Robust High-Resolution Fine OCT Needle for Side-Viewing Interstitial Tissue Imaging

Abstract

Fine optical coherence tomography (OCT) imaging needles that can be integrated with a standard biopsy needle have been developed with a new optics design to improve the optical quality and mechanical robustness, where a fiber-optic lens (that is spliced to a single-mode fiber) and a microreflector are encased within a microglass tube. The design also minimizes the cylindrical lens effect induced by the glass tube and eases the needle assembly process. Real-time cross-sectional OCT imaging of various tissue samples were performed ex vivo using the miniature-imaging needle along with a 1300-nm swept-source OCT system. The preliminary results demonstrate the improved mechanical and optical performance and suggest the potential of the fine OCT needle for minimally invasive interstitial imaging and image-guided biopsy.

I. Introduction

Cancer is the second leading cause of death in the United States [1]. The current standard procedure for cancer diagnosis is biopsy followed by histopathological examination [2]. With the advances in the imaging guidance methods such as ultrasound, computed tomography (CT), and magnetic resonance imaging (MRI) that direct the biopsy needle to the suspicious tissue, malignancies are now being caught earlier and treatments of the diseases are becoming more effective. However, it remains challenging to identify where to take biopsies or decide whether to take biopsies over a suspicious area in order to maximize the diagnostic yield. Therefore, a fine needle technology, which can be integrated with a biopsy needle (11–18 gauge for core needle biopsy or 19–21 gauge for fine needle aspiration) and permit in vivo and real-time assessment of the tissue by providing its structural and functional information before tissue removal, would potentially improve tissue sampling accuracy and reduce the number of unnecessary biopsies.

Optical coherence tomography (OCT) is a recently developed technology that can perform real-time cross-sectional imaging of tissue microanatomy with micrometer-scale resolution [3], [4]. The image contrast in OCT arises from the optical scattering and absorption properties of biological tissues. The use of low coherence gating permits more effective rejection of multiply scattered photons, which enables 1–3 mm imaging depth when near-infrared light (800–1300 nm) is used. The transverse resolution of OCT is governed by the focusing optics, whereas the axial resolution is determined by the coherence length of the light source, which is inversely proportional to the spectral bandwidth of the light source (generally providing an axial resolution of ~3–20 μm). Real-time OCT imaging speed at 4–30 frames/s has been achieved by using the rapid-scanning optical phase-controlled delay line, frequency-sweeping and spectral-domain techniques [5]–[11]. Overall, 1–3 mm imaging depth along with the micrometer-scale resolution and fast imaging speed makes OCT very attractive for 3-D assessments of biological tissues. Recently, new methods/algorithms have been developed to automate OCT data analysis, improving the accuracy of OCT guidance of biopsy [12]–[14].

Extensive ex vivo and in vivo studies have demonstrated that OCT imaging has the potential for diagnosing metaplasia or neoplasia in organs such as the esophagus, colon and pancreas [15]–[17]. Associated tissue morphological changes can be identified by the disruption of normal layered structures revealed by OCT imaging as well as the depth-dependent decay of OCT signals [16], [18]. Various miniature OCT probes with diameters of ~1–2 mm have been developed for in vivo imaging of the epithelial or endothelial lining of internal organs [4], [18]–[25]. Recently, OCT imaging needles with an ultrathin form factor (e.g., ~400 μm) have been developed [26]–[28]. The small diameter of the OCT imaging needle permits its direct integration with a biopsy needle (i.e., going through the biopsy needle core) to provide a “first look” of suspicious tissues before tissue biopsy [26], [28]–[30]. It can be envisioned that under conventional image guidance, the OCT imaging needle is inserted through the biopsy needle into the lesion and high-resolution interstitial image are then taken along the insertion path. Upon identification of a suspicious region by structural imaging or quantitative tissue optical properties, the core biopsy needle can then slide over the OCT imaging needle to the target region. The OCT imaging needle will then be retracted and standard tissue biopsy can then be performed. In addition, an OCT imaging needle can also be used as a stand-alone diagnostic tool for interstitial high-resolution imaging of solid tissues/organs beyond the 1–3 mm OCT imaging depth. Interstitial imaging can be performed by rotating the needle using a dc motor with the image plane perpendicular to the rotation axis. Imaging at various planes can be achieved by inserting or pulling back the OCT needle and a 3-D image can be constructed

