential of this imaging modality has been investigated in many clinical fields (Fercher et al., 2003) including neurology and neurosurgery, i.e., neural development (Boppart et al., 1996), laser ablation (Boppart et al., 1999), and melanoma detection (Boppart et al., 1998). The purpose of this study was to determine the ability of OCT to differentiate among tissues in an in vitro rat brain preparation.

2. Materials and methods

The study was performed under a research protocol approved by the Animal Research Committee of the Cleveland Clinic Foundation. Two male Sprague–Dawley rats weighing 250–300 g (Harlan, Indianapolis, Indiana) were prepared for imaging. The rats were deeply sedated and decapitated using a commercial guillotine. Three coronal plane samples were derived from the two rat brains for OCT imaging. After dissection, the whole brains were immediately chilled on dry ice and subsequently cut into 4 mm sections using a rat brain matrix. Each sample was transferred to a 5 mm cuvette filled with 0.9% NaCl and imaged over the next 3 h. The experiment was conducted at normal room temperature (23 °C). Transverse scan paths were pre-selected to cross-regions with white and gray matter junctions. To identify the ends of the scanning sites, steel pins (Fig. 1) were inserted just before the coronal sections were placed in the cuvette. The pins facilitated subsequent comparison with the sections after they were stained for microscopy.

Fig. 1
Photographs depicting rat brain OCT scanning sites. Steel pins were inserted in the coronal section to facilitate alignment of OCT images with histological images. Dashed squares in the photographs correspond to OCT images (Figs. (Figs.33 and ...

We used polarization insensitive OCT (Fig. 2) based on the classical Michelson interferometer. It featured a high-power, low-coherence superluminescent diode (SLD) centered at 1294 nm with a 33 nm full-width half-maximum bandwidth. The 1294 nm wavelength light source was selected to optimize penetration of the OCT beam into tissue. The light source was divided into the sample arm and the reference arm by a 50/50 beam splitter. Interference occurred only when path lengths of beams returned from the two arms matched to within the coherence length of the light source. Coherence length was inversely proportional to full-width half-maximum bandwidth, and axial resolution was defined as coherence length divided by refractive index of the tissues (Fercher et al., 2003). The coherence length of our system was 18-μm, and the estimated refractive index for soft tissue was 1.4 (Tearney et al., 1995). The interference signal was modulated at the Doppler frequency caused by the reference-arm movement.

Fig. 2
Schematic drawing of OCT system. SLD, superluminescent diode; PC, polarization controller; BPF, band pass filter; and A/D, analog to digital converter.

A pigtailed linear polarization splitter separated the horizontal and vertical polarization signals. Balancing the horizontal and the vertical direct current bias in the reference arm with the sample arm blocked was essential to the polarization insensitive measurement (Sorin and Heffner, 1993). A polarization controller was used to balance the direct current signal of the two channels in the reference arm. The two polarized OCT signals were detected separately by two independent photo-detectors, band-pass filtered and digitized. The signal in each channel was proportional to the amplitude of the scattered electric field, so computing root-mean-square of the two signals resulted in a polarization-independent reflectivity measurement. Lateral scanning of the mirror in the sample arm generated the two-dimensional OCT image.

The OCT image covered 5.1 mm transversely with 240 evenly spaced axial scans (A-scans). The A-scan range determined from the galvanometer mirror in the reference arm was 3.3 mm. Therefore, the OCT image depth was 2.36 mm based upon the estimated refractive index. Ten frame images were obtained at small offsets perpendicular to the transverse scan direction and were averaged to suppress speckle.

Immediately following the scanning procedure, the sample sections were transferred to a 10% paraformaldehyde solution for 24 h. Subsequently, the sections were embedded in paraffin, cut into 30-μm coronal sections and stained with H&E. The stained sections were photographed at low magnification and matched to corresponding locations in the rat brain atlas (Paxinos and Watson, 1998).

3. Results

In the OCT and corresponding histology images (Figs. (Figs.33 and and4),4), the clearly visible steel pins enabled alignment of the contents. Whereas OCT was scanned perpendicular to the coronal plane of the brain, the brain was sectioned parallel to the coronal plane for histology processing. To make the comparison between OCT and histology, coronal plane histology slices were stacked sequentially. Each slice had an interval of 0.3 mm, and the first starts from approximately 0.3 mm below the surface shown in the OCT images. Note that the top part of the sectioned rat brain is uneven in the OCT images although the brain is cut flat. The unevenness is caused by tissue swelling after immersion in the saline solution.

