Ultrahigh-resolution Optical Coherence Tomography System Research Articles

Classifying breast cancer in ultrahigh-resolution optical coherence tomography images using convolutional neural networks.

Optical coherence tomography (OCT) is being investigated in breast cancer diagnostics as a real-time histology evaluation tool. We present a customized deep convolutional neural network (CNN) for classification of breast tissues in OCT B-scans. Images of human breast samples from mastectomies and breast reductions were acquired using a custom ultrahigh-resolution OCT system with 2.72 µm axial resolution and 5.52 µm lateral resolution. The network achieved 96.7% accuracy, 92% sensitivity, and 99.7% specificity on a dataset of 23 patients. The usage of deep learning will be important for the practical integration of OCT into clinical practice.

Ultrahigh resolution spectral-domain optical coherence tomography using the 1000-1600 nm spectral band.

Ultrahigh resolution optical coherence tomography (UHR-OCT) can image microscopic features that are not visible with the standard OCT resolution of 5-15 µm. In previous studies, high-speed UHR-OCT has been accomplished within the visible (VIS) and near-infrared (NIR-I) spectral ranges, specifically within 550-950 nm. Here, we present a spectral domain UHR-OCT system operating in a short-wavelength infrared (SWIR) range from 1000 to 1600 nm using a supercontinuum light source and an InGaAs-based spectrometer. We obtained an axial resolution of 2.6 µm in air, the highest ever recorded in the SWIR window to our knowledge, with deeper penetration into tissues than VIS or NIR-I light. We demonstrate imaging of conduction fibers of the left bundle branch in freshly excised porcine hearts. These results suggest a potential for deep-penetration, ultrahigh resolution OCT in intraoperative applications.

The Association between Tear Film Thickness as Measured with OCT and Symptoms and Signs of Dry Eye Disease: A Pooled Analysis of 6 Clinical Trials

Purpose: To determine the association between tear film thickness (TFT) as measured with ultra-high resolution optical coherence tomography (UHR-OCT) and signs and symptoms of dry eye disease (DED). Methods: A total of 450 eyes from 225 patients with DED from six different randomized clinical trials were included in this pooled analysis. In all subjects, TFT was measured with a custom-built UHR-OCT system. Symptoms of DED were quantified using a standardized Ocular Surface Disease Index (OSD)I questionnaire and clinical signs including tear film break up time (TFBUT) and Schirmer I test were assessed. Associations of the average TFT with OSDI, TFBUT, and Schirmer I test were calculated using a linear regression analysis. Results: The average TFT of the included sample (mean age, 45.0 ± 13.3 years; 65% female) was 4.2 ± 0.5 µm and the OSDI 36.2 ± 10.4. A significant negative correlation was found between TFT and OSDI (r = −0.36 to −0.31; p < 0.001). Tear break up time and Schirmer I test were not correlated with OSDI. Significant albeit weak correlations were found between TFT and TFBUT (r = 0.17 to 0.25; p < 0.01) as well as Schirmer I (r = 0.36 to 0.37; p < 0.001). Subgroup analysis revealed that the correlation was stronger in the subjects with abnormal Schirmer I (<15 mm; r = 0.50 to 0.54; p < 0.001). Conclusions: The present study demonstrates an objective measurement of TFT using a novel OCT approach for DED that correlates with symptoms and signs of DED. Our data are consistent with the idea that TFT represents the aqueous-deficient component of DED.

CorneaNet: fast segmentation of cornea OCT scans of healthy and keratoconic eyes using deep learning.

Deep learning has dramatically improved object recognition, speech recognition, medical image analysis and many other fields. Optical coherence tomography (OCT) has become a standard of care imaging modality for ophthalmology. We asked whether deep learning could be used to segment cornea OCT images. Using a custom-built ultrahigh-resolution OCT system, we scanned 72 healthy eyes and 70 keratoconic eyes. In total, 20,160 images were labeled and used for the training in a supervised learning approach. A custom neural network architecture called CorneaNet was designed and trained. Our results show that CorneaNet is able to segment both healthy and keratoconus images with high accuracy (validation accuracy: 99.56%). Thickness maps of the three main corneal layers (epithelium, Bowman's layer and stroma) were generated both in healthy subjects and subjects suffering from keratoconus. CorneaNet is more than 50 times faster than our previous algorithm. Our results show that deep learning algorithms can be used for OCT image segmentation and could be applied in various clinical settings. In particular, CorneaNet could be used for early detection of keratoconus and more generally to study other diseases altering corneal morphology.