The reported imaging needle designs involved the use of either a gradient-index (GRIN) rod lens or a ball-lens [26], [27]. The major challenge in these designs is the lack of mechanical robustness and high-quality optical performance in addition to the complex needle engineering procedure. First of all, the miniature lens was assembled manually within the needle, making it extremely difficult to control the position of the lens relative to the other miniature optical components. As a result, the imaging beam parameters in terms of the spot size and the working distance were difficult to adjust and optimize. Second, the single-mode fiber, the miniature lens and the microreflector were simply glued together, which could not provide sufficient mechanical robustness. Third, the optical window was made out of optical cement, which could be easily damaged when the needle was introduced into soft tissue, causing severe degradation of image quality. Fourth, the engineering protocol was delicate and complicated, resulting in high fabrications costs and making it impractical for mass production.

In order to overcome the challenges mentioned above, a new OCT imaging needle design, which will be reported in this paper, has been developed with improved optical and mechanical performance. In essence, the distal end optics of the new imaging needle employed a 140-μm diameter GRIN fiber lens of a properly chosen pitch number and a glass rod spacer to fully utilize the available numerical aperture (NA) of the miniature optics. A microreflector with a 220 μm diameter was utilized to deflect the imaging beam by ~90° for side-view imaging. The fiber lens and micro-reflector was encased in a small glass tube (of inner diameter ~250–350 μm) to create a robust optical window and then the glass tube was assembled in a hypodermic tube. To eliminate the cylindrical lens effect of the small glass tube (which would severely diverge the imaging beam) [25], a practical and effective approach was employed by filling the inner glass tube with refractive index matching fluid and flattening the outer surface of the glass tube. Two OCT imaging needle prototypes were developed based on the above new design: one has a 22-gauge diameter which can go through a 19-gauge biopsy needle (without correcting the cylindrical lens effect), and the other has a 25-gauge diameter which can go through a 21-gague biopsy needle (with cylindrical lens effect corrected). As we will discuss in the following sections, the 25-gauge needle had to employ a glass tube with smaller inner diameter (~250 μm) than the one used in the 22-gauge needle (~350 μm). The smaller glass tube would induce more severe distortion of the imaging beam, and thus cylindrical lens effect was corrected for the 25-gauge needle. The final imaging beam profiles in the two needles were similar with an astigmatism ratio of ~1.5. The performances of the two imaging needles were tested by imaging various tissue samples including esophagus, spleen, and pancreas along with a high-speed swept-source OCT system. The preliminary results suggests that the new OCT needle designs offer excellent mechanical and optical robustness suitable for continuous interstitial tissue imaging and imaging guidance of tissue biopsy when integrated with a standard biopsy needle.

II. Methods and Materials

Fig. 1(a) shows the schematic of the new prototype imaging needle design. Beam focusing was achieved by a fiber-optic lens, which was robustly constructed by thermal fusion of a single-mode fiber (SMF, 9/125-μm core/cladding), a multimode fiber (MMF, 102/125-μm core/cladding), and a GRIN fiber (100/140-μm core/cladding diameter). The multimode fiber fused in between the SMF and GRIN fiber lens acted as a glass rod spacer to enlarge the input beam size at the entrance of the GRIN fiber lens for fully utilizing the available fiber lens NA. After fixing the length of the spacer (~450 μm long), the focused spot size (i.e., transverse resolution) and the working distance can be precisely tuned by controlling the pitch number of the GRIN fiber lens, as shown in Fig. 1(b). In this prototype, the GRIN fiber lens had an index gradient constant of 3.8 mm⁻¹. When choosing a length of 0.12 pitch (i.e., 200 μm long), a focused spot size of 2ω₀ ~17 μm was achieved at a working distance of ~650 μm. This working distance was sufficient to have the focal spot outside the needle, and it can be fine tuned by adjusting the distance between the fiber-optic lens and microreflector. The use of an arc fusion splicer to thermally fuse all the junctions (SMF-MMF-fiber lens) and a fiber cleaver to accurately control the fiber length greatly improved the mechanical stability and simplified the engineering procedure.