Fig. 3
OCT and histological images. The OCT image (top) revealed internal structures that were confirmed in the corresponding composite stack of light microscopic images (below). ec, External capsule; opt, optic tract; ic, internal capsule; CA2(3), field CA2(3) ...
Fig. 4
OCT and histological images. These images are similar in organization to those in Fig. 3, but are from a different area. S1BF, primary somatosensory cortex, barrel field and ic, internal capsule.

Having identified tissue types in OCT images, we made quantitative comparisons of attenuation coefficients and measured maximum intensities in homogeneous regions of interests (ROIs) in the OCT images. Each ROI spanned 0.11 mm and the five A-scans in the ROI were averaged. Smoothing was performed in the axial dimension as well to suppress speckle noise further. The smoothing window size was set to 18 μm, which is equal to the coherence length of the OCT system.

The averaged A-scan from each ROI was used to calculate a maximum reflectance intensity and attenuation rate (Fig. 5). The maximum intensity was the log-scaled maximum signal intensity in the averaged A-scan. The attenuation coefficient was the rate of signal attenuation with depth. It was measured by the least-square linear regression of the log signal intensity with depth in the linear portion of the A-scan. According to Beer's law, light attenuates with tissue depths in a log-linear fashion due to loss from absorption and scattering. The shallower segment of the OCDR A-scan, where signal from single scattering events predominates, obeyed Beer's law. At longer delays, light scattered from the deeper levels was mixed with multiple-scattered light from shallower levels and the apparent attenuation slowed (Fig. 5). The onset of multiple scattering effectively limited the depth of OCT analysis of tissue optical properties. Therefore, we limited the fitting of the attenuation rate to the shallower single-scattering segment as determined by visual inspection of the averaged A-scan in each ROI.

Fig. 5
Averaged ROIs or averaged A-scans or gray (hippocampus) and white (external capsule). In-averaged A-scans of ROIs in gray (upper) and white (lower) matter, the slope of the robust least-square fit (straight line) was proportional to the attenuation coefficient ...

Nineteen different ROIs were analyzed using three OCT images of the three brain samples. They included five hippocampus regions, four primary somatosensory cortex barrel field (S1BF) regions, two external capsule regions, five internal capsule regions, and three optic tract areas. Attenuation coefficients and maximum intensities of the averaged A-scans were calculated for each ROI (Fig. 6, Table 1). The internal and external capsules showed very similar optical properties and were clustered together as the capsules group. The optical properties of gray matter, i.e., hippocampus and S1BF, were distinctively different from those of white matter, i.e., capsules and optic tract. Within the white matter, the capsules and optic tract were clearly different from one another in optical properties as well.

Fig. 6
Clustering of different tissue types. White (circle) and gray (square) matter formed clusters when attenuation coefficients were plotted against maximum log intensities. Even within the white matter, capsules (open circles) and optic tracts (closed circles) ...
Table 1
Tissue classification results based on attenuation coefficient and maximum log intensity

4. Discussion

Unlike the standard OCT setups (Huang et al., 1991; Fercher et al., 2003), our OCT system adapted polarization diversity detection for polarization-insensitive measurement. Brain white matter contains tracts of parallel nerve fibers that are highly birefringent (Ducros et al., 2001). Birefringent tissue changes the polarization of the transmitted light. In a standard OCT setup that detects only one polarization channel, highly birefringent tissue typically appears as alternating bright and dark layers. The tissue appears dark when the polarization is mismatched and bright when it is matched. Our OCT system detects both polarization channels and adds the optical intensity in the two channels. This allows the system to detect the total reflected optical signal and measure signal attenuation characteristics of different tissues without interference from polarization shifts.

Our OCT images were highly correlated with the histological images, and different tissue optical characteristics were evident. White matter is brighter at the surface than gray matter because it has a higher degree of scattering. The increased scattering in white matter is presumably due to myelination. Scattering causes attenuation of signal with depth. As expected, signal attenuation is also more rapid in white matters. Taken together (Fig. 6), the signal strength and attenuation characteristic allows perfect differentiation of the white and gray matter tissue we studied.

