Neutron stimulated emission computed tomography (NSECT) is being developed as a non-invasive spectroscopic imaging technique to determine element concentrations in the human body. NSECT uses a beam of fast neutrons that scatter inelastically from atomic nuclei in tissue, causing them to emit characteristic gamma photons that are detected and identified using an energy-sensitive gamma detector. By measuring the energy and number of emitted gamma photons, the system can determine the elemental composition of the target tissue. Such determination is useful in detecting several disorders in the human body that are characterized by changes in element concentration, such as breast cancer. In this paper we describe our experimental implementation of a prototype NSECT system for the diagnosis of breast cancer and present experimental results from sensitivity studies using this prototype. Results are shown from three sets of samples: (a) excised breast tissue samples with unknown element concentrations, (b) a multi-element calibration sample used for sensitivity studies, and (c) a small-animal specimen, to demonstrate detection ability from in-vivo tissue. Preliminary results show that NSECT has the potential to detect elements in breast tissue. Several elements were identified common to both benign and malignant samples, which were confirmed through neutron activation analysis (NAA). Statistically significant differences were seen for peaks at energies corresponding to 37 Cl, 56 Fe, 58 Ni, 59 Co, 79 Br and 87 Rb. The spectrum from the small animal specimen showed the presence of 12 C from tissue, from bone, and elements 39 K, 27 Al, 37 Cl, 56 Fe, 68 Zn and 25 Mg. Threshold sensitivity for the four elements analyzed was found to range from 0.3 grams to 1 gram, which is higher than the microgram sensitivity required for cancer detection. Patient dose levels from NSECT were found to be comparable to those of screening mammography.

Dedicated breast computed tomography (CT) imaging possesses the potential for improved lesion detection over conventional mammograms, especially for women with dense breasts. The breast CT images are acquired with a glandular dose comparable to that of standard two‐view mammography for a single breast. Due to dose constraints, the reconstructed volume has a non‐negligible quantum noise when thin section CT slices are visualized. It is thus desirable to reduce noise in the reconstructed breast volume without loss of spatial resolution. In this study, partial diffusion equation (PDE) based denoising techniques specifically for breast CT were applied at different steps along the reconstruction process and it was found that denoising performed better when applied to the projection data rather than reconstructed data. Simulation results from the contrast detail phantom show that the PDE technique outperforms Wiener denoising as well as adaptive trimmed mean filter. The PDE technique increases its performance advantage relative to Wiener techniques when the photon fluence is reduced. With the PDE technique, the sensitivity for lesion detection using the contrast detail phantom drops by less than when the dose is cut down to of the two‐view mammography. For subjective evaluation, the PDE technique was applied to two human subject breast data sets acquired on a prototype breast CT system. The denoised images had appealing visual characteristics with much lower noise levels and improved tissue textures while maintaining sharpness of the original reconstructed volume.

This paper describes the implementation of neutron-stimulated emission computed tomography (NSECT) for non-invasive imaging and reconstruction of a multi-element phantom. The experimental apparatus and process for acquisition of multi-spectral projection data are described along with the reconstruction algorithm and images of the two elements in the phantom. Independent tomographic reconstruction of each element of the multi-element phantom was performed successfully. This reconstruction result is the first of its kind and provides encouraging proof of concept for proposed subsequent spectroscopic tomography of biological samples using NSECT.

