Hyoun-Joong Kong

A finger-vein verification system using mean curvature

The finger-vein pattern is one of the human biometric signatures that can be used for personal verification. The first task of a verification process using finger-vein patterns is extracting the pattern from an infrared finger image. As a robust extraction method, we propose the mean curvature method, which views the vein image as a geometric shape and finds the valley-like structures with negative mean curvatures. When the matched pixel ratio is used in matching vein patterns, experimental results show that, while maintaining low complexity, the proposed method achieves 0.25% equal error rate, which is significantly lower than what existing methods can achieve.

Segmentation of Optic Nerve Head using Warping and RANSAC

Glaucoma is the second leading cause of blindness, and the retinal nerve fiber layer (RNFL) defect is the early sign of the glaucomatous optic nerve damage. To evaluate the RNFL, segmentation of the optic nerve head on the RNFL photograph should be the first step. This paper presents segmentation of optic nerve head using warping and random sample consensus (RANSAC). Sensitivity and positive predictability of the proposed method were 91% and 78% respectively.

Heart Rate Tracking Using a Doppler Radar With the Reassigned Joint Time-Frequency Transform

Heart rate variations are tracked using an RF Doppler signal by applying a reassigned joint time-frequency transform (RJTFT). In time-frequency analysis, RJTFT improves the readability of the heart rate on a spectrogram and the heart rate is continuously tracked with it. To verify the result, a heartbeat signal was acquired from a stationary human subject using a Doppler radar unit. The sensor operating at 2.45 GHz was located a distance of 0.5 m from the subject. Using the RJTFT, the heart rate was clearly extracted, thereby improved the readability. A reference heart rate was measured by a photoplethysmography (PPG) device and compared with the heart rate tracked by the RJTFT. The resulting root mean square (RMS) difference of the heart rate measured by the two methods was 5%.

Three dimensional reconstruction of conventional stereo optic disc image

Stereo disc photograph was analyzed and reconstructed as 3 dimensional contour image to evaluate the status of the optic nerve head for the early detection of glaucoma and the evaluation of the efficacy of treatment. Stepwise preprocessing was introduced to detect the edge of the optic nerve head and retinal vessels and reduce noises. Paired images were registered by power cepstrum method and zero-mean normalized cross-correlation. After Gaussian blurring, median filter application and disparity pair searching, depth information in the 3 dimensionally reconstructed image was calculated by the simple triangulation formula. Calculated depth maps were smoothed through cubic B-spline interpolation and retinal vessels were visualized more clearly by adding reference image. Resulted 3 dimensional contour image showed optic cups, retinal vessels and the notching of the neural rim of the optic disc clearly and intuitively, helping physicians in understanding and interpreting the stereo disc photograph.

Development and evaluation of an automatic acne lesion detection program using digital image processing

Existing acne grading methods, which depend on overall impression, require a long training period and there is a high degree of variability among raters, including trained dermatologists. The use of lesion count provides fair reproducibility but the method is time consuming. New technologies in photographic equipment and software allow solutions to the problem of acne evaluation. This study was conducted to develop the automatic acne lesion program and evaluation of its usefulness.

Registration of finger vein image using skin surface information for authentication

The finger vein image acquired with an acquisition system should be properly aligned to proceed with comparing algorithm. However it is not easy to find control the points since the images are naturally blurred with an inherent scattering property. To overcome this problem, we propose a novel finger vein registration method utilizing skin surface information (i.e. wrinkles and outlines). We assumed that finger crooking was insignificant. Images were sampled with intended translation and rotation. Each time, two images were acquired successively by switching the light source; one with infrared light and the other with white light. Degree of rotation and translation of sampled image were calculated using outline features in the white light image and then the infrared image was transformed according to the calculated data. To validate our method, correlation values were computed between identical subjects and different subjects. High correlation values were shown between identical subjects whereas low values were shown between different subjects.