Specific gray and white matter regions in the brain also exhibited differences in optical characteristics. The white matter of the optic tract showed a higher reflectance and scattering than the white matter in the capsule regions. The difference is probably caused by the degree of myelination (Hamano et al., 1998). We can distinguish the S1BF from the hippocampus by the mean value of attenuation coefficient. However, the standard deviation of attenuation coefficient in the hippocampus is relatively large due to one outlier in the hippocampus data cluster. This might be due to the inclusion of the field CA2(3) of hippocampus, which is clear in the histological sections. The field CA2(3) has a denser cellular structure (stratum pyramidale) than the rest of hippocampus; thus, it should have different optical properties than the remainder of the hippocampus.

The range of tissue distinction using OCT is limited by the “single scattering” regime. Light reflected from deeper layers is overwhelmed by multiple scattered light from more superficial layers that is detected by OCT as having the same delay. Only single scattered light carries information on the local tissue property. The depth at which multiple scattering becomes dominant can be seen on the OCDR A-scan where attenuation slows down and deviates from log linearity. This depth defines the penetration range of our OCT system. The average penetration range in gray matter is 0.77 mm and that of white matter is 0.31 mm. The penetration range from nine different ROIs in gray matter is from 1.2 to 0.4 mm, and the range in white matter is varied from 0.5 to 0.2 mm.

The detection of different cellular layers in the same tissue type indicates potential for precise positioning of the DBS probe during surgery. More consistent and accurate measurement can be achieved by increasing the spatial averaging sample number, which is also an effective way to suppress speckles noise (Schmitt et al., 1994). One potential application of this technology would be to replace or complement current microelectrode recording techniques. The size of the probe should be the same as or smaller than currently used microelectrodes, which are typically 0.5–1.0 mm in the diameter. OCDR is best suited for DBS lead placement since OCT systems require scanning hardware, which is difficult to build within a 1-mm space. With OCDR there is no transverse scanning, and the information on tissue texture (transverse variation) that is apparent on OCT images is lost. OCDR does capture signal strength and variation in depth. Our in vitro OCT data suggest that this depth variation alone is sufficient for distinguishing tissue types. In the next step, we will measure brain tissue in vivo using an OCDR probe that is integrated into a brain probe. We aim to find similar contrast between brain tissue types even though we expect the absolute signal amplitude and attenuation rate to differ due to differences between in vivo and in vitro tissue characteristics. In addition, the optical setup will be different.

The use of optical properties to differentiate among tissues during DBS has been explored by others. Johns et al. (1998) and Qian et al. (2003) utilized diffuse reflectance spectroscopy to differentiate gray and white matter and to assist during neurosurgery. They found that gray matter, white matter, and crebrospinal fluid within the brain scatter light quite distinctively. Optical spatial profiles and the actual anatomical profiles based on the postoperative MRI images showed good correlation from their previous study (Giller et al., 2000).

The advantage of OCDR/OCT over spectroscopy is the ability to resolve signals over a range of depths. In spectroscopy and other summed reflectivity measurements, the measured signal comes predominantly from tissue immediately ahead of the probe. OCDR/OCT adds range to the measurement by resolving the tissue optical property as a function of depth up to the limit where multiple scattering predominates. Within this range, it is possible to detect tissue type and tissue boundaries with OCDR/OCT. Thus, OCDR/OCT provides the ability to look ahead, measure the distance to the next tissue boundary, and determine the type of tissue in the next deeper layer.

OCDR brain probe will also be able to detect blood vessels. Flow induces Doppler shift in OCT and OCDR signals. Doppler OCT has been successfully used to visualize blood vessels within various tissue types and measure blood flow dynamics (Chen et al., 1998; Rollins et al., 2002). With the addition of Doppler capability to the OCDR brain probe, large blood vessels ahead of the probe can be visualized and avoided, thus reducing the risk of hemorrhages during probe insertion.

In summary, we showed that the tissue classification essential for neurosurgical treatments is possible using optical properties. OCT images were highly correlated with histology images, and quantitative analysis based on signal attenuation and maximum intensity confirmed clear tissue classification among different tissue type. This supports the feasibility for using OCT or OCDR in guiding the insertion of DBS probes.


