Breast composition density has been identified to be a risk factor of developing breast cancer and an indicator of lesion diagnostic obstruction due to masking effect in x-ray mammography images. Volumetric density measurement evaluates fibro-glandular volume, breast volume, and breast volume density measures that have potential advantages over area density measurement in risk assessment. Compared to traditional x-ray absorption computing based areal and volumetric tissue density assessments, image feature detection based density classification approaches emulate the clinical density evaluation process by radiologists instead of using indirect information (e.g., percentage density values). We have modeled breast density assessment as a machine intelligence task which automatically extract the image features and dynamically improves density classification performance in clinical environment: (1) a bank of deep learning networks are explored to automatically extract the image features that emulate the radiologists’ image review process; (2) the pretrained networks are retrained with clinical 2D digital mammography images (for processing and for presentation DICOM images) using transfer learning; (3) a deep reinforcement network is incorporated through human-machine gaming process. The data preprocessing, trained models / processes have been described, and the classification inference have been evaluated with the predicted breast density category values of the clinical validation 2D digital mammographic images in terms of statistic measures. The experimental results have shown that the method is promising for breast density assessment.
Breast density has been identified to be a risk factor of developing breast cancer and an indicator of lesion diagnostic
obstruction due to masking effect. Volumetric density measurement evaluates fibro-glandular volume, breast volume,
and breast volume density measures that have potential advantages over area density measurement in risk assessment.
One class of volume density computing methods is based on the finding of the relative fibro-glandular tissue attenuation
with regards to the reference fat tissue, and the estimation of the effective x-ray tissue attenuation differences between
the fibro-glandular and fat tissue is key to volumetric breast density computing. We have modeled the effective
attenuation difference as a function of actual x-ray skin entrance spectrum, breast thickness, fibro-glandular tissue
thickness distribution, and detector efficiency. Compared to other approaches, our method has threefold advantages: (1)
avoids the system calibration-based creation of effective attenuation differences which may introduce tedious
calibrations for each imaging system and may not reflect the spectrum change and scatter induced overestimation or
underestimation of breast density; (2) obtains the system specific separate and differential attenuation values of fibroglandular
and fat for each mammographic image; and (3) further reduces the impact of breast thickness accuracy to
volumetric breast density. A quantitative breast volume phantom with a set of equivalent fibro-glandular thicknesses has
been used to evaluate the volume breast density measurement with the proposed method. The experimental results have
shown that the method has significantly improved the accuracy of estimating breast density.
Single Mo target, Mo / Rh, or Mo / W bi-track targets with corresponding Mo and Rh filters have provided optimal
target / filter combinations for traditional screen / film systems. In the advent of full-field digital mammography, similar
target / filter combinations were adopted directly for digital imaging systems with direct and indirect conversion based
detectors. To reduce the average glandular dose while maintaining the clinical image quality of FFDMs, alternative
target / filter combinations have been investigated extensively to take advantages of the digital detectors with high
dynamic range, high detection dose efficiency, and low noise level. This paper reports the development of a digital
FFDM system that is equipped with single tungsten target and rhodium and silver filters. A mathematical model was
constructed to quantitatively simulate x-ray spectra, breast compositions, contrast objects, x-ray scatter distribution, grid
performance, and characteristics of a-Se flat panel detector. Computer simulations were performed to select kV/filter for
different breast thickness and breast compositions through maximizing the contrast object detection dose efficiency. A
set of phantom experiments were employed to optimize the x-ray techniques within the constraints of exposure time and
required dose levels. A 50-micrometer rhodium filter was applied for thin and average breasts and a 50-micrometer
silver filter for thicker breasts. To meet our design requirements and EUREF protocol specifications, we finely adjusted
x-ray techniques for 0.45, 0.75, 1.0, 1.35 mGy dose modes with regards to ACR phantom scoring and PMMA phantom
SNR/CNR performance, respectively. The optimal x-ray techniques significantly reduce average glandular dose while
maintaining imaging performance.
