Comparative performance of multiview stereoscopic and mammographic display modalities for breast lesion detection.

PURPOSE: Mammography is known to be one of the most difficult radiographic exams to interpret. Mammography has important limitations, including the superposition of normal tissue that can obscure a mass, chance alignment of normal tissue to mimic a true lesion and the inability to derive volumetric information. It has been shown that stereomammography can overcome these deficiencies by showing that layers of normal tissue lay at different depths. If standard stereomammography (i.e., a single stereoscopic pair consisting of two projection images) can significantly improve lesion detection, how will multiview stereoscopy (MVS), where many projection images are used, compare to mammography? The aim of this study was to assess the relative performance of MVS compared to mammography for breast mass detection. METHODS: The MVS image sets consisted of the 25 raw projection images acquired over an arc of approximately 45 degrees using a Siemens prototype breast tomosynthesis system. The mammograms were acquired using a commercial Siemens FFDM system. The raw data were taken from both of these systems for 27 cases and realistic simulated mass lesions were added to duplicates of the 27 images at the same local contrast. The images with lesions (27 mammography and 27 MVS) and the images without lesions (27 mammography and 27 MVS) were then postprocessed to provide comparable and representative image appearance across the two modalities. All 108 image sets were shown to five full-time breast imaging radiologists in random order on a state-of-the-art stereoscopic display. The observers were asked to give a confidence rating for each image (0 for lesion definitely not present, 100 for lesion definitely present). The ratings were then compiled and processed using ROC and variance analysis. RESULTS: The mean AUC for the five observers was 0.614 +/- 0.055 for mammography and 0.778 +/- 0.052 for multiview stereoscopy. The difference of 0.164 +/- 0.065 was statistically significant with a p-value of 0.0148. CONCLUSIONS: The differences in the AUCs and the p-value suggest that multiview stereoscopy has a statistically significant advantage over mammography in the detection of simulated breast masses. This highlights the dominance of anatomical noise compared to quantum noise for breast mass detection. It also shows that significant lesion detection can be achieved with MVS without any of the artifacts associated with tomosynthesis.

Advanced Techniques for Image Quality Assessment of Modern X-ray Computed Tomography Systems

X-ray computed tomography (CT) imaging constitutes one of the most widely used diagnostic tools in radiology today with nearly 85 million CT examinations performed in the U.S in 2011. CT imparts a relatively high amount of radiation dose to the patient compared to other x-ray imaging modalities and as a result of this fact, coupled with its popularity, CT is currently the single largest source of medical radiation exposure to the U.S. population. For this reason, there is a critical need to optimize CT examinations such that the dose is minimized while the quality of the CT images is not degraded. This optimization can be difficult to achieve due to the relationship between dose and image quality. All things being held equal, reducing the dose degrades image quality and can impact the diagnostic value of the CT examination.

A recent push from the medical and scientific community towards using lower doses has spawned new dose reduction technologies such as automatic exposure control (i.e., tube current modulation) and iterative reconstruction algorithms. In theory, these technologies could allow for scanning at reduced doses while maintaining the image quality of the exam at an acceptable level. Therefore, there is a scientific need to establish the dose reduction potential of these new technologies in an objective and rigorous manner. Establishing these dose reduction potentials requires precise and clinically relevant metrics of CT image quality, as well as practical and efficient methodologies to measure such metrics on real CT systems. The currently established methodologies for assessing CT image quality are not appropriate to assess modern CT scanners that have implemented those aforementioned dose reduction technologies.

Thus the purpose of this doctoral project was to develop, assess, and implement new phantoms, image quality metrics, analysis techniques, and modeling tools that are appropriate for image quality assessment of modern clinical CT systems. The project developed image quality assessment methods in the context of three distinct paradigms, (a) uniform phantoms, (b) textured phantoms, and (c) clinical images.