The work has been supported by NIH 5R21EB002718. The authors express our gratitude to Massud Turbay and Nagi Hatoum for the valuable discussion.


  • Benabid AL, Pollak P, Gervason C, Hoffmann D, Gao DM, Hommel M, et al. Long-term suppression of tremor by chronic stimulation of the ventral intermediate thalamic nucleus. Lancet. 1991;337:403–6. [PubMed]
  • Boppart SA, Bouma BE, Brezinski ME, Tearney GJ, Fujimoto JG. Imaging developing neural morphology using optical coherence tomography. J Neurosci Methods. 1996;70:65–72. [PubMed]
  • Boppart SA, Brezinski ME, Pitris C, Fujimoto JG. Optical coherence tomography for neurosurgical imaging of human intracortical melanoma. Neurosurgery. 1998;43:834–41. [PubMed]
  • Boppart SA, Herrmann J, Pitris C, Stamper DL, Brezinski ME, Fujimoto JG. High-resolution optical coherence tomography-guided laser ablation of surgical tissue. J Surg Res. 1999;82:275–84. [PubMed]
  • Chen Z, Milner TE, Wang X, Srinivas S, Nelson JS. Optical doppler tomography: imaging in vivo blood flow dynamics following pharmacological intervention and photodynamic therapy. Photochem Photobiol. 1998;67:56–60. [PubMed]
  • Ducros MG, Marsack MG, Rylander HG, Thomsen SL, Milner TE. Primate retina imaging with polarization-sensitive optical coherence tomography. J Opt Soc Am A Opt Image Sci Vis. 2001;18:2945–56. [PubMed]
  • Fercher AF, Drexler W, Hitzenberger CK, Lasser T. Optical coherence tomography-principles and applications. Rep Prog Phys. 2003;66:239–303.
  • Giller CA, Johns M, Liu H. Use of an intracranial near-infrared probe for localization during stereotatic surgery for movement disorders. J Neurosurg. 2000;93:498–505. [PubMed]
  • Hamano K, Takeya T, Iwasaki N, Nakayama J, Ohto T, Okada Y. A quantitative study of the progress of myelination in the rat central nervous system, using immunohistochemical method for proteolipid protein. Brain Res Dev Brain Res. 1998;108:287–93. [PubMed]
  • Huang D, Swanson EA, Lin CP, Schuman JS, Stinson WG, Chang W, et al. Optical coherence tomography. Science. 1991;254:1178–81. [PubMed]
  • Johns M, Giller CA, Liu H. Computational and in vivo investigation of optical reflectance from human brain to assist neurosurgery. J Biomed Opt. 1998;3:437–45. [PubMed]
  • Limousin P, Krack P, Pollak P, Benazzouz A, Ardouin C, Hoffmann D, et al. Electrical stimulation of the subthalamic nucleus in advanced Parkinson's disease. N Engl J Med. 1998;339:1105–11. [PubMed]
  • Paxinos G, Watson C. The rat brain in stereotaxic coordinates. Academic Press; San Diego: 1998.
  • Qian Z, Victor SS, Gu Y, Giller CA, Liu H. Look-ahead distance of a fiber probe used to assist neurosurgery: Phantom and Monte Carlo study. Opt Expr. 2003;11:1844–55. [PubMed]
  • Rollins AM, Kulkarni MD, Yazdanfar S, Ung-arunyawee R, Izatt JA. In vivo video rate optical coherence tomography. Opt Expr. 1998;3:219–29. [PubMed]
  • Rollins AM, Yazdanfar S, Barton JK, Izatt JA. Real-time in vivo color Doppler optical coherence tomography. J Biomed Opt. 2002;7:123–9. [PubMed]
  • Schmitt JM, Knuttel A, Yadlowsky M, Eckhaus MA. Optical-coherence tomography of a dense tissue: statistics of attenuation and backscattering. Phys Med Biol. 1994;39:1705–20. [PubMed]
  • Sorin WV, Heffner BL. Polarization independent optical coherence domain reflectometry. Hewlett-Packard Company; 1993.
  • Tearney GJ, Brezinski ME, Southern JF, Bouma BE, Hee MR, Fujimoto JG. Determination of the refractive index of highly scattering human tissue by optical coherence tomography. Opt Lett. 1995;20:2258–60. [PubMed]