Clinical studies have correlated a high breast density to a women's risk of breast cancer. A breast density measurement
that can quantitatively depict the volume distribution and percentage of dense tissues in breasts would be very useful for
risk factor assessment of breast cancer, and might be more predictive of risks than the common but subjective and
coarse 4-point BIRADS scale. This paper proposes to use a neural-network mapping to compute the breast density
information based upon system calibration data, x-ray techniques, and Full Field Digital Mammography (FFDM)
images. The mapping consists of four modules, namely, system calibration, generator of beam quality, generator of
normalized absorption, and a multi-layer feed-forward neural network. As the core of breast density mapping, the
network accepts x-ray target/filter combination, normalized x-ray absorption, pixel-wise breast thickness map, and x-ray
beam quality during image acquisition as input elements, and exports a pixel-wise breast density distribution and a
single breast density percentage for the imaged breast. Training and testing data sets for the design and verification of
the network were formulated from calibrated x-ray beam quality, imaging data with a step wedge phantom under a
variety x-ray imaging techniques, and nominal breast densities of tissue equivalent materials. The network was trained
using a Levenberg-Marquardt algorithm based back-propagation learning method. Various thickness and glandular
density phantom studies were performed with clinical x-ray techniques. Preliminary results showed that the neural
network mapping is promising in accurately computing glandular density distribution and breast density percentage.
Dual-energy contrast enhanced digital mammography (DE-CEDM), which is based upon the digital subtraction of low/high-energy image pairs acquired before/after the administration of contrast agents, may provide physicians physiologic and morphologic information of breast lesions and help characterize their probability of malignancy. This paper proposes to use only one pair of post-contrast low / high-energy images to obtain digitally subtracted dual-energy contrast-enhanced images with an optimal weighting factor deduced from simulated characteristics of the imaging chain. Based upon our previous CEDM framework, quantitative characteristics of the materials and imaging components in the x-ray imaging chain, including x-ray tube (tungsten) spectrum, filters, breast tissues / lesions, contrast agents (non-ionized iodine solution), and selenium detector, were systemically modeled. Using the base-material (polyethylene-PMMA) decomposition method based on entrance low / high-energy x-ray spectra and breast thickness, the optimal weighting factor was calculated to cancel the contrast between fatty and glandular tissues while enhancing the contrast of iodized lesions. By contrast, previous work determined the optimal weighting factor through either a calibration step or through acquisition of a pre-contrast low/high-energy image pair. Computer simulations were conducted to determine weighting factors, lesions' contrast signal values, and dose levels as functions of x-ray techniques and breast thicknesses. Phantom and clinical feasibility studies were performed on a modified Selenia full field digital mammography system to verify the proposed method and computer-simulated results. The resultant conclusions from the computer simulations and phantom/clinical feasibility studies will be used in the upcoming clinical study.
Contrast enhanced digital mammography (CEDM), which is based upon the analysis of a series of x-ray projection images acquired before/after the administration of contrast agents, may provide physicians critical physiologic and morphologic information of breast lesions to determine the malignancy of lesions. This paper proposes to combine the kinetic analysis (KA) of contrast agent uptake/washout process and the dual-energy (DE) contrast enhancement together to formulate a hybrid contrast enhanced breast-imaging framework. The quantitative characteristics of materials and imaging components in the x-ray imaging chain, including x-ray tube (tungsten) spectrum, filter, breast tissues/lesions, contrast agents (non-ionized iodine solution), and selenium detector, were systematically modeled. The contrast-noise-ration (CNR) of iodinated lesions and mean absorbed glandular dose were estimated mathematically. The x-ray techniques optimization was conducted through a series of computer simulations to find the optimal tube voltage, filter thickness, and exposure levels for various breast thicknesses, breast density, and detectable contrast agent concentration levels in terms of detection efficiency (CNR2/dose). A phantom study was performed on a modified Selenia full field digital mammography system to verify the simulated results. The dose level was comparable to the dose in diagnostic mode (less than 4 mGy for an average 4.2 cm compressed breast). The results from the computer simulations and phantom study are being used to optimize an ongoing clinical study.