Cellular resolution corneal imaging with extended imaging range.

Current optical coherence tomography (OCT) technology, which is used for imaging the eye's anterior segment, has been established as a clinical gold standard for the diagnosis of corneal diseases. However, the cellular resolution level information that is critical for many clinical applications is still not available. The major technical challenges toward cellular resolution OCT imaging are the limited ranging depth and depth of focus (DOF). In this work, we present a novel ultrahigh resolution OCT system that achieves an isotropic spatial resolution of <2 µm in tissue. The proposed system could approximately double the ranging depth and extend the DOF using the dual-spectrometer design and the forward-model based digital refocusing method, respectively. We demonstrate that the novel system is capable of visualizing the full thickness of the pig cornea over the ranging depth of 3.5 mm and the border of the corneal endothelial cells 8 times Rayleigh range away from the focal plane. This technology has the potential to realize cellular resolution corneal imaging in vivo.

Effect of Single Instillation of Two Hyaluronic Acid-Based Topical Lubricants on Tear Film Thickness in Patients with Dry Eye Syndrome.

This study investigated the effect of single administration of 2 different hyaluronic acid-based topical lubricants on tear film thickness (TFT) in patients with moderate-to-severe dry eye disease (DED). Sixty patients were randomized to receive eye drops containing unpreserved sodium hyaluronate, triglycerides, and phospholipids (HTP), unpreserved sodium hyaluronate (HA), or unpreserved sodium chloride (NaCl). TFT and lipid layer thickness (LLT) were assessed before and at defined time points after single instillation with a custom-built ultrahigh-resolution optical coherence tomography system and a white light interferometer. Standard tests for DED such as a visual analogue scale (VAS), corneal fluorescein staining (CFS), Schirmer I test, and tear break-up time (BUT) were performed. The time course in TFT was significantly different between the 3 administered agents (P < 0.001). Single administration of HTP significantly increased TFT over a period of 40 min, an increase in TFT over 20 min was observed after instillation of HA, whereas NaCl had no effect. No significant difference in the change in BUT, CFS, or LLT was observed between the 3 groups (P = 0.57, 0.97, and 0.86, respectively). A significant improvement in subjective symptoms (VAS) was found after single instillation of HTP (P = 0.03 vs. baseline) and HA, but not after administration of NaCl (P = 0.03, P < 0.01, and P = 0.08 vs. baseline, respectively, P = 0.57 between groups). Single administration of HTP and HA significantly increased TFT showing a longer residence time and improved subjective symptoms, whereas instillation of NaCl had no effect. Long-term studies with multiple instillations will be necessary to further investigate the effect of the tested products on the clinical outcome. ClinicalTrials.gov ID: NCT03161080.

Effect of changing from preserved prostaglandins to preservative-free tafluprost in patients with glaucoma on tear film thickness

Long-term glaucoma therapy with preservative-containing eye drops may impact ocular surface health. This study was performed to investigate whether a switch from preserved topical prostaglandin therapy to preservative-free tafluprost therapy improves precorneal tear film thickness in patients with glaucoma or ocular hypertension. A total of 30 patients who were under topical preservative-containing prostaglandin monotherapy for at least 6 months were included. Patients were then switched from preserved prostaglandin therapy to unpreserved tafluprost drops once daily. Tear film thickness was measured at baseline and 4 and 12 weeks after therapy change with an ultrahigh-resolution optical coherence tomography system. Furthermore, clinical measures of ocular surface disease were determined and symptoms were assessed using the Dry Eye-Related Quality-of-Life Score. After switching to unpreserved tafluprost, tear film thickness significantly increased from 4.7 ± 0.5 to 5.0 ± 0.6 µm 4 weeks after therapy change and still tended to be increased after 12 weeks (4.8 ± 0.7 µm). Breakup time significantly increased from 5.1 ± 2.3 to 7.2 ± 3.4 s and to 10.1 ± 3.6 s after therapy change. In addition, a significant decrease in corneal staining score from 1.8 ± 0.7 to 1.4 ± 0.8 after 4 weeks and to 0.7 ± 0.7 after 12 weeks treatment was observed. Switching to preservative-free drops reduced Dry Eye-Related Quality-of-Life Score from 11.4 ± 11.0 to 5.7 ± 6.4 and to 4.7 ± 7.5. Our data show that switching to preservative-free tafluprost leads to an increase in tear film thickness, breakup time, and an improvement of Dry Eye-Related Quality-of-Life Score. Our results therefore indicate that a switch to unpreserved tafluprost is beneficial for ocular surface health in patients under long-term preserved prostaglandin eye drops.

