Manual Segmentation Procedure Research Articles

Using an anomaly detection approach for the segmentation of colorectal cancer tumors in whole slide images

Colorectal cancer (CRC) is the second most commonly diagnosed cancer in the United States. Genetic testing is critical in assisting in the early detection of CRC and selection of individualized treatment plans, which have shown to improve the survival rate of CRC patients. The tissue slide review (TSR), a tumor tissue macro-dissection procedure, is a required pre-analytical step to perform genetic testing. Due to the subjective nature of the process, major discrepancies in CRC diagnostics by pathologists are reported, and metrics for quality are often only qualitative. Progressive context encoder anomaly detection (P-CEAD) is an anomaly detection approach to detect tumor tissue from whole slide images (WSIs), since tumor tissue is by its nature, an anomaly. P-CEAD-based CRC tumor segmentation achieves a 71% 26% sensitivity, 92% 7% specificity, and 63% 23% F1 score. The proposed approach provides an automated CRC tumor segmentation pipeline with a quantitatively reproducible quality compared with the conventional manual tumor segmentation procedure.

A detailed manual segmentation procedure for the hypothalamus for 3T T1-weighted MRI

The hypothalamus is a small grey matter structure which plays a crucial role in many physiological functions. Some studies have found an association between hypothalamic volume and psychopathology, which stresses the need for a standardized method to maximize segmentation accuracy. Here, we provide a detailed step-by-step method outlining the procedures to manually segment the hypothalamus using anatomical T1w images from 3T scanners, which many neuroimaging studies collect as a standard anatomical reference image. We compared volumes generated by manual segmentation and those generated by an automatic algorithm, observing a significant difference between automatically and manually segmented hypothalamus volumes on both sides (left: U = 222842, p-value < 2.2e-16; right: U = 218520, p- value < 2.2e-16).•Significant difference exists between existing automatic segmentation methods and the manual segmentation procedure.•We discuss potential drift effects, segmentation quality issues, and suggestions on how to mitigate them.•We demonstrate that the present manual segmentation procedure using standard T1-weighted MRI may be significantly more accurate than automatic segmentation outputs.

Fully automated deep learning-based localization and segmentation of the locus coeruleus in aging and Parkinson\u2019s disease using neuromelanin-sensitive MRI

PurposeDevelopment and performance measurement of a fully automated pipeline that localizes and segments the locus coeruleus in so-called neuromelanin-sensitive magnetic resonance imaging data for the derivation of quantitative biomarkers of neurodegenerative diseases such as Alzheimer’s disease and Parkinson’s disease.MethodsWe propose a pipeline composed of several 3D-Unet-based convolutional neural networks for iterative multi-scale localization and multi-rater segmentation and non-deep learning-based components for automated biomarker extraction. We trained on the healthy aging cohort and did not carry out any adaption or fine-tuning prior to the application to Parkinson’s disease subjects.ResultsThe localization and segmentation pipeline demonstrated sufficient performance as measured by Euclidean distance (on average around 1.3mm on healthy aging subjects and 2.2mm in Parkinson’s disease subjects) and Dice similarity coefficient (overall around 71% on healthy aging subjects and 60% for subjects with Parkinson’s disease) as well as promising agreement with respect to contrast ratios in terms of intraclass correlation coefficient of ge 0.80 for healthy aging subjects compared to a manual segmentation procedure. Lower values (ge 0.48) for Parkinson’s disease subjects indicate the need for further investigation and tests before the application to clinical samples.ConclusionThese promising results suggest the usability of the proposed algorithm for data of healthy aging subjects and pave the way for further investigations using this approach on different clinical datasets to validate its practical usability more conclusively.

Learning From Mouse CT-Scan Brain Images To Detect MRA-TOF Human Vasculatures.