Conventional screen film mammography is the most effective tool for the early detection of breast cancer currently available. However, conventional mammography has relatively low sensitivity for detecting small breast cancers (under several millimeters). Specificity and the positive predictive value of mammography remain limited owing to an overlap in the appearance of benign and malignant lesions, and surrounding structure. We propose to address the limitations accompanying conventional mammography by incorporating a cone beam volume CT (CBVCT) reconstruction technique with a recently developed flat panel detector. In this study, we present the results obtained from a computer simulation study and a breast-imaging phantom experimental study to find out the reconstruction accuracy of different cone beam volume scanning orbits for CBVCTBI, and to determine if different partial scan protocols (less than 360°) are appropriate for breast cancer detection. Three types of CBVCTBI scanning orbits were simulated using an uncompressed breast phantom. The reconstruction accuracy was evaluated as a function of different scanning orbits assuming reconstruction with 360° was the gold standard. The results indicate that with 180° plus cone angle orbit, the reconstruction error is below 4% that is in the acceptable range. In addition, a preliminary phantom study using both 360° and 180° plus cone angle orbits, was conducted on the current flat panel detector-based CBVCT prototype scanner. The companion CBVCT reconstruction images of an uncompressed breast-image phantom are presented.
Conventional film-screen mammography is the most effective tool for the early detection of breast cancer currently available. However, conventional mammography has relatively low sensitivity to detect small breast cancers (under several millimeters) owing to an overlap in the appearances of benign and malignant lesions, and surrounding structure. The limitations accompanying conventional mammography is to be addressed by incorporating a cone beam volume CT imaging technique with a recently developed flat panel detector. A computer simulation study has been performed to prove the feasibility of developing a flat panel detector-based cone beam volume CT breast imaging (FPD-CBVCTBI) technique. In this study, a phantom and specimen experiment is performed to confirm the findings in the computer simulation using the current prototype cone beam volume CT scanner. The results indicate that the CBVCTBI technique effectively removes structure overlap and significantly improves the detectability of small breast tumors. More importantly, the results also demonstrate the patient dose level required for FPD-based CBVCTBI to detect a small tumor (under 5 mm) and a small calcification is less than or equal to that of conventional mammography. The results from this study suggest that FPD-CBVCTBI is a potentially powerful breast-imaging tool.
This paper presents a wavelet analysis-based multi-resolution cone-beam volume CT breast imaging technique that is adaptive for high-resolution and ultra-high resolution reconstructions. Wavelet analysis-based de-noising techniques are employed to improve image quality and further reduce the required absorbed dose. The following steps can summarize this technique. First, in the high-resolution mode, the high spatial resolution projections are rebinned into lower resolution projections through a wavelet decomposition/synthesis procedure while in the ultra-high-resolution mode the original spatial resolution of the projection data is kept. Second, a wavelet analysis-based de-noising technique is applied upon the projection data with quantum fluctuations to suppress the noise level in the reconstructed images. Third, a de-noising method through an adaptation of a wavelet shrinkage approach for noise reduction is utilized in the reconstructed data to improve the image quality in terms of the signal-to-noise ratio and dose efficiency. The computer simulations show that the wavelet analysis-based multi-resolution rebinning approach provides the flexibility to adjust spatial the reconstruction resolution and noise level for various imaging tasks. Also, the wavelet analysis-based de-noising technique efficiently suppresses the quantum mottle induced noise, and contributes to a better low contrast object reconstruction in terms of the signal-to-noise ratio (SNR) improvement. In addition, the reconstruction of a high contrast object, for example, a tiny calcification grain, is obtained with less density spread. The noise level in the reconstructed image is reduced, which means the necessary dose level can be further reduced while the image quality is not compromised.
Cone beam reconstruction has attracted a great deal of attention in the medical imaging community. However, high-resolution cone beam reconstruction (CBR) involves a huge set of data and very time consuming computing. It usually needs customized hardware or a large-scale computer to achieve acceptable speed. Although the Feldkamp algorithm is an approximate CBR algorithm, it is a practical and efficient 3D reconstruction algorithm and is a basic component in several exact cone-beam reconstruction algorithms (CBRA). In this paper, we present a practical implementation for high-speed CBR on a commercially available PC based on hybrid computing (HC). We implement Feldkamp CBR with multi-level acceleration. We use HC utilizing single instruction multiple data (SIMD) and making execution units (EU) in the processor work effectively. We also utilize the multi-thread and fiber support on the operating system, which automatically enable the reconstruction parallelism in the multi-processor environment, and makes data I/O to the hard disk more effective. Memory and cache access optimization is done by properly data partition. This approach was tested on an Intel Pentium III 500Mhz computer and was compared to the traditional implementation. It decreases more than 75% the filtering time for 288 pieces projections, saves more than 60% of the reconstruction time for the 5123 cube, and maintains good precision with less than 0.08% average error. Our system is cost-effective and high-speed. An effective reconstruction engine can be built with a market-available Symmetric Multi-processor (SMP) computer. This is an easy and cheap upgrade and is compatible with newer PC processors.