Distinguishing Keratoconic Eyes and Healthy Eyes Using Ultrahigh-Resolution Optical Coherence Tomography–Based Corneal Epithelium Thickness Mapping

To find differences in epithelial thickness (ET) maps of eyes with keratoconus (KC) and healthy eyes. Institutional cross-sectional study. In this study 40 keratoconic eyes and 76 healthy eyes were scanned using a custom-built ultrahigh-resolution optical coherence tomography system. Automated segmentation ET maps with 17 subsectors were calculated (central, temporal inferior, temporal superior, nasal inferior, and nasal superior area). The thinnest point of the epithelium (minET), the thickest point of the epithelium (maxET), and the thinnest point diagonally opposing the thickest point (ETmax/op) were additional parameters. Ratios were calculated as follows: minET/diagonally opposing point (R1), maxET/diagonally opposing point (R2), inferior temporal area/superior nasal area (RTI/NS), and inferior/superior hemisphere (RI/S). Furthermore, collected parameters were analyzed regarding their diagnostic accuracy (area under the curve; AUC). Statistically significant differences were as follows: central ET, 46.25 ± 2.56/50.91 ± 1.66; minET, 38.50 ± 2.10/46.79 ± 1.27; ETmax/op, 47.14 ± 2.45/49.60 ± 1.57; temporal inferior area: 43.93 ± 2.95/51.04 ± 1.51 (all mean ± standard deviation, μm); R1, 0.76 ± 0.09/0.93 ± 0.04; R2, 1.08 ± 0.04/1.21 ± 0.16; RTI/NS, 0.85 ± 0.08/1.02 ± 0.04; RI/S: 0.92 ± 0.07/0.99 ± 0.02. AUC values were R1: 0.979 (confidence interval [CI]: 0.957-1.000), RTI/NS: 0.977 (CI: 0.951-1.000), and minET: 0.928 (CI: 0.880-0.977). Epithelial thickness maps could clearly visualize different ET patterns. Parameters with the highest potential of diagnostic discrimination between eyes with KC and healthy eyes were, in descending order, R1, RTI/NS, and minET. Consequently, epithelial thickness irregularity and asymmetry seem to be the most promising diagnostic factor in terms of discriminating between keratoconic eyes and healthy eyes.

A Controlled, Randomized Double-Blind Study to Evaluate the Safety and Efficacy of Chitosan-N-Acetylcysteine for the Treatment of Dry Eye Syndrome.

This study was designed to evaluate the effect of chitosan-N-acetylcysteine (C-NAC) eye drops on tear film thickness (TFT) in patients with dry eye syndrome (DES). This was a controlled, randomized, double-blind clinical investigation with patients assigned to 2 cohorts. In Cohort I, 21 patients were randomized to receive 1 instillation of C-NAC eye drops in 1 eye and placebo (normal saline solution) in the contralateral eye. In Cohort II, 17 patients were randomized to receive C-NAC eye drops once (QD) or twice (BID) daily for 5 days. TFT was assessed with a custom-built ultrahigh-resolution optical coherence tomography system. In Cohort I, mean TFT increased from 3.9 ± 0.5 μm predose to 4.8 ± 1.1 μm 10 min postdose after treatment with C-NAC. The increase was significantly different from placebo over time (P < 0.0001) and remained stable until 24 h postdose. In Cohort II, TFT increased with QD and BID instillation, with no significant difference between regimens. In both groups, Ocular Surface Disease Index scores improved, fewer patients presented with corneal damage, and symptoms of ocular discomfort/conjunctival redness were reduced. A single instillation of C-NAC significantly increased mean TFT in patients with DES as early as 10 min after instillation and lasted for 24 h. The magnitude of the increase in TFT following a single instillation was comparable with that after instillation twice daily over 5 days. Corneal damage improved in >60% of patients. C-NAC could be a viable treatment option for DES.