Fig. 1.

(a) Design schematic of the robust OCT imaging needle. The fiber-optic GRIN lens and microreflector were encased in a small glass tube and then assembled in a hypodermic tube. (b) Spot size and working distance as a function of the pitch number of the GRIN fiber lens with a spacer of 450 μm long. The simulation is based on ABCD matrix calculation. The dashed lines on the plots indicate our target parameters.

For side-view imaging, a gold coated microreflector with a diameter of 220 μm was used to deflect the imaging beam by ~90°. To create a durable optical window, the fiber-optic lens and the microreflector was housed within a small glass tube of an inner diameter ~250–350 μm. The glass tube also eased the assembly process making all the optics as a single piece. The positions of the fiber-optic lens and the microreflector inside the glass tube were controlled by two 3-D precision translation stages, respectively. This glass tube, however, would introduce a cylindrical lens effect (i.e., astigmatism) along the azimuthal direction, as shown in Fig. 2(a). Owing to the cylindrical lens effect, the beam focus and working distance along the azimuthal direction would be different from those along the longitudinal direction, and the beam profile will degrade from a circular to an elliptical shape [25]. Ray tracing analysis was conducted to investigate the astigmatism. Fig. 2(b) shows the astigmatism ratio (i.e., the ratio of the transverse resolution in the azimuthal direction to the one in the longitudinal direction) as a function of inner diameter of the glass tube for a given wall thickness (i.e., 45 μm), whereas Fig. 2(c) shows the astigmatism ratio as a function of the wall thickness of the glass tube for a given inner diameter (i.e., 250 μm). It is clear that the astigmatism become more severe with a smaller inner diameter and larger wall thickness. One extreme case of severe astigmatism is cleaving/polishing the GRIN fiber with ~50° to deflect the beam from the fiber side where no additional micro-reflector is needed [28]. In this case (i.e., the inner diameter is zero and the outer diameter is 70 μm), the beam profile will be poor with an astigmatism ratio of ~16.

Fig. 2.

(a) Schematic of the cylindrical lens effect (astigmatism) caused by a small glass tube. (b and c) Simulated astigmatism ratio before and after correction: (b) as a function of inner diameter for a given tube wall thickness (i.e., 45 μm) and (c) as a function of wall thickness for a given inner tube diameter (i.e., 250 μm). The ray tracing analysis shows that the two-step method (i.e., filling the tube with index matching fluid and flattening the outer surface) could well correct the cylindrical lens effect.

To minimize the astigmatism and restore a nearly circular beam profile, the glass tube can be firstly filled with an index-matching fluid (e.g., Zeiss Immersol with a refractive index of 1.518) to reduce the distortion caused by the tube inner surface. Then the outer surface of the glass tube can be polished to have a flat optical window for the beam to pass so that the distortion caused by the outer surface will not take place. Ray tracing simulation shows the astigmatism ratio can be decreased by a factor of 2–3 after the correction. As shown in Fig. 2(b) and (c), the distortion of the beam profile caused by the glass tube can be well corrected by utilizing this two-step and easy-to-implement method.

