Four videokeratographic methods for keratoconus detection were compared with a neural network approach.
A classification neural network for keratoconus screening was designed to detect the presence ...of keratoconus (KC) or keratoconus suspects (KCS); a separate cone severity network graded the severity of conelike topography patterns consistent with KC or KCS. Three hundred TMS-1 examinations (Tomey) were randomly divided into training and test sets. Ten topographic indexes were network inputs. Nine categories were used: normal, astigmatism, KC, KCS, contact lens-induced warpage, pellucid marginal degeneration, photorefractive keratectomy, radial keratotomy, and penetrating keratoplasty. KC was subdivided into KC1 (mild), KC2 (moderate), and KC3 (advanced). There were three outputs for the classification network (KC, KCS, and OTHER); target output values of 0 = OTHER, 0.25 = KCS, 0.5 = KC1, 0.75 = KC2, and 1.0 = KC3 were used for the severity network.
The best-trained classification network had 100% accuracy, specificity, and sensitivity for the test set. The severity network had mean outputs (+/-standard deviation) of OTHER = 0.02+/-0.02, KCS = 0.21+/-0.05, KC1 = 0.52+/-0.17, KC2 = 0.74+/-0.12, and KC3 = 0.91+/-0.15. The severity network output for all categories was well correlated to the keratoconus prediction index (R = 0.892, P < 0.0001). The classification network had an overall accuracy and specificity significantly better (P < or = 0.005) than the Klyce/Maeda keratoconus index (KCI) test, the Rabinowitz test (K & I-S), and simulated keratometry (average Sim K). However, there were no significant differences in keratoconus sensitivity between the classification network, KCI, and K & I-S. The sensitivity and specificity of average Sim K were significantly worse than those of the other tests. The classification network had significantly better sensitivity (P < 0.001) and specificity (P = 0.025) for KCS detection than the K & I-S.
The neural networks completely distinguished KC from KCS and from topographies that resembled KC. The network approach equaled the sensitivity of currently used tests for keratoconus detection and outperformed them in terms of accuracy and specificity.
ABSTRACT We report radio SETI observations on a large number of known exoplanets and other nearby star systems using the Allen Telescope Array (ATA). Observations were made over about 19000 hr from ...2009 May to 2015 December. This search focused on narrowband radio signals from a set totaling 9293 stars, including 2015 exoplanet stars and Kepler objects of interest and an additional 65 whose planets may be close to their habitable zones. The ATA observations were made using multiple synthesized beams and an anticoincidence filter to help identify terrestrial radio interference. Stars were observed over frequencies from 1 to 9 GHz in multiple bands that avoid strong terrestrial communication frequencies. Data were processed in near-real time for narrowband (0.7-100 Hz) continuous and pulsed signals with transmitter/receiver relative accelerations from −0.3 to 0.3 m s−2. A total of 1.9 × 108 unique signals requiring immediate follow-up were detected in observations covering more than 8 × 106 star-MHz. We detected no persistent signals from extraterrestrial technology exceeding our frequency-dependent sensitivity threshold of 180-310 × 10−26 W m−2.
Although visual inspection of corneal topography maps by trained experts can be powerful, this method is inherently subjective. Quantitative classification methods that can detect and classify ...abnormal topographic patterns would be useful. An automated system was developed to differentiate keratoconus patterns from other conditions using computer-assisted videokeratoscopy.
This system combined a classification tree with a linear discriminant function derived from discriminant analysis of eight indices obtained from TMS-1 videokeratoscope data. One hundred corneas with a variety of diagnoses (keratoconus, normal, keratoplasty, epikeratophakia, excimer laser photorefractive keratectomy, radical keratotomy, contact lens-induced warpage, and others) were used for training, and a validation set of 100 additional corneas was used to evaluate the results.
In the training set, all 22 cases of clinically diagnosed keratoconus were detected with three-false-positive cases (sensitivity 100%, specificity 96%, and accuracy 97%). With the validation set, 25 out of 28 keratoconus cases were detected with one false-positive case, which was a transplanted cornea (sensitivity 89%, specificity 99%, and accuracy 96%).
This system can be used as a screening procedure to distinguish clinical keratoconus from other corneal topographies. This quantitative classification method may also aid in refining the clinical interpretation of topographic maps.
To determine the goodness-of-prediction of the fitting routine by measuring the difference between topographic corneal surfaces and their Zernike reconstructions as a function of polynomial order and ...optical zone size for various corneal conditions.
Corneal research laboratory in a university eye center.
Corneal topography maps (N = 253) were obtained from the Louisiana State University Eye Center. A variety of corneal conditions were used: normals; astigmatism; laser in situ keratomileusis, photorefractive keratoplasty (PRK), and radial keratotomy (RK) postoperative cases (myopic spherical corrections); keratoconus suspect; mild, moderate, and severe keratoconus; pellucid marginal degeneration; contact lens-induced corneal warpage; and penetrating keratoplasty. The root-mean-square (RMS) error of the goodness-of-prediction of the Zernike representation of corneal surface elevation was extracted for 4, 6, and 10 mm optical zones, whereas Zernike radial orders were varied from 3 to 14 in 1-order steps. The mean +/- SEM of the RMS error was plotted as a function of Zernike order and compared with criteria for normal surface fits.