Effect of different lubricant eye gels on tear film thickness as measured with ultrahigh-resolution optical coherence tomography.

To compare the effect of a single drop of different lubricant eye gels on tear film thickness (TFT) as measured with ultrahigh-resolution optical coherence tomography (UHR-OCT) in patients with mild-to-moderate dry eye disease (DED). The study followed a randomized, single-masked, observer-blinded parallel group design. Patients received a single dose of either unpreserved trehalose 3%+hyaluronic acid 0.15% (TH), hyaluronic acid 0.2% (HA) or polyethylene glycol 0.4%+propylene glycol 0.3% (PP) eye drops. In total, 60 patients were included and TFT was measured using a custom-built UHR-OCT system. The mean TFT at baseline was 3.5±0.7μm. There was a significant difference in the time-course of TFT between the three groups (p=0.001 between groups). Ten minutes after instillation, TFT increased by 155.8±86.6%, 65.7±71.5% and 33.4±19.6% in the PP, TH and HA group, respectively (p<0.001 between groups). Thirty minutes after instillation, the effect of all three different agents on TFT was comparable. Sixty and 120min after administration, a significant increase in TFT was only seen for the TH group, but not for the other products (60min: p<0.021 between groups; 120min: p<0.037 between groups). In conclusion, we observed differences in TFT after administration of the lubricant gels. Ten minutes after instillation, a pronounced increase in TFT was observed in all groups. As compared to the other products, the combination of trehalose 3%+hyaluronic acid 0.15% offers a significantly longer increase in TFT indicating for a longer residence time.

Optical edge effects create conjunctival indentation thickness artefacts.

Conjunctival compression observed in ultrahigh resolution optical coherence tomography (UHR-OCT) images of contact lens edges could be actual tissue alteration, may be an optical artefact arising from the difference between the refractive indexes of the lens material and the conjunctival tissue, or could be a combination of the two. The purpose of this study is to image the artefact with contact lenses on a non-biological (non-indentable) medium and to determine the origins of the observed conjunctival compression. Two-dimensional cross-sectional images of the edges of a selection of marketed silicone hydrogel and hydrogel lenses (refractive index ranging from 1.40 to 1.43) were acquired with a research grade UHR-OCT system. The lenses were placed on three continuous surfaces, a glass sphere (refractive index n = 1.52), a rigid contact lens (n = 1.376) and the cornea of a healthy human subject (average n = 1.376). The displacement observed was analysed using ImageJ. The observed optical displacement ranged between 5.39(0.06) μm with Acuvue Advance and 11.99(0.18) μm with Air Optix Night & Day when the lens was imaged on the glass reference sphere. Similarly, on a rigid contact lens displacement ranged between 5.51(0.03) and 9.72(0.12) μm. Displacement was also observed when the lenses were imaged on the human conjunctiva and ranged from 6.49(0.80) μm for the 1-day Acuvue Moist to 17.4(0.22) μm for the Pure Vision contact lens. An optical displacement artefact was observed when imaging a contact lens on two rigid continuous surfaces with UHR-OCT where compression or indentation of the surface could not have been a factor. Contact lenses imaged in situ also exhibited displacement at the intersection of the contact lens edge and the conjunctiva, likely a manifestation of both the artefact and compression of the conjunctiva.

Ultra-high resolution optical coherence tomography for encapsulation quality inspection

We present the application of ultra-high resolution optical coherence tomography (UHR-OCT) in evaluation of thin, protective films used in printed electronics. Two types of sample were investigated: microscopy glass and organic field effect transistor (OFET) structure. Samples were coated with thin (1–3 μm) layer of parylene C polymer. Measurements were done using experimental UHR-OCT device based on a Kerr-lens mode locked Ti:sapphire femtosecond laser, photonic crystal fibre and modified, free-space Michelson interferometer. Sub-micron resolution offered by the UHR-OCT system applied in the study enables registration of both interfaces of the thin encapsulation layer. Complete, volumetric characterisation of protective layers is presented, demonstrating possibility to use OCT for encapsulation quality inspection.