The earlier studies on brain vasculature semantic segmentation used classical image analysis methods to extract the vascular tree from images. Nowadays, deep learning methods are widely exploited for various image analysis tasks. One of the strong restrictions when dealing with neural networks in the framework of semantic segmentation is the need to dispose of a ground truth segmentation dataset, on which the task will be learned. It may be cumbersome to manually segment the arteries in a 3D volumes (MRA-TOF typically). In this work, we aim to tackle the vascular tree segmentation from a new perspective. Our objective is to build an image dataset from mouse vasculatures acquired using CT-Scans, and enhance these vasculatures in such a way to precisely mimic the statistical properties of the human brain. The segmentation of mouse images is easily automatized thanks to their specific acquisition modality. Thus, such a framework allows to generate the data necessary for the training of a Convolutional Neural Network - i.e. the enhanced mouse images and there corresponding ground truth segmentation - without requiring any manual segmentation procedure. However, in order to generate an image dataset having consistent properties (strong resemblance with MRA images), we have to ensure that the statistical properties of the enhanced mouse images do match correctly the human MRA acquisitions. In this work, we evaluate at length the similarities between the human arteries as acquired on MRA-TOF and the "humanized" mouse arteries produced by our model. Finally, once the model duly validated, we experiment its applicability with a Convolutional Neural Network.

Deep learning based joint segmentation and characterization of multi-class retinal fluid lesions on OCT scans for clinical use in anti-VEGF therapy.

In anti-vascular endothelial growth factor (anti-VEGF) therapy, an accurate estimation of multi-class retinal fluid (MRF) is required for the activity prescription and intravitreal dose. This study proposes an end-to-end deep learning-based retinal fluids segmentation network (RFS-Net) to segment and recognize three MRF lesion manifestations, namely, intraretinal fluid (IRF), subretinal fluid (SRF), and pigment epithelial detachment (PED), from multi-vendor optical coherence tomography (OCT) imagery. The proposed image analysis tool will optimize anti-VEGF therapy and contribute to reducing the inter- and intra-observer variability. The proposed RFS-Net architecture integrates the atrous spatial pyramid pooling (ASPP), residual, and inception modules in the encoder path to learn better features and conserve more global information for precise segmentation and characterization of MRF lesions. The RFS-Net model is trained and validated using OCT scans from multiple vendors (Topcon, Cirrus, Spectralis), collected from three publicly available datasets. The first dataset consisted of OCT volumes obtained from 112 subjects (a total of 11,334B-scans) is used for both training and evaluation purposes. Moreover, the remaining two datasets are only used for evaluation purposes to check the trained RFS-Net's generalizability on unseen OCT scans. The two evaluation datasets contain a total of 1572 OCT B-scans from 1255 subjects. The performance of the proposed RFS-Net model is assessed through various evaluation metrics. The proposed RFS-Net model achieved the mean F1 scores of 0.762, 0.796, and 0.805 for segmenting IRF, SRF, and PED. Moreover, with the automated segmentation of the three retinal manifestations, the RFS-Net brings a considerable gain in efficiency compared to the tedious and demanding manual segmentation procedure of the MRF. Our proposed RFS-Net is a potential diagnostic tool for the automatic segmentation of MRF (IRF, SRF, and PED) lesions. It is expected to strengthen the inter-observer agreement, and standardization of dosimetry is envisaged as a result.

Fully automatic model‐based segmentation and classification approach for MRI brain tumor using artificial neural networks

SummaryThe accuracy of brain tumor diagnosis based on medical images is greatly affected by the segmentation process. The segmentation determines the tumor shape, location, size, and texture. In this study, we proposed a new segmentation approach for brain tissues using MR images. The method includes three computer vision fiction strategies which are enhancing images, segmenting images, and filtering out non ROI based on the texture and HOG features. A fully automatic model‐based trainable segmentation and classification approach for MRI brain tumour using artificial neural networks to precisely identifying the location of the ROI. Therefore, the filtering out non ROI process have used in view of histogram investigation to avert the non ROI and select the correct object in brain MRI. However, identification the tumor kind utilizing the texture features. A total of 200 MRI cases are utilized for the comparing between automatic and manual segmentation procedure. The outcomes analysis shows that the fully automatic model‐based trainable segmentation over performs the manual method and the brain identification utilizing the ROI texture features. The recorded identification precision is 92.14%, with 89 sensitivity and 94 specificity.