The clinical goal of breast imaging is to detect tumor masses when they are as small as possible, preferably less than 10 mm in diameter. Conventional film-screen mammography is the most effective tool for the early detection of breast cancer currently available. However, conventional mammography has relatively low sensitivity to detect small breast cancers (under several millimeters). Specificity and the positive predictive value of mammography remain limited owing to an overlap in the appearance of benign and malignant lesions, and surrounding structure. The limitations accompanying conventional mammography is to be addressed by incorporating a cone beam volume CT reconstruction technique with a recently developed flat panel detector. A computer simulation study has been performed to prove the feasibility of developing a flat panel detector-based cone beam volume CT breast imaging (FPD-CBVCTBI) technique. In this study, a preliminary phantom experiment is conducted to verify the findings in the computer simulation using a prototype flat panel detector-based cone beam volume CT scanner. The results indicate that the FPD-CBVCTBI technique effectively removes structure overlap and significantly improves the detectability of small breast tumors. This suggests that FPD-CBVCTBI is a potentially powerful breast-imaging tool.
X-ray projection mammography, using a film/screen combination or digital techniques, has proven to be the most effective imaging modality for early detection of breast cancer currently available. However, the inherent superimposition of structures makes small carcinoma (a few millimeters in size) difficult to detect in the occultation case or in dense breasts, resulting in a high false positive biopsy rate. The cone-beam x-ray projection based volume imaging using flat panel detectors (FPDs) makes it possible to obtain three-dimensional breast images. This may benefit diagnosis of the structure and pattern of the lesion while eliminating hard compression of the breast. This paper presents a novel cone-beam volume CT mammographic imaging protocol based on the above techniques. Through computer simulation, the key issues of the system and imaging techniques, including the x-ray imaging geometry and corresponding reconstruction algorithms, x-ray characteristics of breast tissues, x-ray setting techniques, the absorbed dose estimation and the quantitative effect of x-ray scattering on image quality, are addressed. The preliminary simulation results support the proposed cone-beam volume CT mammographic imaging modality in respect to feasibility and practicability for mammography. The absorbed dose level is comparable to that of current two-view mammography and would not be a prominent problem for this imaging protocol. Compared to traditional mammography, the proposed imaging protocol with isotropic spatial resolution will potentially provide significantly better low contrast detectability of breast tumors and more accurate location of breast lesions.
The purpose of this study is to characterize a real time flat panel detector (FPD)-based imaging system for cone beam volume tomographic digital angiography (CBVTDA) applications. A prototype FPD-based imaging system has been designed and constructed on a modified GE 8800 CT scanner. This system is evaluated for CBVTDA using two phantoms. The system is first characterized in terms of linearity and dynamic range of the detector, the effect of image lag and scatter on the image quality, low contrast resolution and high contrast spatial resolution. The results indicate that the FPD-based imaging system can achieve 2lp/mm spatial resolution and provide appropriate low contrast resolution for intravenous CBVTDA angiography with clinically acceptable entrance exposure level.
Recent development of large area flat panel solid state detector arrays indicates that flat panel image sensors have some common potential advantages: compactness, absence of geometric distortion and veiling glare with the benefits of high resolution, high DQE, high frame rate and high dynamic range, small image lag (less than 1%) and excellent linearity (approximately 1%). The advantages of the new flat-panel detector make it a promising candidate for cone beam volume tomographic angiography imaging. The purpose of this study is to characterize a Selenium thin film transistor (STFT) flat panel detector-based imaging system for cone beam volume tomographic angiography imaging applications. A prototype STFT detector-based cone beam volume tomographic angiography imaging system has been designed and constructed based on the modification of a GE 8800 CT scanner. This system is evaluated using a vascular phantom with different x-ray spectra, different sizes of vessels and different iodine concentration levels. The results indicate that with the currently available STFT flat panel detector, 90 kVp is the optimal kVp to achieve the highest signal-to-noise ratio for volume tomographic angiography imaging and the low contrast resolution of the system is 4 mg/ml iodine for a 2 mm vessel.