Mutual information is a popular intensity-based image similarity measure mainly used in image registration. This measure has been also very successful as the similarity metric in our knowledge-based computer-assisted detection (CADe) system for the detection of masses in screening mammograms. Our CADe system is designed to assess a new, query case based on its similarity with known cases stored in the knowledge database. However, intensity-based mutual information captures only relationships between the gray level values of corresponding pixels. This study presents a novel advancement of our CADe system by incorporating neighborhood textural information when estimating the mutual information of two images. Specifically, an entropy filter is applied to the images, effectively replacing each image pixel value with its neighborhood entropy. This pixel-based entropy is a localized measure of image texture. Then, the information-theoretic CAD system is asked to make a decision regarding the query case using the texture-based mutual information similarity metric. The entropy-based image enhancement and MI-based decision making processes are repeated at different neighborhood scales. Finally, an artificial network merges intensity-based and texture-based decisions to investigate possible improvements in mass detection performance. Given a database of 1,820 regions of interest (ROIs) extracted from screening mammograms (901 depicting a biopsy-proven mass and 919 depicting normal parenchyma) and a leave-one out sampling scheme, the study showed that our CADe system achieves an ROC area of 0.87±0.01 using the intensity-based ROC. The ROC performance for the texture-based CADe system ranges from 0.69±0.01 to 0.83±0.01 depending on the scale of analysis. The synergistic approach of the ANN using both intensity-based and texture-based information resulted in statistically significantly better performance with an ROC area index of 0.93±0.01.

Previously, we developed a simple Laguerre–Gauss (LG) channelized Hotelling observer (CHO) for incorporation into our mass computer-aided detection (CAD) system. This LG-CHO was trained using initial detection suspicious region data and was empirically optimized for free parameters. For the study presented in this paper, we wish to create a more optimal mass detection observer based on a novel combination of LG channels. A large set of LG channels with differing free parameters was created. Each of these channels was applied to the suspicious regions, and an output test statistic was determined. A stepwise feature selection algorithm was used to determine which LG channels would combine best to detect masses. These channels were combined using a HO to create a single template for the mass CAD system. Results from free-response receiver operating characteristic curves demonstrated that the incorporation of the novel LG-CHO into the CAD system slightly improved performance in high-sensitivity regions.

Neutron Stimulated Emission Computed Tomography (NSECT) was evaluated as a potential technique for breast cancer diagnosis. NSECT can form a 3D tomographic image with an elemental (isotopic) spectrum provided at each reconstructed voxel. The target is illuminated (in vivo) by a neutron beam that scatters in-elastically producing characteristic gamma emission that is acquired tomographically with a spectrograph. Images are reconstructed of each element in the acquired spectrum. NSECT imaging was simulated for benign and malignant breast masses. A range of the number of incident neutrons was simulated from 19 million to 500k neutrons. Simulation included all known primary and secondary physical interactions in both the breast as well as in the spectrometer. Characteristic energy spectra were acquired by simulation and were analyzed for statistically significant differences between benign and malignant breasts. For 1 million incident neutrons, there were 61 differences in the spectra that were statistically significant (p < 0.05). Of these, 23 matched known characteristic emission from 6 elements that have been found in the breast (Br, Cs, K, Mn, Rb, Zn). The dose to two breasts was less than 3% of the dose of a 4 view screening mammogram. Increasing the dose to 52% of the mammogram (19 million neutrons) provided 89 significant spectral differences that matched 30 known emissions from 7 elements that have been found in the breast (Br, Co, Cs, K, Mn, Rb, Zn). Decreasing the dose to 1.4% (500K neutrons) eliminated all statistically significant matches to known elements. This study suggests that NSECT may be a viable technique for detecting human breast cancer in vivo at a reduced dose compared to 4 view screening mammography.

Multiplatform application of CAD systems in mammography is often limited due to image preprocessing steps that are tailored to the acquisition protocol such as the digitizer. The purpose of this study was to validate our knowledge-based CAD system across two different digitizers. Our system relies on the similarity of a query image with known cases stored in a knowledge database. Image similarity is assessed using information theory, without any image preprocessing. Therefore, we hypothesize that our CAD system can operate robustly across digitizers. We tested the hypothesis using two different datasets of mammographic regions of interest (ROIs) for mass detection. The two databases consisted of 1,820 and 1,809 ROIs extracted from DDSM mammograms digitized using a Lumisys and a Howtek scanner respectively. Three experiments were performed. First, we evaluated the CAD system on each dataset independently. Then, we evaluated the system on each dataset when the other dataset was used as the knowledge database. Finally, we assessed the CAD detection performance when the knowledge database contained mixed cases. Our CAD system had similar performance across digitizers (Az=0.87±0.01 for Lumisys vs. Az=0.8±0.01 for Howtek) when assessed independently. When the system was tested on one dataset while the other was used as the knowledge database, ROC performance declined marginally, mainly based on the partial ROC area index. This result suggests that blind translation of the system without some experience with cases digitized with the same digitizer is not recommended when the system is expected to operate at high sensitivity decision thresholds. When the system operated with a knowledge database of mixed cases, its performance across digitizers was robust yet slightly inferior to what observed independently.

