In vivo navigation of capsule endoscopy

Capsule endoscopy is like Sun Wukong drilling into the belly of Princess Tiefan in his travels to the west. It can show the intestines and stomach at a glance. Modern capsule endoscopy is a very popular research direction in the diagnosis of digestive tract diseases. It has the advantages of convenient examination, no trauma, no lead, no pain, no cross infection and no influence on the normal work of patients. It expands the field of vision of digestive tract examination and overcomes the shortcomings of traditional insertion endoscopy, such as poor tolerance, not suitable for the elderly, infirmity and critical illness. It can be used as the first choice for diagnosis of digestive tract diseases, especially small intestinal diseases. The principle is that the examinee can take the capsule of the built-in camera and signal transmission device orally, make it move in the digestive tract and take pictures by means of the peristalsis of the digestive tract. The doctor can use the image recorder and image workstation in vitro to understand the whole digestive tract of the examinee, so as to make a diagnosis of the patient's condition. Pixel number is an important performance parameter of optical imaging surface sensor products, which represents the level of product imaging resolution performance. For medical products, such as capsule endoscopy, which integrates optical systems into an integral and non-removable system, the number of pixels often implies the quality of optical resolution. Ideal design is matched one by one according to the scale of pixel unit and optical resolution unit. As the optical system with aberration correction can use numerical aperture to represent the optical resolution level, the number of pixels can represent the optical resolution performance of the product. However, in order to prevent the Moire effect when the optical resolution is higher than Nyquist frequency, the optical resolution based on the center of field of view is mostly lower than Nyquist frequency. Due to the lack of testing and evaluation methods and the lack of understanding of the matching level of the full field of view, the matching level of the field of view center has to be further reduced due to the phenomenon of Moire fringes caused by over-matching of the local field of view, resulting in a significant decline in the resolution of the number of pixels, and the efficiency of the actual number of pixels is extremely low. At present, the resolution of products with the same number of pixels on the market is quite different. Therefore, it is meaningful to pay attention to the matching design level between the number of pixels and the optical resolution of full field of view, and to evaluate the resolution effectiveness of the number of pixels.

    Jia Xiaohang's research team started with capsule endoscopy and devoted to the research of matching level and validity evaluation method of full field of view with pixel number. This method selects the horizontal optical axis section as the meridian plane to analyze. The section passing through the planar array sensor is row or column scanned. The analytical method of the optical resolution angle continuation arrangement was adopted. Then the algorithm of line resolution element number was derived. The number of effective pixels of the line was described by the number of resolution elements. Then, the effective rate of line pixel number was obtained. The pixel element projection angle and the optical resolution angle were taken as the analytical units. The ratio between the pixel element projection angle and the optical resolution angle in the same field of view position was regarded as the line-matching ratio. Three basic functions for analysis were further extended, including the resolution angle density function, the relative function of line matching ratio and the density eigenfunction of pixel element projection angle. A simple measurement method was derived by using the simple conversion relations between a line segment and an included angle on a spherical field of view. Those three basic functions for analysis were obtained by fitting, so that the situation of the pixel element and the optical resolution element in every position, the matching details between two elements and the matching level between two elements in the full field of view were obtained. The parameters, including the line resolution element number, effective rate of line pixel number, center field of view matching ratio, maximum matching ratio of the full field of view and its position and other parameters, not only constitute the evaluation basis for matching ratio and resolution validity of the pixel number of the capsule endoscope, but also offer a reference for product design, analysis and modification.

The function curve of manufacturer A 's product

About author
Jia Xiaohang has been engaged in the research of optical performance testing, evaluation methods and standards of medical optical products for a long time. For endoscope products with large field of view and depth of field, the author has done a lot of research work on the optoelectronic performance of the combination of clinical application scenarios and photoelectric array sensor imaging. The Z-sphere mathematical model, which can simulate all kinds of non-conduit clinical field of view, and the field conversion method based on the analysis of main ray tracing, were first developed by the author. The relative distortion of unit in the clinical field of view of endoscope was solved, including the difficulty of measuring and evaluating the edge light effect of endoscope self-illumination, imaging light effect and field of view light effect. A method of color reduction test and evaluation based on the color rendering of virtual light source and a method of measuring the efficiency of light energy transfer based on the concept of surface illumination transfer are developed. The evaluation theory of endoscope optical technology system is constructed in combination with the depth of field and other research contents. The results were applied to YY0068.1-2008 Medical Endoscope Hard Endoscopy Part 1: Industrial Standards for Optical Performance and Testing Methods, which won the "China Standards Innovation Contribution Award" in 2014. In recent years, Jia Xiaohang has guided his research team to complete the measurement methods of luminance response, signal-to-noise ratio, imaging SFR (MTF) and tolerance of endoscope optoelectronic performance. The results have been applied to YY/T 1603-2018 camera system industry standard and YY/T 1587-2018 electronic endoscope industry standard.


Jia Xiaohang, Zhang Qinyuan, Yan Qinglai, et al. The study of the methods for evaluating the matching rate of capsule endoscope pixels and its resolution effectiveness[J]. Opto-Electronic Engineering, 2019, 46(1): 180112.
DOI: 10.12086/oee.2019.180112