scholarly journals Modeling Combined Ultrasound and Photoacoustic Imaging: Simulations aiding Device Development and Deep Learning

2020 ◽  
Author(s):  
Sumit Agrawal ◽  
Ajay Dangi ◽  
Sri-Rajasekhar Kothapalli
Keyword(s):  
Deep Learning ◽  
Photoacoustic Imaging ◽  
Design Parameters ◽  
Simulation Platform ◽  
Detector Arrays ◽  
Device Development ◽  
Effective Implementation ◽  
Optical Attenuation ◽  
Molecular Information ◽  
Imaging Performance

AbstractCombined ultrasound and photoacoustic (USPA) imaging has attracted several clinical applications due to its ability to simultaneously display structural and molecular information of deep biological tissue in real time. However, the depth dependent optical attenuation and the unknown optical and acoustic heterogeneities, limit the USPA imaging performance, especially from deeper tissue regions. Novel instrumentation, image reconstruction and deep learning methods are currently being explored to improve the USPA image quality. Effective implementation of these approaches requires a reliable USPA simulation tool capable of generating US based anatomical and PA based molecular information. Here, we developed a hybrid USPA simulation platform by integrating finite element models of light and ultrasound propagation. The feasibility of modeling US combined with optical fluence dependent multispectral PA imaging is demonstrated using in silico homogeneous and heterogeneous prostate tissue. The platform allows optimization of device design parameters, such as the aperture size and frequency of light source and ultrasound detector arrays. In addition, the potential of this simulation platform to generative massive USPA datasets aiding the data driven deep-learning enhanced USPA imaging has been validated using both simulations and experiments.

OPTIMIZATION OF DESIGN PARAMETERS FOR FLUORESCENCE LAMINAR OPTICAL TOMOGRAPHY

2011 ◽  
Vol 04 (03) ◽  
pp. 309-323 ◽  
Author(s):  
CHAO-WEI CHEN ◽  
YU CHEN
Keyword(s):  
Imaging Technique ◽  
Optical Tomography ◽  
Design Parameters ◽  
Optimization Strategy ◽  
Scattering Medium ◽  
Value Analysis ◽  
Molecular Information ◽  
Depth Sensitivity ◽  
Detector Configurations ◽  
Imaging Performance

Laminar optical tomography (LOT) is a mesoscopic tomographic imaging technique ranging between confocal microscopy and diffuse optical tomography (DOT). Fluorescence LOT (FLOT) provides depth-resolved molecular information with 100–200 μm resolution over 2–3 mm depth. In this study, we use Monte Carlo simulation and singular-value analysis (SVA) to optimize the source-detector configurations for potential enhancement of FLOT imaging performance. The effects of different design parameters, including source incidence and detector collection angles, detector number, and sampling density, are presented. The results indicate that angled incidence/detection configuration might improve the imaging resolution and depth sensitivity, especially for low-scattering medium. Increasing the number of detectors and the number of scanning steps will also result in enhanced imaging performance. We also demonstrate that the optimal imaging performance depends upon the background scattering coefficient. Our result might provide an optimization strategy for FLOT or LOT experimental setup.

The influence of confocal microscope design on imaging performance

1993 ◽  
Vol 51 ◽  
pp. 144-145
Author(s):  
C J R Sheppard
Keyword(s):  
Image Quality ◽  
Fluorescence Imaging ◽  
Confocal Microscope ◽  
Industrial Applications ◽  
Three Dimensions ◽  
Design Parameters ◽  
Weak Signals ◽  
Size And Shape ◽  
Imaging Performance ◽  
Fundamental Limitation

The confocal microscope is now widely used in both biomedical and industrial applications for imaging, in three dimensions, objects with appreciable depth. There are now a range of different microscopes on the market, which have adopted a variety of different designs. The aim of this paper is to explore the effects on imaging performance of design parameters including the method of scanning, the type of detector, and the size and shape of the confocal aperture.It is becoming apparent that there is no such thing as an ideal confocal microscope: all systems have limitations and the best compromise depends on what the microscope is used for and how it is used. The most important compromise at present is between image quality and speed of scanning, which is particularly apparent when imaging with very weak signals. If great speed is not of importance, then the fundamental limitation for fluorescence imaging is the detection of sufficient numbers of photons before the fluorochrome bleaches.

scholarly journals Review of deep learning for photoacoustic imaging

2021 ◽  
Vol 21 ◽  
pp. 100215
Author(s):  
Changchun Yang ◽  
Hengrong Lan ◽  
Feng Gao ◽  
Fei Gao
Keyword(s):  
Deep Learning ◽  
Photoacoustic Imaging