To further improve the mechanical robustness, the entire glass tube with the fiber-optic lens and micro-reflector was housed in a stainless hypodermic tube that has a precut window for the imaging beam to pass through. Two OCT imaging needles were fabricated with tradeoffs between the needle diameter and robustness as well as the beam profile. The first one was a 22-gauge OCT needle (~560 μm inner diameter and ~720 μm outer diameter) using a glass tube with an inner diameter of 350 μm and an outer diameter of 440 μm. The simulated astigmatism ratio as shown in Fig. 2(b) is ~1.7. No astigmatism correction was performed to this needle. The second one was a 25-gauge OCT needle (~350 μm inner diameter and ~500 μm outer diameter) by using a glass tube with an inner diameter of 250 μm and an outer diameter of 340 μm. The simulated astigmatism ratio as shown in Fig. 2(b) is about 2.4. Astigmatism correction was carried out on this needle, resulting in a lower astigmatism ratio of ~1.3. Table I summarizes the calculated beam spot size of a needle (with the glass tube) at the longitudinal and azimuthal focal positions before and after astigmatism correction. As can be seen, the transverse resolution in the azimuthal direction of an imaging needle with a glass tube will be degraded from 17 to 34 μm and the focal position will greatly shift from 650 μm to about 1 mm. After the two steps correction, the transverse resolution and focal positions along azimuthal direction are almost recovered to those along the longitudinal direction. The small discrepancy is caused by the nonideal index matching fluid, which is not exactly the same as the glass tube.

TABLE I.

Calculated Spot Size and Working Distance Before and After the Correction of the Cylindrical Astigmatism

	Spot size (Longitude × Azimuth) @ focal position
With glass tube (250-μm inner diameter, 340-μm outer diameter)	17 μm × 42 μm @ L focus: 650 μm 88 μm × 34 μm @ A focus: 1020 μm
Fill the inner tube with index matching fluid and flatten the outer surface	17 μm × 22 μm @ L focus: 650 μm 20 μm × 17 μm @ A focus: 610 μm

L: longitudinal direction; A: azimuthal direction.

Finally, the distal end of the hypodermic tube was sharpened to facilitate insertion into solid tissues/organs. A hollow speedometer cable was used to encase the rest of the single-mode fiber from the hypodermic needle to a fiber connector at the proximal end of imaging needle.

Real-time imaging of various tissues including pig esophagus, spleen and pancreas was performed ex vivo with the fine needles along with a swept source OCT system [24]. The swept source had a center wavelength of 1310 nm, a full-width at half-maximum (FWHM) spectral bandwidth of 106 nm and a roundtrip sweeping rate of 8 kHz. The measured axial resolution was ~12 μm in air and the detection sensitivity was ~120 dB. The imaging frame rate was four frames per second with each frame consisting of 3000 axial scans and each axial scan of 2048 pixels. Rectangular OCT imaging was performed by linearly pushing/pulling the needle through the hollow speedometer cable with a linear translation stage. Circumferential imaging was performed by continuously rotating the needle by a compact fiber-optic rotary joint which had ~90% one-way throughput, and ±1.5% rotation speed fluctuation and ±3.5% coupling efficiency fluctuation over 360° rotation.

III. Results and Discussions

A. 22-Gauge OCT Needle Without Astigmatism Correction

Fig. 3(a) and (b), respectively, show the photos of the distal end of the 22-gauge OCT imaging needle based on the previously mentioned design and the entire OCT needle going through a 19-gauge biopsy needle. The prototype imaging needle had a measured confocal parameter of 480 μm, corresponding to a focused spot size of ~20 μm with the focus about 200 μm outside the hypodermic needle.

Fig. 3.

(a) Distal end of the 22-gauge OCT imaging needle without astigmatism correction. (b) Entire 22-gauge OCT imaging needle (including the rigid part—25-gauge hypodermic tube and the flexible part—hollow speedometer cable) going through a 19-gauge biopsy needle.

The imaging performance of the needle was tested by sliding the needle on the tissue surface and then inserting/retracing the needle within the tissue. Fig. 4(a) and (b) show representative cross-sectional OCT images of human fingertip and pig esophagus, respectively. In Fig. 4(a), the stratum corneum, sweat ducts, dermis and fingerprint ridges can be clearly delineated. In Fig. 4(b), the epithelium, lamina propria and muscularis mucosae can be identified. Fig. 5(a) and (b) show two typical ex vivo interstitial OCT images of fresh pig spleen. In Fig. 5(a), the OCT image was taken when the needle was slid on the tissue surface and the signal is dominated by strong back-reflection from the tissue surface (i.e., capsule) of the spleen. In Fig. 5(b), the needle was directly inserted into the tissue and the image was taken when the needle was retracted. The OCT signals are mainly from the pulp (the main components inside the spleen). Compared to Fig. 5(a), the sampling depth in Fig. 5(b) is dramatically limited due to the highly scattering and absorption of lymphocytes in the white pulp and red blood cells in the red pulp. Overall, these preliminary OCT images demonstrate that the fine OCT imaging needle is mechanically robust and optically durable for assessment of tissue microstructures from the surface or interstitially.