The work in this dissertation used the “task-based” definition of image quality. That is, image quality was broadly defined as the effectiveness by which an image can be used for its intended task. Under this definition, any assessment of image quality requires three components: (1) A well defined imaging task (e.g., detection of subtle lesions), (2) an “observer” to perform the task (e.g., a radiologists or a detection algorithm), and (3) a way to measure the observer’s performance in completing the task at hand (e.g., detection sensitivity/specificity).

First, this task-based image quality paradigm was implemented using a novel multi-sized phantom platform (with uniform background) developed specifically to assess modern CT systems (Mercury Phantom, v3.0, Duke University). A comprehensive evaluation was performed on a state-of-the-art CT system (SOMATOM Definition Force, Siemens Healthcare) in terms of noise, resolution, and detectability as a function of patient size, dose, tube energy (i.e., kVp), automatic exposure control, and reconstruction algorithm (i.e., Filtered Back-Projection– FPB vs Advanced Modeled Iterative Reconstruction– ADMIRE). A mathematical observer model (i.e., computer detection algorithm) was implemented and used as the basis of image quality comparisons. It was found that image quality increased with increasing dose and decreasing phantom size. The CT system exhibited nonlinear noise and resolution properties, especially at very low-doses, large phantom sizes, and for low-contrast objects. Objective image quality metrics generally increased with increasing dose and ADMIRE strength, and with decreasing phantom size. The ADMIRE algorithm could offer comparable image quality at reduced doses or improved image quality at the same dose (increase in detectability index by up to 163% depending on iterative strength). The use of automatic exposure control resulted in more consistent image quality with changing phantom size.

Based on those results, the dose reduction potential of ADMIRE was further assessed specifically for the task of detecting small (<=6 mm) low-contrast (<=20 HU) lesions. A new low-contrast detectability phantom (with uniform background) was designed and fabricated using a multi-material 3D printer. The phantom was imaged at multiple dose levels and images were reconstructed with FBP and ADMIRE. Human perception experiments were performed to measure the detection accuracy from FBP and ADMIRE images. It was found that ADMIRE had equivalent performance to FBP at 56% less dose.

Using the same image data as the previous study, a number of different mathematical observer models were implemented to assess which models would result in image quality metrics that best correlated with human detection performance. The models included naïve simple metrics of image quality such as contrast-to-noise ratio (CNR) and more sophisticated observer models such as the non-prewhitening matched filter observer model family and the channelized Hotelling observer model family. It was found that non-prewhitening matched filter observers and the channelized Hotelling observers both correlated strongly with human performance. Conversely, CNR was found to not correlate strongly with human performance, especially when comparing different reconstruction algorithms.

The uniform background phantoms used in the previous studies provided a good first-order approximation of image quality. However, due to their simplicity and due to the complexity of iterative reconstruction algorithms, it is possible that such phantoms are not fully adequate to assess the clinical impact of iterative algorithms because patient images obviously do not have smooth uniform backgrounds. To test this hypothesis, two textured phantoms (classified as gross texture and fine texture) and a uniform phantom of similar size were built and imaged on a SOMATOM Flash scanner (Siemens Healthcare). Images were reconstructed using FBP and a Sinogram Affirmed Iterative Reconstruction (SAFIRE). Using an image subtraction technique, quantum noise was measured in all images of each phantom. It was found that in FBP, the noise was independent of the background (textured vs uniform). However, for SAFIRE, noise increased by up to 44% in the textured phantoms compared to the uniform phantom. As a result, the noise reduction from SAFIRE was found to be up to 66% in the uniform phantom but as low as 29% in the textured phantoms. Based on this result, it clear that further investigation was needed into to understand the impact that background texture has on image quality when iterative reconstruction algorithms are used.