In vivo demonstration of reflection artifact reduction in LED-based photoacoustic imaging using deep learning

2021 ◽  
Author(s):  
Sumit Agrawal ◽  
Kerrick Johnstonbaugh ◽  
Thaarakh Suresh ◽  
Ankit Garikipati ◽  
Mithun Kuniyil Ajith Singh ◽  
...  
Keyword(s):  
Deep Learning ◽  
Photoacoustic Imaging ◽  
Artifact Reduction

Intelligent Control Emotion Mechanism Based on Artificial Life

2014 ◽  
Vol 494-495 ◽  
pp. 1220-1224
Author(s):  
Guo Feng Zhang
Keyword(s):  
Intelligent Control ◽  
Artificial Life ◽  
Control Algorithm ◽  
Inverted Pendulum ◽  
Design Parameters ◽  
Simulation Platform ◽  
Original Algorithm ◽  
Simulation Results ◽  
Improved Algorithm

The emotion mechanism being introduced to Human-Simulated Intelligent Control algorithm, a fresh Intelligent Control algorithm is proposed, to solve the parameters design and tuning difficulty of the original algorithm. The verification of proposed control algorithm is completed on the car-inverted pendulum simulation platform. After the simulation results of the original and its improved algorithm is compared with, the new control algorithm can be found to obtain better control results with less design parameters. Thus, the suggested idea confirms its feasibility and effectiveness preliminarily.

scholarly journals Photoacoustic 3D imaging detects potential microvascular injuries

2020 ◽  
Author(s):  
Yoshihiro Ishida ◽  
Yoshiaki Matsumoto ◽  
Yasufumi Asao ◽  
Aya Yoshikawa ◽  
Masako Kataoka ◽  
...  
Keyword(s):  
3D Imaging ◽  
Photoacoustic Imaging ◽  
Imaging Modality ◽  
Diagnostic Testing ◽  
Vascular Involvement ◽  
Hemoglobin Content ◽  
Vascular Change ◽  
Detector Arrays ◽  
Blood Clots ◽  
Written Informed Consent

AbstractThe photoacoustic imaging (PAI) system is an emerging imaging modality that can be useful in clinical diagnostic testing. Previous studies have shown utility of PAI in diagnosing breast cancer, visualizing vascular change in ageing, and planning flap reconstruction surgeries. In this study, we show that PAI can be used to non-invasively visualize microvascular injuries in smokers, and in a vasculitis patient.We used two prototype PAI systems, PAI-03, and PAI-04 in this study. These systems were equipped with hemisphere-detector arrays, and used lasers at wavelengths of 795 nm (PAI-03), 756 and 797nm (PAI-04). Healthy volunteers and a patient diagnosed with vasculitis were enrolled upon obtaining written informed consent. The whole hand of the volunteers was scanned.We noted there were unique circular structures that we coined “minute signals”. There were higher number of minute signals in smokers than in non-smokers. Enhanced minute signals were noted in the patient with vasculitis. In this patient, minute signals were suggested to have lower oxygenated hemoglobin content, likely suggesting blood clots or hemorrhage. Our study shows PAI systems can visualize unappreciated abnormalities in fine vasculatures non-invasively. Use of PAI can extend to other clinical entities with suspected vascular involvement.

scholarly journals A Study into Math Document Classification using Deep Learning

2020 ◽  
Author(s):  
Fatimah Alshamari ◽  
Abdou Youssef
Keyword(s):  
Deep Learning ◽  
Classification Performance ◽  
Document Classification ◽  
Design Parameters ◽  
Technological Advancement ◽  
Scientific Publications ◽  
Specific Domain ◽  
Decision Choices ◽  
And Mathematics

Document classification is a fundamental task for many applications, including document annotation, document understanding, and knowledge discovery. This is especially true in STEM fields where the growth rate of scientific publications is exponential, and where the need for document processing and understanding is essential to technological advancement. Classifying a new publication into a specific domain based on the content of the document is an expensive process in terms of cost and time. Therefore, there is a high demand for a reliable document classification system. In this paper, we focus on classification of mathematics documents, which consist of English text and mathematics formulas and symbols. The paper addresses two key questions. The first question is whether math-document classification performance is impacted by math expressions and symbols, either alone or in conjunction with the text contents of documents. Our investigations show that Text-Only embedding produces better classification results. The second question we address is the optimization of a deep learning (DL) model, the LSTM combined with one dimension CNN, for math document classification. We examine the model with several input representations, key design parameters and decision choices, and choices of the best input representation for math documents classification.

Sign in / Sign up

Export Citation Format

Share Document