Performance Evaluation of Dorsal Vein Network of Hand Imaging Using Relative Total Variation-Based Regularization for Smoothing Technique in a Miniaturized Vein Imaging System: A Pilot Study

Abstract

Vein puncture is commonly used for blood sampling, and accurately locating the blood vessel is an important challenge in the field of diagnostic tests. Imaging systems based on near-infrared (NIR) light are widely used for accurate human vein puncture. In particular, segmentation of a region of interest using the obtained NIR image is an important field, and research for improving the image quality by removing noise and enhancing the image contrast is being widely conducted. In this paper, we propose an effective model in which the relative total variation (RTV) regularization algorithm and contrast-limited adaptive histogram equalization (CLAHE) are combined, whereby some major edge information can be better preserved. In our previous study, we developed a miniaturized NIR imaging system using light with a wavelength of 720–1100 nm. We evaluated the usefulness of the proposed algorithm by applying it to images acquired by the developed NIR imaging system. Compared with the conventional algorithm, when the proposed method was applied to the NIR image, the visual evaluation performance and quantitative evaluation performance were enhanced. In particular, when the proposed algorithm was applied, the coefficient of variation was improved by a factor of 15.77 compared with the basic image. The main advantages of our algorithm are the high noise reduction efficiency, which is beneficial for reducing the amount of undesirable information, and better contrast. In conclusion, the applicability and usefulness of the algorithm combining the RTV approach and CLAHE for NIR images were demonstrated, and the proposed model can achieve a high image quality.

Comparison of Gabor Filter Parameter Characteristics for Dorsal Hand Vein Authentication Using Artificial Neural Networks

“…Biometrik merupakan sebuah metode otentikasi dan menjadi sistem yang paling efektif untuk melakukan pengenalan Personal, karena biometrik memiliki karakteristik yang unik, tidak dapat dibagikan, tidak bisa dilupakan, dan tidak bisa hilang ataupun dirampas [6]. Teknologi saat ini telah menggunakan berbagai bentuk sistem biometrik diantaranya yaitu Pengenal Wajah [7], Sidik Jari [8], Iris Mata [9], Garis Telapak Tangan [10], Pembuluh Darah Punggung Tangan [11], untuk melakukan otentikasi pengenalan Personal, seperti sistem keamanan sidik jari dan pengenalan wajah pada smartphone. Dalam penggunaan otentikasi sebuah biometrik harus memenuhi beberapa persyaratan diantaranya yaitu universal setiap manusia harus memiliki biometrik tersebut, memiliki keunikan dan mampu membedakan banyak manusia, tidak bisa hilang atau pun berubah [12].…”

Design and Implementation of Embedded-Based Vein Image Processing System with Enhanced Denoising Capabilities

“…The time and physical cost associated with the intravenous angiography procedure are high. A vein image processing system for instantly determining the location of a vein was developed to address this issue [ 1 , 2 , 5 ]. Figure 1 shows NIR rays with wavelengths of 700–1000 nm characterized by a lower absorption rate for water, melanin, and hemoglobin.…”

“…Skin tissue and subcutaneous fat without veins appear brighter because NIR rays that pass through veins under the skin appear darker as they absorb NIR rays. Vein image processing technology improves the visibility of images by enhancing the contrast of dark vein components and strengthening outlines using an image processing algorithm with images considered at wavelengths in the range of the optical window [ 2 , 6 ].…”

“…In addition, the vein image processing algorithm used in the proposed study constructed the entire process by referring to the traditional vein image processing techniques [ 1 , 2 , 8 ] previously studied.…”

“…Medical accidents can arise because of such failures, resulting in human casualties and adverse effects, and there is a pressing need for a more precise angiography technique. Techniques for ensuring the success of injection procedures that apply an image processing algorithm to the vein image with a near-infrared (NIR) camera and use a vein projector for projecting it back onto the skin surface improve vein visibility, and related research is actively underway [ 1 , 2 ].…”