Fitting accuracy improved as more Zernike terms were included, but some conditions showed significant errors (when compared with normal surfaces), even with many added terms. For a 6 mm optical zone, the normal cornea group had the lowest RMS error and did not require terms above the 4th order to achieve <0.25 microm RMS error. Astigmatism met the 0.25 microm threshold at the 5th order, whereas keratoconus suspect required 7 orders. Laser in situ keratomileusis and PRK met the 0.25 microm threshold at the 8th order, whereas RK required 10 orders. Contact lens-induced corneal warpage and mild keratoconus needed 12 orders to meet the 0.25 microm threshold, whereas pellucid marginal degeneration, moderate and severe keratoconus, and keratoplasty categories were not well fitted even at 14 orders.
A 4th-order Zernike polynomial appeared reliable for modeling the normal cornea, but using a 4th-order fitting routine with an abnormal corneal surface caused a loss of fine-detail shape information. As more Zernike terms were added, the accuracy of the fit improved, and the result approached the minimum error found with normal corneas. Unless sufficient higher-order Zernike terms are included when analyzing irregular surfaces, some diagnostic applications of Zernike coefficients may not be rigorous. This conclusion also suggests that wavefront shape analysis is similarly dependent on the number of orders used. Current surgical corrections based on normal-eye wavefronts may fail to capture all visually relevant aberrations in abnormal eyes, such as those having laser retreatments or experiencing corneal warpage from contact lens wear. A clinical goodness-of-fit or goodness-of-prediction index would indicate whether the number of terms in use has fully accounted for all of the visually significant aberrations present in the eye.
The purpose of this study was to determine whether a combination of vitamins B6, B9, and B12 is an effective intervention for reducing the signs and symptoms of nonproliferative diabetic retinopathy.
...Ten subjects with type 2 diabetes mellitus (n = 20 eyes) with clinically diagnosed mild to moderate nonproliferative diabetic retinopathy were recruited from a private practice ophthalmology clinic for this open-label, uncontrolled, prospective six-month study. Metanx® vitamin tablets (containing 3 mg L-methylfolate calcium, 35 mg pyridoxal-5'-phosphate, and 2 mg methylcobalamin) were administered at a dosage of two tablets daily. Primary outcome indicators were the percent change in mean retinal sensitivity threshold measured by macular microperimetry and the percent change in mean central retinal thickness measured by spectral-domain optical coherence tomography.
Three subjects were lost to follow-up. In the remaining seven subjects, two of 14 eyes had foveal edema that prevented microperimetry measurements due to poor fixation. The remaining 12 eyes showed a nonlinear improvement in mean threshold retinal sensitivity (P < 0.001). Overall change in mean central retinal thickness in 14 eyes was linear (R(2) = 0.625; P = 0.034), with a significant reduction between one and six months (P = 0.012).
In this pilot study, the Metanx intervention appeared to have some beneficial effects with respect to reducing retinal edema and increasing light sensitivity in subjects with nonproliferative diabetic retinopathy.
The air/tear interface contributes 70% of the vergence in the eye and, because of this, even minor variations in its shape can produce significant visual deficit. Placido disc-based corneal ...topographers measure the precise characteristics of the corneal surface, transforming shape into color-coded dioptric power maps and topography indexes related to optical quality and specific patterns associated with pathology. Artificial intelligence-based methods are used to classify corneal topography and these are used as screening tools. Coupling corneal topography measurements with aberrometry measurements permits the display of the internal aberrations of the eye. Together, these data provide the opportunity to extend refractive correction beyond sphere and cylinder to the higher order aberrations as well.
It is assumed that wavefront error data arising from aberrometry are adequately described by a Zernike polynomial function, although this assumption has not been extensively tested. Inaccuracies in ...wavefront error may compromise clinical testing and refractive correction procedures. The current retrospective study correlates visual acuity with corneal wavefront error and with the residual surface elevation error after fitting with the Zernike method.
Corneal topography maps were obtained from 32 keratoconus cases, 27 postoperative penetrating keratoplasty cases, and 29 postoperative conductive keratoplasty cases (88 total). The best spectacle-corrected visual acuity (BSCVA) for each case ranged from -0.2 to 1.3 logarithm of the minimum angle of resolution (logMAR) units (20/12.5-20/400). Topography was analyzed to determine wavefront error and the elevation fit error for a 4-mm optical zone. The 4th and 10th expansion series were analyzed with the 0th-order (piston) and 1st order (tip and tilt) removed. Linear regression analysis was performed. The difference in root mean square (RMS) error between the 4th- and 10th-order analyses was assessed for both wavefront and elevation fit error.