To further investigate this phenomenon with more realistic textures, two anthropomorphic textured phantoms were designed to mimic lung vasculature and fatty soft tissue texture. The phantoms (along with a corresponding uniform phantom) were fabricated with a multi-material 3D printer and imaged on the SOMATOM Flash scanner. Scans were repeated a total of 50 times in order to get ensemble statistics of the noise. A novel method of estimating the noise power spectrum (NPS) from irregularly shaped ROIs was developed. It was found that SAFIRE images had highly locally non-stationary noise patterns with pixels near edges having higher noise than pixels in more uniform regions. Compared to FBP, SAFIRE images had 60% less noise on average in uniform regions for edge pixels, noise was between 20% higher and 40% lower. The noise texture (i.e., NPS) was also highly dependent on the background texture for SAFIRE. Therefore, it was concluded that quantum noise properties in the uniform phantoms are not representative of those in patients for iterative reconstruction algorithms and texture should be considered when assessing image quality of iterative algorithms.

The move beyond just assessing noise properties in textured phantoms towards assessing detectability, a series of new phantoms were designed specifically to measure low-contrast detectability in the presence of background texture. The textures used were optimized to match the texture in the liver regions actual patient CT images using a genetic algorithm. The so called “Clustured Lumpy Background” texture synthesis framework was used to generate the modeled texture. Three textured phantoms and a corresponding uniform phantom were fabricated with a multi-material 3D printer and imaged on the SOMATOM Flash scanner. Images were reconstructed with FBP and SAFIRE and analyzed using a multi-slice channelized Hotelling observer to measure detectability and the dose reduction potential of SAFIRE based on the uniform and textured phantoms. It was found that at the same dose, the improvement in detectability from SAFIRE (compared to FBP) was higher when measured in a uniform phantom compared to textured phantoms.

The final trajectory of this project aimed at developing methods to mathematically model lesions, as a means to help assess image quality directly from patient images. The mathematical modeling framework is first presented. The models describe a lesion’s morphology in terms of size, shape, contrast, and edge profile as an analytical equation. The models can be voxelized and inserted into patient images to create so-called “hybrid” images. These hybrid images can then be used to assess detectability or estimability with the advantage that the ground truth of the lesion morphology and location is known exactly. Based on this framework, a series of liver lesions, lung nodules, and kidney stones were modeled based on images of real lesions. The lesion models were virtually inserted into patient images to create a database of hybrid images to go along with the original database of real lesion images. ROI images from each database were assessed by radiologists in a blinded fashion to determine the realism of the hybrid images. It was found that the radiologists could not readily distinguish between real and virtual lesion images (area under the ROC curve was 0.55). This study provided evidence that the proposed mathematical lesion modeling framework could produce reasonably realistic lesion images.

Based on that result, two studies were conducted which demonstrated the utility of the lesion models. The first study used the modeling framework as a measurement tool to determine how dose and reconstruction algorithm affected the quantitative analysis of liver lesions, lung nodules, and renal stones in terms of their size, shape, attenuation, edge profile, and texture features. The same database of real lesion images used in the previous study was used for this study. That database contained images of the same patient at 2 dose levels (50% and 100%) along with 3 reconstruction algorithms from a GE 750HD CT system (GE Healthcare). The algorithms in question were FBP, Adaptive Statistical Iterative Reconstruction (ASiR), and Model-Based Iterative Reconstruction (MBIR). A total of 23 quantitative features were extracted from the lesions under each condition. It was found that both dose and reconstruction algorithm had a statistically significant effect on the feature measurements. In particular, radiation dose affected five, three, and four of the 23 features (related to lesion size, conspicuity, and pixel-value distribution) for liver lesions, lung nodules, and renal stones, respectively. MBIR significantly affected 9, 11, and 15 of the 23 features (including size, attenuation, and texture features) for liver lesions, lung nodules, and renal stones, respectively. Lesion texture was not significantly affected by radiation dose.