Fig. 4.

Representative OCT images of (a) fingertip and (b) pig esophagus with the 22-gauge OCT needle slid on the tissue surface. SD: sweat duct; SC: stratum corneum; D: sermis; E: spithelium; LP: lamina propria; MM: luscularis mucosae.

Fig. 5.

Representative OCT images of pig spleen acquired with a 22-gauge OCT needle. (a) Needle was slid on the tissue capsule. (b) Needle was inserted and retracted from the tissue. The inset shows a two times magnified area marked by the dashed rectangle. The OCT signals exhibit different depth-dependent decay in (a) capsule and (b) pulp, indicating different optical properties of different tissue components.

B. 25-Gauge OCT Needle With Astigmatism Correction

Another prototype OCT imaging needle was built with the astigmatism corrected following the above discussed method. Fig. 6(a) shows a photo of the distal end of the 25-gauge imaging needle. This prototype had a measured confocal parameter of 380 μm, corresponding to a focused spot size of ~18 μm with the focus about 300 μm outside the hypodermic needle. Fig. 6(b) shows the beam profiles near the focus (the upper one) and about 3 mm away from the focus (the lower one). As can be seen, a nearly round beam profile was achieved, suggesting the cylindrical lens effect caused by the small glass tube was well corrected.

Fig. 6.

(a) Distal end of the 25-gauge OCT imaging needle with astigmatism correction. (b) Beam profiles at the focus and ~3 mm away from the focus. The round beam profiles suggest that the cylindrical lens effect was well corrected.

Fig. 7(a) shows a representative ex vivo cross-sectional OCT image of mouse esophageal tissue. In this paper, a 20-gauge hypodermictube (of an ~900 μm outer diameter), which mimicked the biopsy needle was inserted into the lumen of the esophagus. The 20-gauge tube had an ~240° open window and the imaging needle was rotated inside the 20-gauge tube. Normal layered structure is evident, including mucosa, submucosa, and muscularis. The thin epithelium can also be identified as indicated by the arrows. Fig. 7(b) shows a typical ex vivo cross-sectional OCT image of pig pancreas tissue. The OCT needle was directly inserted into the pancreas tail and the OCT image was taken when the needle was pulled back. The structure indicated by the circle, approximately 200 μm in diameter, may be related to islets of Langerhans, which are highly scattering granules within the pancreatic parenchyma tissue [31]. These initial OCT needle-imaging results demonstrate the ability of the fine needle for the assessment of solid tissues/organs.

Fig. 7.

(a) Representative ex vivo OCT image of fresh mouse esophageal tissue when the needle was rotated in a 20-gauge biopsy needle. M1: mucosa; S: submucosa; M2: muscularis. The structure indicated by the arrows is epithelium. (b) Typical ex vivo OCT image of fresh pig pancreatic tail when the needle was directly inserted in the sample and retracted. The structure indicated by the circle may be related to islets of Langerhans. The inset is a two times magnified area marked by the dashed rectangle.

IV. Conclusion

In summary, two fine OCT imaging needles have been developed with improved optical quality and mechanical robustness. The distal end optics of the new imaging needle design employed a GRIN fiber lens of a properly chosen pitch and a glass rod spacer to fully utilize the available NA of the miniature optics. The fiber-optic lens and the microreflector were housed within a small glass tube to create a durable optical window. The cylindrical lens effect caused by the glass tube was analyzed using ray tracing and was well corrected by a two-step method. In the both needles, the cylindrical lens effect was controlled to an acceptable astigmatism ratio (~1.3–1.7). The preliminary real-time OCT needle imaging results acquired with a swept source OCT system suggested the feasibility of in vivo interstitial imaging of solid (soft) tissues/organs. The small size permits direct integration of an OCT imaging needle with standard excisional biopsy devices such as a core biopsy needle (11–18 gauge) or a fine aspiration needle (19–21 gauge) for minimally invasive interstitial diagnosis and image-guided biopsy in solid organs such as the spleen, pancreas and prostate.