Several studies have shown that the concentration of certain elements may be a disease indicator. We are developing a spectroscopic imaging technique, Neutron Stimulated Emission Computed Tomography (NSECT), to non-invasively measure and image elemental concentrations within the body. The region of interest is interrogated via a beam of high-energy neutrons that excite elemental nuclei through inelastic scatter. These excited nuclei then relax by emitting characteristic gamma radiation. Acquiring the gamma energy spectrum in a tomographic geometry allows reconstruction of elemental concentration images. Our previous studies have demonstrated NSECT's ability to obtain spectra and images of known elements and phantoms, as well as, initial interrogations of biological tissue. Here, we describe the results obtained from NSECT interrogation of a fixed mouse specimen. The specimen was interrogated via a 5MeV neutron beam for 9.3 hours in order to ensure reasonable counting statistics. The gamma energy spectrum was obtained using two High-Purity Germanium (HPGe) clover detectors. A background spectrum was obtained by interrogating a specimen container containing 50mL of 0.9% NaCl solution. Several elements of biological interest including 12C, 40Ca, 31P, and 39K were identified with greater then 90% confidence. This interrogation demonstrates the feasibility of NSECT interrogation of small animals. Interrogation with a commercial neutron source that provides higher neutron flux and lower energy (~2.5MeV) neutrons would reduce scanning time and eliminate background from certain elements.

A new imaging technique, neutron stimulated emission computed tomography (NSECT), is being developed that has potential for utilization in breast cancer imaging. NSECT is a spectroscopic imaging technique that is able to produce elemental concentration images and previous studies have identified differences in trace element concentrations between malignant and benign tissues. NSECT illuminates the body via a beam of neutrons causing elemental nuclei to become excited and emit characteristic gamma radiation. By imaging the gamma rays in a tomographic manner it is possible to reconstruct elemental composition images. This method requires high-resolution spectroscopy, thereby eliminating the use conventional scintillation gamma cameras; in this case, spectral information is obtained from high-purity germanium (HPGe) semiconductor detectors, providing only 1D spatial information. To obtain 2D elemental concentration images, we are adapting high-energy solar spectroscopy technology. A rotating modulation collimator (RMC) consisting of two parallel-slat collimators is placed in front of the detector and modulates the incoming signal in a manner predicted by its geometry. Reconstruction of 2D images is possible by counting the number incident gammas at each rotation angle. A significant challenge is presented when attempting to modify the RMC for use in the near field and a prototype camera has been constructed to verify the geometric validity of a RMC for this use. Herein we present the progress to date in the design and development of a high-energy spectroscopic gamma camera for use with NSECT imaging of the breast.

The underlying mechanism in projection radiography as well as in computed tomography (CT) is the accumulative attenuation of a pencil x-ray beam along a straight line. However, when a portion of photons is deviated from their original path by scattering, it is not valid to assume that these photons are the survival photons along the lines connecting the x-ray source and the individual locations where they are detected. Since these photons do not carry the correct spatial information, the final image is contaminated. Researchers are seeking techniques to reduce scattering, and hence, improve image quality, by scatter compensation. Previously, we presented a post-acquisition scatter compensation technique based on an underlying statistical model. We used the Poisson noise model, which assumed that the signals in the detector individually followed the Poisson process. Since most x-ray detectors are energy integrating rather than photon counting, the Poisson noise model can be improved by taking this property into account. In this study, we developed a Gaussian noise model by the matching-of-the-first-two-moments method. The Maximum Likelihood Estimator of the scatter-free image was derived via the expectation maximization (EM) technique. The maximum a posteriori estimate was also calculated. The Gaussian noise model was preliminarily evaluated on a full-field digital mammography system.