The second study demonstrating the utility of the lesion modeling framework focused on assessing detectability of very low-contrast liver lesions in abdominal imaging. Specifically, detectability was assessed as a function of dose and reconstruction algorithm. As part of a parallel clinical trial, images from 21 patients were collected at 6 dose levels per patient on a SOMATOM Flash scanner. Subtle liver lesion models (contrast = -15 HU) were inserted into the raw projection data from the patient scans. The projections were then reconstructed with FBP and SAFIRE (strength 5). Also, lesion-less images were reconstructed. Noise, contrast, CNR, and detectability index of an observer model (non-prewhitening matched filter) were assessed. It was found that SAFIRE reduced noise by 52%, reduced contrast by 12%, increased CNR by 87%. and increased detectability index by 65% compared to FBP. Further, a 2AFC human perception experiment was performed to assess the dose reduction potential of SAFIRE, which was found to be 22% compared to the standard of care dose.

In conclusion, this dissertation provides to the scientific community a series of new methodologies, phantoms, analysis techniques, and modeling tools that can be used to rigorously assess image quality from modern CT systems. Specifically, methods to properly evaluate iterative reconstruction have been developed and are expected to aid in the safe clinical implementation of dose reduction technologies.

Prospective Estimation of Radiation Dose and Image Quality for Optimized CT Performance

X-ray computed tomography (CT) is a non-invasive medical imaging technique that generates cross-sectional images by acquiring attenuation-based projection measurements at multiple angles. Since its first introduction in the 1970s, substantial technical improvements have led to the expanding use of CT in clinical examinations. CT has become an indispensable imaging modality for the diagnosis of a wide array of diseases in both pediatric and adult populations [1, 2]. Currently, approximately 272 million CT examinations are performed annually worldwide, with nearly 85 million of these in the United States alone [3]. Although this trend has decelerated in recent years, CT usage is still expected to increase mainly due to advanced technologies such as multi-energy [4], photon counting [5], and cone-beam CT [6].

Despite the significant clinical benefits, concerns have been raised regarding the population-based radiation dose associated with CT examinations [7]. From 1980 to 2006, the effective dose from medical diagnostic procedures rose six-fold, with CT contributing to almost half of the total dose from medical exposure [8]. For each patient, the risk associated with a single CT examination is likely to be minimal. However, the relatively large population-based radiation level has led to enormous efforts among the community to manage and optimize the CT dose.

As promoted by the international campaigns Image Gently and Image Wisely, exposure to CT radiation should be appropriate and safe [9, 10]. It is thus a responsibility to optimize the amount of radiation dose for CT examinations. The key for dose optimization is to determine the minimum amount of radiation dose that achieves the targeted image quality [11]. Based on such principle, dose optimization would significantly benefit from effective metrics to characterize radiation dose and image quality for a CT exam. Moreover, if accurate predictions of the radiation dose and image quality were possible before the initiation of the exam, it would be feasible to personalize it by adjusting the scanning parameters to achieve a desired level of image quality. The purpose of this thesis is to design and validate models to quantify patient-specific radiation dose prospectively and task-based image quality. The dual aim of the study is to implement the theoretical models into clinical practice by developing an organ-based dose monitoring system and an image-based noise addition software for protocol optimization.

More specifically, Chapter 3 aims to develop an organ dose-prediction method for CT examinations of the body under constant tube current condition. The study effectively modeled the anatomical diversity and complexity using a large number of patient models with representative age, size, and gender distribution. The dependence of organ dose coefficients on patient size and scanner models was further evaluated. Distinct from prior work, these studies use the largest number of patient models to date with representative age, weight percentile, and body mass index (BMI) range.

With effective quantification of organ dose under constant tube current condition, Chapter 4 aims to extend the organ dose prediction system to tube current modulated (TCM) CT examinations. The prediction, applied to chest and abdominopelvic exams, was achieved by combining a convolution-based estimation technique that quantifies the radiation field, a TCM scheme that emulates modulation profiles from major CT vendors, and a library of computational phantoms with representative sizes, ages, and genders. The prospective quantification model is validated by comparing the predicted organ dose with the dose estimated based on Monte Carlo simulations with TCM function explicitly modeled.