Heterogeneous choroidal thickness pattern in diabetic patients without diabetic retinopathy

PurposeThe authors intend to present an analysis and comparison of the choroidal thickness in more distant superior regions. Choroidal thickness is increased in a heterogeneous pattern superiorly to the fovea in diabetic patients without diabetic retinopathy.MethodsObservational case‐control study of 60 eyes of 60 patients with diabetes mellitus type 2 without diabetic retinopathy. A control group of 31 eyes of 31 patients without diabetes with similar demographic features was also created. All the patients were recruited from the same outpatient clinic. Enhanced depth imaging spectral‐domain optical coherence tomography (EDI SD‐OCT SPECTRALIS®) was performed and high‐resolution macular scans were obtained. Choroidal thickness was evaluated 2000 μm superiorly to the fovea by manual layer segmentation. Previous studies have validated the validity of using a manual layer segmentation procedure in OCT. Statistical analysis was done with SPSS Statistics. A p value &lt;0.05 was considered statistically significant.ResultsDiabetic patients were in average 67.0 ± 9.6 years old and 56.7% were male. Average choroidal thickness 2000 μm superiorly to the fovea was 287.77 ± 74.32 μm in diabetic patients and 249.81 ± 53.96 μm in non‐diabetic patients. There was a significant increase in choroidal thickness in diabetic patients when compared to non‐diabetic patients (T‐student, p = 0.014).ConclusionsChoroidal thickness can be accurately measured using high‐definition macular OCT scans. This study documented structural differences in the choroid associated with diabetes mellitus type 2 that are not only present in the central foveal area as reported in previous studies. Diabetes Mellitus is associated with multiple microvascular dysfunctions that are probably coupled with choroidal vascular modifications that induce changes in thickness.

Magnetic resonance scanning and image segmentation procedure at 3 T for volumetry of human hippocampal subfields

BACKGROUND: Recent evidence suggests that dementia affects hippocampal substructures differentially and thus identifying anatomical details of this structure is potentially clinically important. Visualising details of human hippocampal substructures in vivo is challenging by imaging due to the small size of the medial temporal lobe structure. METHODS: MRI data were acquired with a 3 T MR scanner using a 2D multi-echo spin echo pulse sequence at such a spatial resolution to reveal hippocampal subfield boundaries. These images were used to develop a manual segmentation procedure for the hippocampal subfields based on image contrast within the structures and/or geometric constraints by anatomical landmarks. RESULTS: The T2-images were used to devise a segmentation protocol for Cornu Ammonis (CA) CA1, CA2, CA3, dentate gyrus, subiculum and lumped Stratum Lacunosum + Stratum Moleculare + Stratum Radiatum. The segmentation protocol was applied to MRI data from healthy young and aged controls as well as a small cohort of mild cognitive impairment (MCI) subjects. The reported subfield volumes showed high levels of inter-rater reliability arguing for potential as a tool in documenting subfield volumetry in clinical research. CONCLUSIONS: A comprehensive and robust MRI protocol which allows the labelling of six separate hippocampal subfields from images acquired with a 3 T clinical scanner is presented to promote hippocampal subfield volumetry of clinical cohorts.

Automated methods for hippocampus segmentation: the evolution and a review of the state of the art.

