Font Size: a A A

Principle And Technique Of Photoacoustic And Ultrasonic Dual-Modality Imaging For Diagnosing Early Uterine Neoplasm

Posted on:2020-05-17Degree:DoctorType:Dissertation
Country:ChinaCandidate:Y P LinFull Text:PDF
GTID:1364330620452900Subject:Optical Engineering
Abstract/Summary:PDF Full Text Request
Photoacoustic(PA)and Ultrasonic(US)dual-modality imaging is a hybrid imaging,providing structural information by US imaging and functional information by PA imaging,will be more efficient to image and detect early uterine neoplasm.This dissertation presents a PA and US dual-modality imaging for real-time noninvasive diagnosis of early uterine neoplasm by employing internal pulse light irradiation through vagina and cervix.The US imaging system is outside the body and only the PA excitation device is inside the body.Some studies has been systematic carried out,including implementation of the PA and US dual-modality imaging system,simulation of light delivery in the uterine tissue,phantom experiment,pig uterus ex vivo experiment,PA signal enhancement and computer aided diagnosis system for early endometrial cancer,etc.Innovations in the main research include the following aspects.Firstly,PA and US dual-modality imaging system is presented by the present conditions of the lab.The hardware composition and software design of the system are described in detail.Two blood vessels phantom experiments are conducted in order to valid the feasibility and determine the resolution of the system.By transducer scanning,3D PA and US images are reconstructed which could clearly describe the tissue structure and benefit the detecting in clinical application.Secondly,in order to optimize selection of excitation source and compensation for PA signal of the PA and US dual-modality imaging system,a three dimensional optical model of the human uterus by two diffused light sources is built by the histological structure and light absorption distribution is study by employing Monte Carlo simulations.The results show that the laser with a wavelength of 800 nm is good for PA imaging detection of early uterine neoplasm.Material likes undiluted raw and homogenized milk which has an extremely small absorption coefficient and an extremely big scattering coefficient is helpful to the illumination of the uterus cavity.Furthermore,a 3D triangular meshes tumor-embedded uterine optical model is established to study the optical absorption energy of tumors in different places,which helps the determination of PA imaging accuracy and depth.Thirdly,PA and US dual-modality imaging experiments of uterine phantom are implemented by a cylindrical diffuser.The results show that cylindrical diffuser has ability to illuminate a larger volume of the uterine tissue.A better result is achieved by filling the uterine phantom cavity an extremely small absorption coefficient and an extremely big scattering coefficient material.Moreover,PA and US dual-modality imaging ex vivo experiments of pig uterus filled with Intralipid-20% are conducted.The position and size of the tumor are detected accurately.Fourthly,in order to improve the lateral resolution of the PA and US dual-modality imaging system filtered delay multiply and sum algorithm is described and coded.PA and US images in previous experiments are reconstructed again by filtered delay multiply and sum algorithm.The results show a better lateral resolution of the system and a better contrast of the images.Finally,a full implementation of computer aided diagnosis for early endometrial cancer based on PA and US dual-modality images is presented.To begin with,lots of early endometrial cancer PA and US dual-modality images are simulated according the endometrial cancer stage FIGO IA and IB.In addition,a classifier of early endometrial cancer which has a testing accuracy of 99.58% is built by multilayer convolutional neural network.Finally,the convolutional neural network classifier is deployed by using a microframework Flask.
Keywords/Search Tags:photoacoustic imaging technology, dual-modality imaging, uterine neoplasm, three dimensional imaging, internal excitation, diffused fiber, computer-aided diagnosis, Monte Carlo simulation
PDF Full Text Request
Related items