锐化
探测器
血肿
材料科学
计算机科学
生物医学工程
偏移量(计算机科学)
计算机视觉
脑组织
人工智能
声学
医学
光学
放射科
物理
程序设计语言
作者
Yanyan Li,Hong Luo,Xinyan Fan,Yi Lv,Xiru Xie,Xin Luo,Ping Guo
标识
DOI:10.1088/1612-202x/ac458f
摘要
Abstract Near-infrared spectroscopy (NIRS) is a noninvasive, portable and low-cost technique, giving it great advantages for the early diagnosis of craniocerebral hematoma. More information on a hematoma can help medical staff to classify a patient’s condition in order to give more timely and effective treatment. Based on the principle of NIRS, in this paper a simulation of brain tissue is carried out in COMSOL finite element simulation software and brain tissue models are constructed of single-target hematoma, dual-target hematoma and hematomas at different depths. An array scanning single-source multi-detector is used to detect optical information and images are reconstructed and optimized by an imaging algorithm and the derivative peak sharpening method. The results show that a single-target hematoma can be accurately imaged, and the offset of the coordinate is within 1.4 mm,. The accuracy of dual-target hematoma imaging is limited, but derivative peak sharpening and multidirectional measurement can optimize the imaging effect. The array scanning detector can detect information about hematomas situated up to a depth of about 14 mm in the subcutaneous tissue.
科研通智能强力驱动
Strongly Powered by AbleSci AI