The segmentation of the hippocampus in Magnetic Resonance Imaging (MRI) has been an important procedure to diagnose and monitor several clinical situations. The precise delineation of the borders of this brain structure makes it possible to obtain a measure of the volume and estimate its shape, which can be used to diagnose some diseases, such as Alzheimer's disease, schizophrenia and epilepsy. As the manual segmentation procedure in three-dimensional images is highly time consuming and the reproducibility is low, automated methods introduce substantial gains. On the other hand, the implementation of those methods is a challenge because of the low contrast of this structure in relation to the neighboring areas of the brain. Within this context, this research presents a review of the evolution of automatized methods for the segmentation of the hippocampus in MRI. Many proposed methods for segmentation of the hippocampus have been published in leading journals in the medical image processing area. This paper describes these methods presenting the techniques used and quantitatively comparing the methods based on Dice Similarity Coefficient. Finally, we present an evaluation of those methods considering the degree of user intervention, computational cost, segmentation accuracy and feasibility of application in a clinical routine.

The Long-Term Course of Functional and Anatomical Recovery After Macular Hole Surgery

To investigate the long-term effect of macular hole surgery, foveal structure and the thickness of retinal layers were analyzed with spectral-domain optical coherence tomography (SD-OCT). The long-term postoperative course of macular thickness and best-corrected visual acuity (BCVA) were followed. In a retrospective cohort study, SD-OCT scans were obtained from the horizontal midline in 51 eyes 54±20 months postoperatively and from 30 control eyes. Retinal layer thickness was measured with a manual segmentation procedure aided by a customized computer program. BCVA was followed and macular thickness was quantified over time with the time-domain (TD) OCT Fast Macular Thickness program for up to 91 months. Median foveal thickness between the outer plexiform and ganglion cell layers was greater than normal while that of the other retinal layers was normal. The median foveal shape remained slightly distorted. The postoperative decrease of central macular thickness toward normal values was delayed to 28 months postoperatively. Nasal macular thickness was decreased to normal at 6 months while superior, temporal, and inferior macular thickness was decreased to normal at 1 to 2 months postoperatively. Preoperative mean BCVA was 20/100±3 lines. Postoperatively, mean BCVA was 20/44±2 lines at 3 to 6 months, 20/40±2 lines at 1 year, 20/32±2 lines at 2 years, and 20/28±1 line after a mean follow-up period of 54±20 months. Long-term postoperatively, the median thickness of retinal layers remains slightly thickened between the outer plexiform and the ganglion cell layer. The process of gradual recovery may continue for several years after macular hole surgery.

Transition Zones between Healthy and Diseased Retina in Choroideremia (CHM) and Stargardt Disease (STGD) as Compared to Retinitis Pigmentosa (RP)

To describe the structural changes across the transition zone (TZ) in choroideremia (CHM) and Stargardt disease (STGD) and to compare these to the TZ in retinitis pigmentosa (RP). Frequency-domain (Fd)OCT line scans were obtained from seven patients with CHM, 20 with STGD, and 12 with RP and compared with those of 30 previously studied controls. A computer-aided manual segmentation procedure was used to determine the thicknesses of the outer segment (OS) layer, the outer nuclear layer plus outer plexiform layer (ONL+), the retinal pigment epithelium plus Bruch's membrane (RPE+BM), and the outer retina (OR). The TZ, while consistent within patient groups, showed differences across disease groups. In particular, (1) OS loss occurred before ONL+ loss in CHM and RP, whereas ONL+ loss occurred before OS loss in STGD; (2) ONL+ was preserved over a wider region of the retina in CHM than in RP; (3) RPE+BM remained normal across the RP TZ, but was typically thinned in CHM. In some CHM patients, it was abnormally thin in regions with normal OS and ONL+ thickness. In STGD, RPE+BM was thinned by the end of the TZ; and (4) the disappearances of the IS/OS and OLM were more abrupt in CHM and STGD than in RP. On fdOCT scans, patients with RP, CHM, and STGD all have a TZ between relatively healthy and severely affected retina. The patterns of changes in the receptor layers are similar within a disease category, but different across categories. The findings suggest that the pattern of progression of each disease is distinct and may offer clues for strategies in the development of future therapies.