We use a contrast-detail observer study to compare performance of a novel 3D computed mammotomography (CmT) system with a commercially developed full-field digital mammography (FFDM) system. A contrast-detail phantom comprised of uniform acrylic spheres of various diameters was developed and placed in a variety of mediums including uniform water (simulating low contrast lesions within a uniform background), water and acrylic yarn (simulating low contrast lesions, with over/under-lying structure), oil only (simulating higher contrast lesions in a uniform background), and oil and acrylic yarn (simulating higher contrast lesions with over/under-lying structure). For CmT, the phantom was placed in a 14.6 cm diameter uncompressed breast phantom and projections acquired using a simple circular orbit, W-target tube, 60 kVp tube potential, 0.05 cm Ce filtration, 4 mAs per projection, and a CSI(T1) digital x-ray detector. Reconstructions used an iterative OSTR algorithm. For FFDM, the phantom was placed in a 5.3-cm-thick compressed breast phantom. Single CC-view mammograms were acquired using a clinical W-target tube with 50 um Rh filtration, 28 kVp, photo-timed mAs per our clinical mammography operation, and a Selenium-based flat-panel detector (Mammomat Novation, Siemens). Six observers evaluated the images in terms of the number of detectable spheres. FFDM performed significantly better for the low contrast lesions in uniform water background (p < 0.05). However, CmT performed significantly better for all other cases. (p < 0.05). Results indicate that CmT shows significant advantage in soft tissue detection over FFDM in otherwise low contrast dense breasts.

Neutron stimulated emission computed tomography (NSECT) is an imaging technique that provides an in-vivo tomographic spectroscopic image of the distribution of elements in a body. To achieve this, a neutron beam illuminates the body. Nuclei in the body along the path of the beam are stimulated by inelastic scattering of the neutrons in the beam and emit characteristic gamma photons whose unique energy identifies the element. The emitted gammas are collected in a spectrometer and form a projection intensity for each spectral line at the projection orientation of the neutron beam. Rotating and translating either the body or the beam will allow a tomographic projection set to be acquired. Images are reconstructed to represent the spatial distribution of elements in the body. Critical to this process is the appropriate removal of background gamma events from the spectrum. Here we demonstrate the equivalence of two background correction techniques and discuss the appropriate application of each.

Neutron stimulated emission computed tomography (NSECT) is being developed as a non-invasive technique to measure element concentration in in-vivo tissue at molecular levels. We have developed a system that performs this task using an incident neutron beam that scatters inelastically with an atomic nucleus causing it to emit a characteristic gamma photon. An energy-sensitive gamma detector is used to detect this energy and identify the target atom. Here we describe an experiment to determine the concentration of natural potassium (39K) in bovine liver without the need for a biopsy. A 5 MeV neutron beam was used to scan a known quantity of bovine liver to obtain a gamma spectrum showing element concentration in the liver. An aqueous KCl solution calibration sample was then scanned to establish a ratio of potassium concentration to gamma counts for the experimental setup. Counts from gamma peaks corresponding to excited states in 39K were summed and compared with counts from the known calibration sample, to give the concentration of 39K in the liver. A high purity germanium (HPGe) clover detector was used to measure the emitted gamma energy. The results were validated through neutron activation analysis (NAA) of the liver sample. The concentration of 39K reported by NSECT was found to be within 13% of the NAA result, clearly demonstrating the ability of NSECT for non-invasive quantification of element concentration in tissue.