Non-contact finger vein acquisition system using NIR laser

Authentication using finger vein pattern has substantial advantage than other biometrics. Because human vein patterns are hidden inside the skin and tissue, it is hard to forge vein structure. But conventional system using NIR LED array has two drawbacks. First, direct contact with LED array raise sanitary problem. Second, because of discreteness of LEDs, non-uniform illumination exists. We propose non-contact finger vein acquisition system using NIR laser and Laser line generator lens. Laser line generator lens makes evenly distributed line laser from focused laser light. Line laser is aimed on the finger longitudinally. NIR camera was used for image acquisition. 200 index finger vein images from 20 candidates are collected. Same finger vein pattern extraction algorithm was used to evaluate two sets of images. Acquired images from proposed non-contact system do not show any non-uniform illumination in contrary with conventional system. Also results of matching are comparable to conventional system. We developed Non-contact finger vein acquisition system. It can prevent potential cross contamination of skin diseases. Also the system can produce uniformly illuminated images unlike conventional system. With the benefit of non-contact, proposed system shows almost equivalent performance compared with conventional system.

Segmentation of Solid Nodules in Ultrasonographic Breast Image Based on Wavelet Transform

An accurate segmentation of solid nodules in ultrasonographic (US) breast image is presented. 1-level 2-dimensional Discrete Wavelet Transform (DWT) is used to create features reflecting the texture information of the original image. Using these features, the texture classification is achieved. Finally, solid nodule region is segmented from the classified texture region. Proper threshold for texture classification is automatically decided. Empirically acquired information about the relationship between the texture characteristic of the original image and the optimal threshold is examined and used. Presented algorithm is applied to 284 malignant solid nodules and 300 benign solid nodules and the resulting images are presented.

Novel Natural Fill Telemetric Pressure Flow Study of Discomfort and Bladder Outlet Obstruction

PURPOSE We evaluated the accuracy of natural fill telemetric pressure flow study performed in a private room, and assessed patient discomfort and experience after the procedure compared to those of standard pressure flow study. MATERIALS AND METHODS In 58 patients with lower urinary tract symptoms/benign prostatic hyperplasia free uroflowmetry, and natural fill telemetric and standard pressure flow studies were prospectively performed. Immediately after each step patients were asked to rate the experience in terms of pain, embarrassment, bother, boredom and repeat testing. Subjective items and objective urodynamic parameters were compared among the 3 tests. RESULTS Maximum urine flow on natural fill telemetric pressure flow study was not different from that on free uroflowmetry. In contrast, maximum flow, detrusor pressure at maximum flow and bladder contractility index on the standard pressure flow study were significantly lower than on the natural fill pressure flow study. However, the bladder outlet obstruction index was not significantly different between the 2 studies, and the bladder outlet obstruction and bladder contractility indexes correlated well. There were also no differences in bother and embarrassment scores. However, natural fill telemetric pressure flow study was superior in terms of pain and boredom scores. Patients were more willing to undergo repeat natural fill telemetric pressure flow study than standard pressure flow study. Time to complete the test was significantly shorter for the natural fill study than for the standard study. CONCLUSIONS Flow rate differences between pressure flow studies and free uroflowmetry are not due to mechanical obstruction by the catheter but to other factors, such as the bladder filling method. Objective parameters on the natural fill telemetric pressure flow study correlated with standard pressure flow study findings. The natural fill telemetric pressure flow study shortened the time needed for the test and may decrease the discomfort of the standard pressure flow study.

Development of Wireless Urodynamic Pressure-Flow Study System

This paper presents a natural-filling wireless pressure-flow study (NFW-PFS) system for the diagnosis of urologic problems which many patients in older population suffer from. For the benign prostatic hyperplasia (BPH) patients, NFW-PFS was performed in a private room. Several features were extracted from the pressure and flow data and used to make the artificial neural network classifier which was able to predict the status of BPH patient’s bladder outlet. Results showed good agreement with theoretical predictions based on clinical evidence and significant improvement over conventional urodynamic study (UDS) system in terms of several subjective parameters. The work presented here has profound implications for future studies of wearable UDS system and may one day help solve the problem of urodynamic application to home healthcare.