Biographies

YicongWu received the B.S. and M.S. degrees from the Department of Biomedical Engineering, Zhejiang University, Hangzhou, China, in 1999 and 2002, respectively, and the Ph.D. degree from the Department of Electronic and Computer Engineering, The Hong Kong University of Science and Technology, Kowloon, Hong Kong, in 2006.

From 2007 to 2008, he was a Postdoctoral Fellow with the Department of Bioengineering, University of Washington, Seattle. In 2009, he joined the Department of Biomedical Engineering, Johns Hopkins University, Baltimore, MD. His research interests include the development of endomicroscopy technologies for nonlinear optical imaging and optical coherence tomography.

Jiefeng Xi received the B.S. and M.S. degrees from the Department of Precision Instruments, Tsinghua University, Beijing, China, in 2002 and 2005, respectively.

In 2006, he joined the Biophotonics Research Laboratory, Department of Bioengineering, University of Washington, Seattle, as a Research Fellow. Since 2009, he has been with the Department of Biomedical Engineering, Johns Hopkins University, Baltimore, MD. His current research interests include using optical coherence tomography and signal processing techniques for the noninvasive assessment of biological tissues.

Li Huo received the B.S., M.S., and Ph.D. degrees in electrical engineering from Tsinghua University, Beijing, China, in 2000, 2003, and 2005, respectively.

From 2005 to 2008, he was a Research Associate with the Department of Information Engineering, Chinese University of Hong Kong, Shatin, Hong Kong, where he was engaged on optical communications. In 2008, he joined the Department of Bioengineering, University of Washington, Seattle, as a Senior Postdoctoral Fellow. Since 2009, he has been with the Department of Biomedical Engineering, Johns Hopkins University, Baltimore, MD. His current research interests include developing novel optical coherence tomography techniques for the noninvasive assessment of biological tissues.

Jason Padvorac, photograph and biography not available at the time of publication.

Eun Ji Shin, photograph and biography not available at the time of publication.

Samuel A. Giday, photograph and biography not available at the time of publication.

Anne Marie Lennon, photograph and biography not available at the time of publication.

Marcia Irene F. Canto, photograph and biography not available at the time of publication.

Joo Ha Hwang (M’09), photograph and biography not available at the time of publication.

Xingde Li (M’06) received the B.S. degree in physics from the University of Science and Technology of China, Hefei, China, in 1990, and the Ph.D. degree in physics from the University of Pennsylvania, Philadelphia, in 1998.

He was a Postdoctoral Researcher of biomedical optics with the Ultrafast Optics Group, Research Laboratory of Electronics, Massachusetts Institute of Technology, Cambridge, for three years. In fall 2001, he joined the Department of Bioengineering, University of Washington, Seattle, as an Assistant Professor, and was promoted to a tenured Associate Professor in 2007. Since 2009, he has been with the Department of Biomedical Engineering, Johns Hopkins University, Baltimore, MD. His current research interests include biophotonics imaging technologies (including optical coherence tomography, multiphoton fluorescence, diffuse light tomography, endomicroscopy, etc.) and their applications in biomedicine, nanobiophotonics such as development of nanocontrast agents for improving optical imaging sensitivity and molecular specificity.

Dr. Li is a member of the Optical Society of America, the International Society for Optical Engineering. He is on the editorial board of the IEEE Transactions on Biomedical Engineering and the Journal of Biomedical Optics. He is currently the Chair of the Emerging Technologies Committee of the IEEE Engineering in Medicine and Biology Society. He was the recipient of the National Science Foundation Career Award in 2004.