Architectural distortion (AD) is a sign of malignancy often missed during mammographic interpretation. The purpose of this study was to explore the application of fractal analysis to the investigation of AD in screening mammograms. The study was performed using mammograms from the Digital Database for Screening Mammography (DDSM). The fractal dimension (FD) of mammographic regions of interest (ROIs) was calculated using the circular average power spectrum technique. Initially, the variability of the FD estimates depending on ROI location, mammographic view and breast side was studied on normal mammograms. Then, the estimated FD was evaluated using receiver operating characteristics (ROC) analysis to determine if it can discriminate ROIs depicting AD from those depicting normal breast parenchyma. The effect of several factors such as ROI size, image subsampling and breast density was studied in detail. Overall, the average FD of the normal ROIs was statistically significantly higher than that of the ROIs with AD. This result was consistent across all factors studied. For the studied set of implementation parameters, the best ROC performance achieved was 0.89 +/- 0.02. The generalizability of these conclusions across different digitizers was also demonstrated.

In this article, we present a characterization of the effect of difference of Gaussians (DoG) filters in the detection of mammographic regions. DoG filters have been used previously in mammographic mass computer-aided detection (CAD) systems. As DoG filters are constructed from the subtraction of two bivariate Gaussian distributions, they require the specification of three parameters: the size of the filter template and the standard deviations of the constituent Gaussians. The influence of these three parameters in the detection of mammographic masses has not been characterized. In this work, we aim to determine how the parameters affect (1) the physical descriptors of the detected regions, (2) the true and false positive rates, and (3) the classification performance of the individual descriptors. To this end, 30 DoG filters are created from the combination of three template sizes and four values for each of the Gaussians' standard deviations. The filters are used to detect regions in a study database of 181 craniocaudal-view mammograms extracted from the Digital Database for Screening Mammography. To describe the physical characteristics of the identified regions, morphological and textural features are extracted from each of the detected regions. Differences in the mean values of the features caused by altering the DoG parameters are examined through statistical and empirical comparisons. The parameters' effects on the true and false positive rate are determined by examining the mean malignant sensitivities and false positives per image (FPpI). Finally, the effect on the classification performance is described by examining the variation in FPpI at the point where 81% of the malignant masses in the study database are detected. Overall, the findings of the study indicate that increasing the standard deviations of the Gaussians used to construct a DoG filter results in a dramatic decrease in the number of regions identified at the expense of missing a small number of malignancies. The sharp reduction in the number of identified regions allowed the identification of textural differences between large and small mammographic regions. We find that the classification performances of the features that achieve the lowest average FPpI are influenced by all three of the parameters.

As more diagnostic testing options become available to physicians, it becomes more difficult to combine various types of medical information together in order to optimize the overall diagnosis. To improve diagnostic performance, here we introduce an approach to optimize a decision-fusion technique to combine heterogeneous information, such as from different modalities, feature categories, or institutions. For classifier comparison we used two performance metrics: The receiving operator characteristic (ROC) area under the curve [area under the ROC curve (AUC)] and the normalized partial area under the curve (pAUC). This study used four classifiers: Linear discriminant analysis (LDA), artificial neural network (ANN), and two variants of our decision-fusion technique, AUC-optimized (DF-A) and pAUC-optimized (DF-P) decision fusion. We applied each of these classifiers with 100-fold cross-validation to two heterogeneous breast cancer data sets: One of mass lesion features and a much more challenging one of microcalcification lesion features. For the calcification data set, DF-A outperformed the other classifiers in terms of AUC (p < 0.02) and achieved AUC=0.85 +/- 0.01. The DF-P surpassed the other classifiers in terms of pAUC (p < 0.01) and reached pAUC=0.38 +/- 0.02. For the mass data set, DF-A outperformed both the ANN and the LDA (p < 0.04) and achieved AUC=0.94 +/- 0.01. Although for this data set there were no statistically significant differences among the classifiers' pAUC values (pAUC=0.57 +/- 0.07 to 0.67 +/- 0.05, p > 0.10), the DF-P did significantly improve specificity versus the LDA at both 98% and 100% sensitivity (p < 0.04). In conclusion, decision fusion directly optimized clinically significant performance measures, such as AUC and pAUC, and sometimes outperformed two well-known machine-learning techniques when applied to two different breast cancer data sets.