Optical coherence tomography as a method of quality inspection for printed electronics products

Application of time domain, ultra high resolution optical coherence tomography (UHR-OCT) in printed electronics products’ quality inspection is demonstrated. Presented study was done using experimental UHR-OCT device based on a Kerr-lens mode locked Ti:sapphire femtosecond laser, photonic crystal fibre and modified, free-space Michelson interferometer. Possibilities of the technique are demonstrated by analysis of an RF antenna—example of printed electronics products. Measurements were done with submicron axial resolution, offered by UHR-OCT system developed in our laboratory. Such high resolution is necessary due to the thickness of material layers used in printed electronics. In addition to tomography imaging, numerical results were compared with data provided by two commercially available measurement devices: Wyko NT3300 optical profiler and Dektak 8 stylus profiler (both Veeco). Comparison of profile heights and their spatial correlation is presented. Ability for full volumetric reconstruction and accuracy justified with reference measurements prove OCT to be a reliable tool in printed electronics product testing.

General Bayesian estimation for speckle noise reduction in optical coherence tomography retinal imagery

An important image post-processing step for optical coherence tomography (OCT) images is speckle noise reduction. Noise in OCT images is multiplicative in nature and is difficult to suppress due to the fact that in addition the noise component, OCT speckle also carries structural information about the imaged object. To address this issue, a novel speckle noise reduction algorithm was developed. The algorithm projects the imaging data into the logarithmic space and a general Bayesian least squares estimate of the noise-free data is found using a conditional posterior sampling approach. The proposed algorithm was tested on a number of rodent (rat) retina images acquired in-vivo with an ultrahigh resolution OCT system. The performance of the algorithm was compared to that of the state-of-the-art algorithms currently available for speckle denoising, such as the adaptive median, maximum a posteriori (MAP) estimation, linear least squares estimation, anisotropic diffusion and wavelet-domain filtering methods. Experimental results show that the proposed approach is capable of achieving state-of-the-art performance when compared to the other tested methods in terms of signal-to-noise ratio (SNR), contrast-to-noise ratio (CNR), edge preservation, and equivalent number of looks (ENL) measures. Visual comparisons also show that the proposed approach provides effective speckle noise suppression while preserving the sharpness and improving the visibility of morphological details, such as tiny capillaries and thin layers in the rat retina OCT images.

Combined optical coherence tomography and electroretinography system for in vivo simultaneous morphological and functional imaging of the rodent retina

A combined ultrahigh resolution optical coherence tomography (UHROCT) and a electroretinography (ERG) system is presented for simultaneous imaging of the retinal structure and physiological response to light stimulation in the rodent eye. The 1060-nm UHROCT system provides approximately 3x5 microm (axialxlateral) resolution in the rat retina and time resolution of 22 micros. A custom-designed light stimulator integrated into the UHROCT imaging probe provides light stimuli with user-selected color, duration, and intensity. The performance of the combined system is demonstrated in vivo in healthy rats, and in a rat model of drug-induced outer retinal degeneration. Experimental results show correlation between the observed structural and physiological changes in the healthy and degenerated retina.

Noninvasive Volumetric Imaging and Morphometry of the Rodent Retina with High-Speed, Ultrahigh-Resolution Optical Coherence Tomography

To demonstrate high-speed, ultrahigh-resolution optical coherence tomography (OCT) for noninvasive, in vivo, three-dimensional imaging of the retina in rat and mouse models. A high-speed, ultrahigh-resolution OCT system using spectral, or Fourier domain, detection has been developed for small animal retinal imaging. Imaging is performed with a contact lens and postobjective scanning. An axial image resolution of 2.8 mum is achieved with a spectrally broadband superluminescent diode light source with a bandwidth of approximately 150 nm at approximately 900-nm center wavelength. Imaging can be performed at 24,000 axial scans per second, which is approximately 100 times faster than previous ultrahigh-resolution OCT systems. High-definition and three-dimensional retinal imaging is performed in vivo in mouse and rat models. High-speed, ultrahigh-resolution OCT enabled high-definition, high transverse pixel density imaging of the murine retina and visualization of all major intraretinal layers. Raster scan protocols enabled three-dimensional volumetric imagingand comprehensive retinal segmentation algorithms allowed measurement of retinal layers. An OCT fundus image, akin to a fundus photograph was generated by axial summation of three-dimensional OCT data, thus enabling precise registration of OCT measurements to retinal fundus features. High-speed, ultrahigh-resolution OCT enables imaging of retinal architectural morphology in small animal models. OCT fundus images allow precise registration of OCT images and repeated measurements with respect to retinal fundus features. Three-dimensional OCT imaging enables visualization and quantification of retinal structure, which promises to allow repeated, noninvasive measurements to track disease progression, thereby reducing the need for killing the animal for histology. This capability can accelerate basic research studies in rats and mice and their translation into clinical patient care.