Automatic and manual segmentation of healthy retinas using high-definition optical coherence tomography

This study compared automatic- and manual segmentation modalities in the retina of healthy eyes using high-definition optical coherence tomography (HD-OCT). Twenty retinas in 20 healthy individuals were examined using an HD-OCT system (Carl Zeiss Meditec, Inc.). Three-dimensional imaging was performed with an axial resolution of 6 μm at a maximum scanning speed of 25,000 A-scans/second. Volumes of 6 × 6 × 2 mm were scanned. Scans were analysed using a matlab-based algorithm and a manual segmentation software system (3D-Doctor). The volume values calculated by the two methods were compared. Statistical analysis revealed a high correlation between automatic and manual modes of segmentation. The automatic mode of measuring retinal volume and the corresponding three-dimensional images provided similar results to the manual segmentation procedure. Both methods were able to visualize retinal and subretinal features accurately. This study compared two methods of assessing retinal volume using HD-OCT scans in healthy retinas. Both methods were able to provide realistic volumetric data when applied to raster scan sets. Manual segmentation methods represent an adequate tool with which to control automated processes and to identify clinically relevant structures, whereas automatic procedures will be needed to obtain data in larger patient populations.

Foveal Structure and Thickness of Retinal Layers Long-Term after Surgical Peeling of Idiopathic Epiretinal Membrane

To better understand the long-term effect of idiopathic epiretinal membrane peeling on retinal anatomy, the foveal structure and the thickness of individual retinal layers were analyzed with frequency-domain optical coherence tomography (fdOCT). The long-term postoperative course of macular thickness was followed. fdOCT scans were obtained from the horizontal midline in 33 eyes long-term (mean 46 ± 13 months) after surgery and in 30 eyes of age-matched controls. Raw images were exported, and the thickness of retinal layers was measured with a manual segmentation procedure aided by a customized computer program. Macular thickness was quantified over time with the time-domain (td) OCT Fast Macular Thickness program. Thickness of retinal layers between the outer nuclear and the ganglion cell plus inner plexiform layers in the horizontal midline of the fovea and the nasal parafovea was greater than normal, whereas that of the RPE, photoreceptor, and retinal nerve fiber layers was not different from controls. Twelve of 33 eyes had a foveal pit though the median foveal shape was distorted. Central macular thickness quantified with tdOCT remained increased, whereas the decrease of nasal macular thickness toward normal values was incomplete and delayed to 35 months after surgery. Superior, temporal, and inferior macular thickness returned to normal 12 to 14 months after surgery. Long-term after surgery, the fovea and the nasal parafovea remain thickened between the outer nuclear layer and the ganglion cell layer, whereas the superior, temporal, and inferior macular thickness returns to normal. Long-term observations are required in the assessment of macular recovery from mechanical stress.

In vivo characterization of morphological properties and contact areas of the rat cartilage derived from high-resolution MRI

Knee osteoarthritis deals with the loss of cartilaginous substances. Biomaterial implant using mesenchymal stromal cells was usually used to perform cartilage regeneration. The bio-integration and biofunctionality assessments of the used biomaterial implant on the injury sites require the longitudinal monitoring of the cartilage over times. The aim of this present study was to develop a characterization framework of the cartilage in knee osteoarthritis especially its morphological properties and contact areas derived from high-resolution magnetic resonance imaging (MRI) technique. Manual segmentation procedure was applied for extracting cartilage from surrounding biological tissues. Geometrical computing techniques were applied to calculate characterization properties of the rat cartilage. As clinical case study, curvature, thickness, volume and relative contact area properties of the rat cartilage in knee osteoarthritis were quantified on each knee full model and/or its region of interest (ROI) models. We found that the cartilage thickness and volume of three cartilage compartments decrease at the beginning (from 7 to 28 days). Tibial cartilage thickness increases at the end of the osteoarthritis process (from 28 to 56 days). This is due to the fact that the cartilage swelling occurred during osteoarthritis process. These findings were confirmed when observing cartilage curvature; we noted that there is an increasing curvature due to the incurvation effect of the cartilage during osteoarthritis process. The relative contact areas of three cartilage compartments increase slightly. However, the difference between control knees and osteoarthritis knees is significant. ROIs analyses of the thickness and curvature properties were also reported on the femoral and tibial cartilages.