The correlation of BSCVA to wavefront error for 4th-order terms was moderately strong and significant (R2=0.581; P<0.001). The 10th-order correlation for wavefront error had a similar result (R2=0.565; P<0.001), but the regression was not significantly different from the 4th-order result. The correlation of BSCVA to the elevation fit error was strong and significant for the 4th order (R2=0.658; P<0.001). The 10th-order data had a similar result (R2=0.509; P<0.001), and there was no significant difference between the two regressions. Only 72% of the cases showed a shift toward increased wavefront error with the 10th-order series, whereas 18% lost wavefront error. All cases showed a shift toward improved elevation fit with the 10th-order expansion.
The wavefront error correlation to acuity was moderately strong, but the corneal elevation fit error also strongly correlated with visual acuity, indicating that Zernike polynomials do not fully characterize the surface shape features that influence vision and that exist in postsurgical or pathologic eyes. In addition, the change in wavefront error when using a larger expansion series was found to increase or diminish somewhat unpredictably. The authors conclude that Zernike polynomials fail to model all the information that influences visual acuity, which may confound clinical diagnosis and treatment.
Previously, human corneal stromal interlamellar cohesive strength in the horizontal meridian was shown to be twice as strong peripherally as centrally (approximately 2.90 x 10(-1) versus 1.40 x ...10(-1) N/mm). In the current study, stromal samples excised from the vertical meridian were studied to determine if meridional differences also exist. Precise knowledge about corneal stromal structure is warranted, because anisotropy can influence postoperative healing strength and corneal shape.
Limbus-to-limbus stromal strips with a constant 2.2-mm width were obtained from the vertical meridians of 52 eyebank corneas (8 single, 22 pairs). Cohesive strength was recorded as samples were split at a constant 1.6 mm/s at 50% stromal depth.
The mean cohesive strength in the inferior periphery was found to be only two thirds the strength observed in the nasal or temporal periphery (1.96 x 10(-1) N/mm versus approximately 2.94 x 10(-1) N/mm @ 5 mm). The mean cohesive strength in the inferior periphery was also significantly less than the strength of the superior periphery (1.85 x 10(-1) N/mm +/- 8.83 x 10(-3) SEM versus 2.34 x 10(-1) +/- 1.37 x 10(-2) @ 4 mm from the central cornea; P = 0.0027). Unlike the samples from the horizontal meridian, which could be described by a profile symmetric about the central cornea, force profiles for the vertical data could be characterized with distinct, classifiable patterns that were generally asymmetric. Fellow corneas from a single donor tended to have strength profiles belonging to the same class (P = 0.035; chi-square), although some paired eyes exhibited profiles from distinctly different classes.
These data strongly support the concept of an anisotropic collagen macrostructure that is more complex than previously believed. This inherent structural anisotropy may become a significant determinant of corneal shape during ectatic disease and some forms of keratorefractive surgery.
Zernike expansion has been selected for use in describing wavefront aberrations in the human eye. The advantages and limitations of this approach are assessed for eyes with varying degrees of ...aberration.
Corneal topography examinations were taken with the Nidek OPD-Scan topographer/aberrometer. These higher data density corneal topography examinations were converted to height data and subsequently to wavefront representations. System noise was evaluated with a 2D frequency analysis of 43-D test balls. Both Zernike polynomials and 2D Fourier transforms were used to evaluate fidelity in the presentation of the point spread function. A display format for potential clinical use was developed based upon Zernike decomposition.
Systematic noise from the corneal topographer was found to be minimal and, when eliminated, produced small changes in the point spread function. Using Zernike decomposition up to the 30th order failed to preserve the higher frequency aberrations present in aberrated eyes. Use of a Zernike decomposition display with a fixed micron scale presented only clinically significant details of spherical aberration, coma, trefoil, irregular components above third order and total higher-order aberrations (above second order).
Zernike polynomials excel in extracting the low-order optical characteristics of visual optics. Zernikes accurately represent both low- and high-order aberrations in normal eyes where high-order aberrations are clinically insignificant. For eyes after corneal surgery or eyes with corneal pathology such as keratoconus that have significant higher-order aberrations, the Zernike method fails to capture all clinically significant higher-order aberrations.
Videokeratography is a powerful tool for the diagnosis of corneal shape abnormalities. However, interpretation of the topographic map is sometimes difficult, especially when pathologies with similar ...topographic patterns are suspected. The neural networks model, an artificial intelligence approach, was applied for automated pattern interpretation in corneal topography, and its usefulness was assessed.
One hundred eighty-three topographic maps were selected and classified by human experts into seven categories: normal, with-the-rule astigmatism, keratoconus (mild, moderate, advanced), postphotorefractive keratectomy, and postkeratoplasty. The maps were divided into a training set (108 maps) and a test set (75 maps). For each map, 11 topography-characterizing indices calculated from the data provided by the TMS-1 videokeratoscope, plus the corresponding diagnosis category, were used to train a neural network.
The correct classification was achieved by a trained neural network for all 108 maps in the training set. In the test set, the neural network correctly classified 60 of 75 maps (80%). For every category, accuracy and specificity were greater than 90%, whereas sensitivity ranged from 44% to 100%.
With further testing and refinement, the neural networks paradigm for computer-assisted interpretation or objective classification of videokeratography may become a useful tool to aid the clinician in the diagnosis of corneal topographic abnormalities.
PDF