The flat panel detector (FPD) has become a highly promising candidate for a wide variety of applications. A prototype selenium thin film transistor (STFT) array-based volume tomographic angiography (VTA) imaging system has been constructed for the feasibility study. This experimental set- up uses a 14' X 17' STFT detector with a 2560 X 3072 array of 14 bit pixels. While an STFT detector offers high resolution digital images, there will always be some defects on the detector. These defects will result in severe streaks and ring artifacts, which have been found in reconstructed images of preliminary phantom studies. It is obvious that the stationary noise sources of the FPD are enhanced by the reconstruction procedure. In this paper, an accurate and efficient FPD calibration method for the VTA imaging system is proposed to reduce the artifacts. An improved gain map and a bad pixel detection method with an adaptive threshold are introduced based on statistical models of the FPD. A more efficient localized and sensitive bad pixel detection ability is obtained by sub-dividing the detector array into sub- arrays, classifying bad pixels as different regional patterns, and then optimizing an interpolation scheme for each pattern. The real-time background correction, gain correction, bad- pixel correction, and methods to generate calibration maps are described in detail. The calibration technique is examined through phantom studies and evaluated by comparing the artifacts and noise in reconstructed images. Improvement of image quality is obtained utilizing the calibration technique. It has been clearly verified that the streaks and ring artifacts in reconstructed VTA images are significantly reduced. Finally, the advantages of our method and future works are also discussed.
This study presents a new intravenous (IV) tomographic angiography imaging technique, called intravenous volume tomographic digital angiography (VTDA) for cross sectional pulmonary angiography. While the advantages of IV-VTDA over spiral CT in terms of volume scanning time and resolution have been validated and reported in our previous papers for head and neck vascular imaging, the superiority of IV-VTDA over spiral CT for cross sectional pulmonary angiography has not been explored yet. The purpose of this study is to demonstrate the advantage of isotropic resolution of IV-VTDA in the x, y and z directions through phantom and animal studies, and to explore its clinical application for detecting clots in pulmonary angiography. A prototype image intensifier-based VTDA imaging system has been designed and constructed by modifying a GE 8800 CT scanner. This system was used for a series of phantom and dog studies. A pulmonary vascular phantom was designed and constructed. The phantom was scanned using the prototype VTDA system for direct 3D reconstruction. Then the same phantom was scanned using a GE CT/i spiral CT scanner using the routine pulmonary CT angiography protocols. IV contrast injection and volume scanning protocols were developed during the dog studies. Both VTDA reconstructed images and spiral CT images of the specially designed phantom were analyzed and compared. The detectability of simulated vessels and clots was assessed as the function of iodine concentration levels, oriented angles, and diameters of the vessels and clots. A set of 3D VTDA reconstruction images of dog pulmonary arteries was obtained with different IV injection rates and isotropic resolution in the x, y and z directions. The results of clot detection studies in dog pulmonary arteries have also been shown. This study presents a new tomographic IV angiography imaging technique for cross sectional pulmonary angiography. The results of phantom and animal studies indicate that IV-VTDA is superior to spiral CT for cross sectional pulmonary angiography.
Access to the requested content is limited to institutions that have purchased or subscribe to SPIE eBooks.
You are receiving this notice because your organization may not have SPIE eBooks access.*
*Shibboleth/Open Athens users─please
sign in
to access your institution's subscriptions.
To obtain this item, you may purchase the complete book in print or electronic format on
SPIE.org.
INSTITUTIONAL Select your institution to access the SPIE Digital Library.
PERSONAL Sign in with your SPIE account to access your personal subscriptions or to use specific features such as save to my library, sign up for alerts, save searches, etc.