The Transition Zone between Healthy and Diseased Retina in Patients with Retinitis Pigmentosa

To describe the structural changes in the transition zone from relatively healthy retinal regions to severely affected regions in patients with retinitis pigmentosa (RP) using frequency domain optical coherence tomography (fdOCT). FdOCT line scans of the horizontal meridian were obtained from one eye of 13 patients with RP and 30 control subjects. The patients had normal or near normal foveal sensitivities and visual field diameters ≥10°. Using a computer-aided manual segmentation procedure, the locations at which the outer segment (OS) and outer nuclear layer plus outer plexiform layer (ONL+) thicknesses fell below the 95% confidence interval of the controls were measured, as were the locations at which the OS layer disappeared and the locations at which the ONL+ was reduced to an asymptotically small thickness. The progression from healthy to severely affected regions followed a common pattern in most patients. Region A, the central region including the foveal center, had normal OS and ONL+ thickness. Region B had abnormal OS but normal ONL+ thickness. Region C had abnormal but measurable OS and ONL+ thicknesses. In Region D, the OS layer disappeared, as did the IS/OS line, and the ONL+ thickness decreased further. In Region E, the ONL+ reached an asymptotic thickness. The structural changes in the transition zone followed an orderly progression from a thinning of the OS layer, to a thinning of the ONL+, to a loss of the OS layer, to an ONL+ reduced to an asymptotically small level.

Reliability of a Computer-Aided Manual Procedure for Segmenting Optical Coherence Tomography Scans

To assess the within- and between-operator agreement of a computer-aided manual segmentation procedure for frequency-domain optical coherence tomography scans. Four individuals (segmenters) used a computer-aided manual procedure to mark the borders defining the layers analyzed in glaucoma studies. After training, they segmented two sets of scans, an Assessment Set and a Test Set. Each set had scans from 10 patients with glaucoma and 10 healthy controls. Based on an analysis of the Assessment Set, a set of guidelines was written. The Test Set was segmented twice with a ≥1 month separation. Various measures were used to compare test and retest (within-segmenter) variability and between-segmenter variability including concordance correlations between layer borders and the mean across scans (n = 20) of the mean of absolute differences between local border locations of individual scans, MEAN{mean( ΔLBL )}. Within-segmenter reliability was good. The mean concordance correlations values for an individual segmenter and a particular border ranged from 0.999 ± 0.000 to 0.978 ± 0.084. The MEAN{mean( ΔLBL )} values ranged from 1.6 to 4.7 μm depending on border and segmenter. Similarly, between-segmenter agreement was good. The mean concordance correlations values for an individual segmenter and a particular border ranged from 0.999 ± 0.001 to 0.992 ± 0.023. The MEAN{mean( ΔLBL )} values ranged from 1.9 to 4.0 μm depending on border and segmenter. The signed and unsigned average positions were considerably smaller than the MEAN{mean( ΔLBL )} values for both within- and between-segmenter comparisons. Measures of within-segmenter variability were only slightly larger than those of between-segmenter variability. When human segmenters are trained, the within-and between-segmenter reliability of manual border segmentation is quite good. When expressed as a percentage of retinal layer thickness, the results suggest that manual segmentation provides a reliable measure of the thickness of layers typically measured in studies of glaucoma.

