Simulation and assessment of realistic breast lesions using fractal growth models

Rashidnasab, Alaleh; Elangovan, Premkumar; Yip, Mary; Díaz, Oliver; Dance, David R.; Young, Kenneth C.; Wells, Kevin

doi:10.1088/0031-9155/58/16/5613

Cited by 48 publications

(57 citation statements)

References 25 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Figure 1 shows examples of spiculated and non-spiculated lesions. In our previous work 3, 4 , we simulated and validated non-spiculated masses using the method known as DLA (Diffusion Limited Aggregation). This is a fractal-like growth process to generate 3D irregular lesion-like structures.…”

Section: Introductionmentioning

confidence: 99%

Simulation of spiculated breast lesions

Elangovan

Alrehily

et al. 2016

SPIE Proceedings

View full text Add to dashboard Cite

Virtual clinical trials are a promising new approach increasingly used for the evaluation and comparison of breast imaging modalities. A key component in such an assessment paradigm is the use of simulated pathology, in particular, simulation of lesions. Breast mass lesions can be generally classified into two categories based on their appearance; nonspiculated masses and spiculated masses. In our previous work, we have successfully simulated non-spiculated masses using a fractal growth process known as diffusion limited aggregation. In this new work, we have extended the DLA model to simulate spiculated lesions by using features extracted from patient DBT images containing spiculated lesions. The features extracted included spicule length, width, curvature and distribution. This information was used to simulate realistic looking spicules which were attached to the surface of a DLA mass to produce a spiculated mass. A batch of simulated spiculated masses was inserted into normal patient images and presented to an experienced radiologist for review. The study yielded promising results with the radiologist rating 60% of simulated lesions in 2D and 50% of simulated lesions in DBT as realistic.

show abstract

Section: Introductionmentioning

confidence: 99%

Simulation of spiculated breast lesions

Elangovan

Alrehily

et al. 2016

SPIE Proceedings

View full text Add to dashboard Cite

show abstract

“…The dataset consisted of 630 images of background tissue only and 630 images containing uniform spheres and simulated lesions [12] of 4 and 6 mm diameter embedded in a validated breast phantom. Local glandularity texture was used to control contrast.…”

Section: Training and Testing Using Simulated Data At Minimum Detectamentioning

confidence: 99%

A deep learning model observer for use in alterative forced choice virtual clinical trials

Alnowami

Mills

Awis

et al. 2018

Medical Imaging 2018: Image Perception, Observer Performance, and Technology Assessment

View full text Add to dashboard Cite

Virtual clinical trials (VCTs) represent an alternative assessment paradigm that overcomes issues of dose, high cost and delay encountered in conventional clinical trials for breast cancer screening. However, to fully utilize the potential benefits of VCTs requires a machine-based observer that can rapidly and realistically process large numbers of experimental conditions. To address this, a Deep Learning Model Observer (DLMO) was developed and trained to identify lesion targets from normal tissue in small (200 x 200 pixel) image segments, as used in Alternative Forced Choice (AFC) studies. The proposed network consists of 5 convolutional layers with 2x2 kernels and ReLU (Rectified Linear Unit) activations, followed by max pooling with size equal to the size of the final feature maps and three dense layers. The class outputs weights from the final fully connected dense layer are used to consider sets of n images in an n-AFC paradigm to determine the image most likely to contain a target. To examine the DLMO performance on clinical data, a training set of 2814 normal and 2814 biopsy-confirmed malignant mass targets were used. This produced a sensitivity of 0.90 and a specificity of 0.92 when presented with a test data set of 800 previously unseen clinical images. To examine the DLMOs minimum detectable contrast, a second dataset of 630 simulated backgrounds and 630 images with simulated lesion and spherical targets (4mm and 6mm diameter), produced contrast thresholds equivalent to/better than human observer performance for spherical targets, and comparable (12 % difference) for lesion targets.

show abstract

“…This breast phantom included five breast tissues, corresponding to (1) adipose tissue, (2) 25% glandular tissue, (3) 50% glandular tissue, (4) 75% glandular tissue and (5) skin. Furthermore, three realistic breast lesions simulated with a procedure developed at the University of Surrey [9] (see Fig. 4(b)) were inserted into the anthropomorphic breast phantom at different locations, corresponding to different breast tissue regions, as shown in Fig.…”

Section: Sample Geometrymentioning

confidence: 99%

“…where Φ( ) represents the photon fluence per energy interval, ( ) the linear attenuation coefficients (taken from (a) Anthropomorphic breast phantom [8] (b) Breast lesion [9] (c) Phantom slice after insertion NIST [12]) and illustrates the thickness of the detector (1 mm for this case). Furthermore, the energy resolution ( ) was modelled using a Gaussian distribution [13].…”

Section: Sample Geometrymentioning

confidence: 99%

Breast CT image simulation framework for optimisation of lesion visualisation

Díaz

Elangovan

Enshaeifar

et al. 2013

2013 IEEE Nuclear Science Symposium and Medical Imaging Conference (2013 NSS/MIC)

View full text Add to dashboard Cite

Abstract-Although X-ray mammography is the gold standard technique for breast cancer detection, it suffers from limitations due to tissue superposition which could either obscure or mimic a breast lesion. Dedicated breast computed-tomography (BrCT) represents an alternative technology with the potential to overcome these limitations. However, this technology is still under investigation in order to study and improve certain parameters (e.g. dose, scattered radiation, etc.). In this work, an image simulation framework is proposed to generate realistic BrCT images and spectral imaging analysis is explored to enhance the contrast of breast lesions. Results illustrated an improvement in contrast between 5 and 10% when the final image is reconstructed using X-ray photons with energies between 21 and 30 keV, in comparison with the reconstructed image from the polychromatic energy spectrum recorded within the image receptor.

show abstract

Simulation and assessment of realistic breast lesions using fractal growth models

Abstract: Abstract. A new method of generating realistic three dimensional simulated breastrealistic masses with more variability in shape compared to the RW method. DLA 24 generated lesions can overcome the lack of complexity in structure and shape in many 25 current methods of mass simulation.

Cited by 48 publications

References 25 publications

Simulation of spiculated breast lesions

Simulation of spiculated breast lesions

A deep learning model observer for use in alterative forced choice virtual clinical trials

Breast CT image simulation framework for optimisation of lesion visualisation

Contact Info

Product

Resources

About