Chapter 5 aims to implement the organ dose-estimation framework in clinical practice to develop an organ dose-monitoring program based on a commercial software (Dose Watch, GE Healthcare, Waukesha, WI). In the first phase of the study we focused on body CT examinations, and so the patient’s major body landmark information was extracted from the patient scout image in order to match clinical patients against a computational phantom in the library. The organ dose coefficients were estimated based on CT protocol and patient size as reported in Chapter 3. The exam CTDIvol, DLP, and TCM profiles were extracted and used to quantify the radiation field using the convolution technique proposed in Chapter 4.

With effective methods to predict and monitor organ dose, Chapters 6 aims to develop and validate improved measurement techniques for image quality assessment. Chapter 6 outlines the method that was developed to assess and predict quantum noise in clinical body CT images. Compared with previous phantom-based studies, this study accurately assessed the quantum noise in clinical images and further validated the correspondence between phantom-based measurements and the expected clinical image quality as a function of patient size and scanner attributes.

Chapter 7 aims to develop a practical strategy to generate hybrid CT images and assess the impact of dose reduction on diagnostic confidence for the diagnosis of acute pancreatitis. The general strategy is (1) to simulate synthetic CT images at multiple reduced-dose levels from clinical datasets using an image-based noise addition technique; (2) to develop quantitative and observer-based methods to validate the realism of simulated low-dose images; (3) to perform multi-reader observer studies on the low-dose image series to assess the impact of dose reduction on the diagnostic confidence for multiple diagnostic tasks; and (4) to determine the dose operating point for clinical CT examinations based on the minimum diagnostic performance to achieve protocol optimization.

Chapter 8 concludes the thesis with a summary of accomplished work and a discussion about future research.

Plate-specific gain map correction for the improvement of detective quantum efficiency in computed radiography.

PURPOSE: The purpose of this work is to improve the noise power spectrum (NPS), and thus the detective quantum efficiency (DQE), of computed radiography (CR) images by correcting for spatial gain variations specific to individual imaging plates. CR devices have not traditionally employed gain-map corrections, unlike the case with flat-panel detectors, because of the multiplicity of plates used with each reader. The lack of gain-map correction has limited the DQE(f) at higher exposures with CR. This current work describes a feasible solution to generating plate-specific gain maps. METHODS: Ten high-exposure open field images were taken with an RQA5 spectrum, using a sixth generation CR plate suspended in air without a cassette. Image values were converted to exposure, the plates registered using fiducial dots on the plate, the ten images averaged, and then high-pass filtered to remove low frequency contributions from field inhomogeneity. A gain-map was then produced by converting all pixel values in the average into fractions with mean of one. The resultant gain-map of the plate was used to normalize subsequent single images to correct for spatial gain fluctuation. To validate performance, the normalized NPS (NNPS) for all images was calculated both with and without the gain-map correction. Variations in the quality of correction due to exposure levels, beam voltage/spectrum, CR reader used, and registration were investigated. RESULTS: The NNPS with plate-specific gain-map correction showed improvement over the noncorrected case over the range of frequencies from 0.15 to 2.5 mm(-1). At high exposure (40 mR), NNPS was 50%-90% better with gain-map correction than without. A small further improvement in NNPS was seen from carefully registering the gain-map with subsequent images using small fiducial dots, because of slight misregistration during scanning. Further improvement was seen in the NNPS from scaling the gain map about the mean to account for different beam spectra. CONCLUSIONS: This study demonstrates that a simple gain-map can be used to correct for the fixed-pattern noise in a given plate and thus improve the DQE of CR imaging. Such a method could easily be implemented by manufacturers because each plate has a unique bar code and the gain-map for all plates associated with a reader could be stored for future retrieval. These experiments indicated that an improvement in NPS (and hence, DQE) is possible, depending on exposure level, over a wide range of frequencies with this technique.