Arabic Speech Segmentation: Automatic Verses Manual Method and Zero Crossing Measurements

In this paper, a comparison between the automatic and manual approach of speech segmentation for the Arabic speech is conducted. In this approach, the automatic method, using the energy level measurement is compared to the manual segmentation procedure. The traditional zero crossing method commonly used for speech processing is also included in this work. The energy measurement method is based on dividing the uttered tokens into different levels. For instance, the Arabic language phonemes are divided into two energy regions: unvoiced phonemes which can be categorized as low energy include the sounds / س / (/s/) and / ه / (/h/). On the other hand, vowels and semi-vowels such as / / ( فتحه ) (/ ‘a /) and / و / (/w/) are labeled as high energy. Voiced fricatives, for instance, the sounds / ز / (/z/) and / /ع (/aa/) are classified as high energy phonemes.

Measurement of Local Retinal Ganglion Cell Layer Thickness in Patients With Glaucoma Using Frequency-Domain Optical Coherence Tomography

To explore the feasibility of obtaining a local measurement of the thickness of the retinal ganglion cell layer in patients with glaucoma using frequency-domain optical coherence tomography (fdOCT) and a computer-aided manual segmentation procedure. The fdOCT scans were obtained from the horizontal midline for 1 eye of 26 patients with glaucoma and 20 control subjects. The thickness of various layers was measured with a manual segmentation procedure aided by a computer program. The patients were divided into low- and high-sensitivity groups based on their foveal sensitivity on standard automated perimetry. The RGC plus inner plexiform and the retinal nerve fiber layers of the low-sensitivity group were significantly thinner than those of the high-sensitivity group. While these layers were thinner in the patients than the controls, the thicknesses of inner nuclear layer and receptor layer were similar in all 3 groups. Further, the thinning of the retinal ganglion cell plus inner plexiform layer in 1 glaucoma-affected eye showed qualitative correspondence to the loss in 10-2 visual field sensitivity. Local measures of RGC layer thickness can be obtained from fdOCT scans using a manual segmentation procedure, and these measures show qualitative agreement with visual field sensitivity.

Multiparameter Segmentation and Quantization of Brain Tumor from MRI Images

In this paper, a method for segmentation of brain tumor has been developed on 2D-MRI data which allows the identification (10-15 minutes operator time) of tumor tissue with accuracy and reproducibility comparable to manual segmentation (2-6 hours operator time) making the automatic segmentation practical reality for malignant tumors. In this scheme, after a manual segmentation procedure the tumor identification has been made for the potential use of MRI data for improving the approximate brain tumor shape and 2D visualization for surgical planning.

Thickness of Receptor and Post-receptor Retinal Layers in Patients with Retinitis Pigmentosa Measured with Frequency-Domain Optical Coherence Tomography

To better understand the effects of retinitis pigmentosa (RP) on post-receptor anatomy, the thicknesses of the receptor, inner nuclear, retinal ganglion cell (RGC), and retinal nerve fiber layers (RNFL) were measured with frequency-domain optical coherence tomography (fdOCT). FdOCT scans were obtained from the horizontal midline in 30 patients with RP and 23 control subjects of comparable age. Raw images were exported and the thicknesses of photoreceptor/RPE, inner nuclear, RGC plus inner plexiform, and nerve fiber layers were measured with a manual segmentation procedure aided by a computer program. The RNFL thickness was also measured in 20 controls and 25 patients using circular peripapillary fdOCT scans. Results from controls were consistent with known anatomy. In patients with RP, the pattern of photoreceptor loss with eccentricity was consistent with the field constriction characteristic of RP. INL and RGC layer measures were comparable to normal subjects, although some patients showed slightly thicker RGC layers. However, RNFL layer thickness was significantly greater than normal; a majority of patients showed a thicker RFNL on both horizontal midline scans and peripapillary scans. To make optimal use of OCT RNFL thickness as a measure of the integrity of RGCs in patients with RP, a better understanding of the causes of the thickening seen in the majority of the patients is needed. As the RGC layer thickness can be measured with fdOCT, RGC layer thickness may turn out to be a more direct and valid indicator of the presence of RGCs in patients with RP.