In vitro measurement of rabbit corneal epithelial thickness using ultrahigh resolution optical coherence tomography

The objective of this study was to reproducibly measure corneal epithelial thickness centrally and at the limbus in the rabbit cornea using ultrahigh resolution optical coherence tomography (OCT). Twelve freshly enucleated New Zealand white rabbit eyes were kept in a moist chamber at 4 degrees C. An ultrahigh resolution OCT system with a spatial resolution of 1.3 microm was used to image the cornea and its component layers. The central and peripheral (limbal) regions of all the samples were scanned within 6 h of harvest in order to minimize the post-mortem degradation of the corneal epithelium. The thickness of the corneal epithelium was determined by measuring the pixel equivalents of the obtained image. Unpaired Student's t-test was used to evaluate differences. The epithelial thickness centrally was found to be 45.8 +/- 2.2 microm, and 37.6 +/- 1.4 microm at the limbus (P < 0.001). Rabbit corneal epithelium is thicker centrally than at the limbus when measured by ultrahigh resolution OCT. This technique will aid in delineating the pathophysiology of diseases of the anterior cornea.

Comparison of ultrahigh- and standard-resolution optical coherence tomography for imaging macular hole pathology and repair

Purpose To compare ultrahigh-resolution optical coherence tomography (UHR-OCT) technology to a standard-resolution OCT instrument for the imaging of macular hole pathology and repair; to identify situations where UHR-OCT provides additional information on disease morphology, pathogenesis, and management; and to use UHR-OCT as a baseline for improving the interpretation of the standard-resolution images. Design Observational and interventional case series. Participants Twenty-nine eyes of 24 patients clinically diagnosed with macular hole in at least one eye. Methods A UHR-OCT system has been developed and employed in a tertiary-care ophthalmology clinic. Using a femtosecond laser as the low-coherence light source, this new UHR-OCT system can achieve an unprecedented 3-μm axial resolution for retinal OCT imaging. Comparative imaging was performed with UHR-OCT and standard 10-μm resolution OCT in 29 eyes of 24 patients with various stages of macular holes. Imaging was also performed on a subset of the population before and after macular hole surgery. Main outcome measures Ultrahigh- and standard-resolution cross-sectional OCT images of macular hole pathologies. Results Both UHR-OCT and standard-resolution OCT exhibited comparable performance in differentiating various stages of macular holes. The UHR-OCT provided improved imaging of finer intraretinal structures, such as the external limiting membrane and photoreceptor inner segment (IS) and outer segment (OS), and identification of the anatomy of successful surgical repair. The improved resolution of UHR-OCT enabled imaging of previously unidentified changes in photoreceptor morphology associated with macular hole pathology and postoperative repair. Visualization of the junction between the photoreceptor IS and OS was found to be an important indicator of photoreceptor integrity for both standard-resolution and UHR-OCT images. Conclusions Ultrahigh-resolution optical coherence tomography improves the visualization of the macular hole architectural morphology. The increased resolution of UHR-OCT enables the visualization of photoreceptor morphology associated with macular holes. This promises to lead to a better understanding of the pathogenesis of macular holes, the causes of visual loss secondary to macular holes, the timing of surgical repair, and the evaluation of postsurgical outcome. Ultrahigh-resolution optical coherence tomography imaging of macular holes that correspond to known alterations in retinal morphology can be used to interpret retinal morphology in UHR-OCT images. Comparisons of UHR-OCT images with standard-resolution OCT images can establish a baseline for the better interpretation of clinical standard-resolution OCT images. The ability to visualize photoreceptors and their integrity or impairment is an indicator of macular hole progression and surgical outcome.