Neutron Stimulated Emission Computed Tomography (NSECT) is being developed for in vivo measurement of the concentration and location of biologically relevant elements. NSECT is a spectroscopic imaging technique whereby the body is illuminated via high-energy neutrons that excite elemental nuclei that then relax through characteristic gamma radiation. This imaging technique requires high-resolution spectroscopy, thereby eliminating the use conventional scintillation gamma cameras. Consequently, high-purity germanium (HPGe) semiconductor detectors are utilized, providing no spatial information. To obtain 2D elemental concentration images, we are adapting high-energy solar spectroscopy technology. A rotating modulation collimator (RMC) consisting of two parallel-slat collimators is placed in front of the detector to modulate the incoming signal in a manner predicted by its geometry. Reconstruction of 2D images is feasible by counting the number of incident gammas at each rotation angle. The challenge is to identify a RMC geometry that allows this method to work in the near-field environment, which has far fewer assumptions and simplifications than the infinite focus of solar imaging. Herein we describe construction of a prototype RMC and experiments conducted with a radioactive Na point source. These experiments verified that the RMC modulates the signal in a manner consistent with its geometric and physical properties.

Certain elements (such as Fe, Cu, Zn, etc.) are vital to the body and an imbalance of such elements can either be a symptom or cause of certain pathologies. Neutron Stimulated Emission Computed Tomography (NSECT) is a spectroscopic imaging technique whereby the body is illuminated via a beam of neutrons causing elemental nuclei to become excited and emit characteristic gamma radiation. Acquiring the gamma energy spectra in a tomographic geometry allows reconstruction of elemental concentration images. Previously we have demonstrated the feasibility of NSECT using first generation CT approaches; while successful, the approach does not scale well and has limited resolution. Additionally, current gamma cameras operate in an energy range too low for NSECT imaging. However, the orbiting Reuven Ramaty High Energy Solar Spectroscopic Imager (RHESSI) captures and images gamma rays over the high-energy range equivalent to NSECT's (3 keV to 17 MeV) by utilizing Collimator-based Fourier transform imaging. A High Purity Germanium (HPGe) detector counts the number of energy events per unit of time, providing spectroscopic data. While a pair of rotating collimators placed in front of the detector modulates the number of gamma events, providing spatial information. Knowledge of the number of energy events at each discrete collimator angle allows for 2D image reconstruction. This method has proven successful at a focus of infinity in the RHESSI application. Our goal is to achieve similar results at a reasonable near-field focus. Here we describe the results of our simulations to implement a rotating modulation collimator (RMC) gamma imager for use in NSECT using simulations in Matlab. To determine feasible collimator setups and the stability of the inverse problem a Matlab environment was created that uses the geometry of the system to generate 1D observation data from 2D images and then to reconstruct 2D images using the MLEM algorithm. Reasonable collimator geometries were determined, successful reconstruction was achieved and the inverse problem was found to be stable.

Neutron stimulated emission computed tomography (NSECT) is presented as a new technique for in vivo tomographic spectroscopic imaging. A full implementation of NSECT is intended to provide an elemental spectrum of the body or part of the body being interrogated at each voxel of a three-dimensional computed tomographic image. An external neutron beam illuminates the sample and some of these neutrons scatter inelastically, producing characteristic gamma emission from the scattering nuclei. These characteristic gamma rays are acquired by a gamma spectrometer and the emitting nucleus is identified by the emitted gamma energy. The neutron beam is scanned over the body in a geometry that allows for tomographic reconstruction. Tomographic images of each element in the spectrum can be reconstructed to represent the spatial distribution of elements within the sample. Here we offer proof of concept for the NSECT method, present the first single projection spectra acquired from multi-element phantoms, and discuss potential biomedical applications.