Artificial intelligence in medicine: mitigating risks and maximizing benefits via quality assurance, quality control, and acceptance testing

The adoption of artificial intelligence (AI) tools in medicine poses challenges to existing clinical workflows. This commentary discusses the necessity of context-specific quality assurance (QA), emphasizing the need for robust QA measures with quality control (QC) procedures that encompass (1) acceptance testing (AT) before clinical use, (2) continuous QC monitoring, and (3) adequate user training. The discussion also covers essential components of AT and QA, illustrated with real-world examples. We also highlight what we see as the shared responsibility of manufacturers or vendors, regulators, healthcare systems, medical physicists, and clinicians to enact appropriate testing and oversight to ensure a safe and equitable transformation of medicine through AI.

Multiparametric MRI in Era of Artificial Intelligence for Bladder Cancer Therapies

This review focuses on the principles, applications, and performance of mpMRI for bladder imaging. Quantitative imaging biomarkers (QIBs) derived from mpMRI are increasingly used in oncological applications, including tumor staging, prognosis, and assessment of treatment response. To standardize mpMRI acquisition and interpretation, an expert panel developed the Vesical Imaging-Reporting and Data System (VI-RADS). Many studies confirm the standardization and high degree of inter-reader agreement to discriminate muscle invasiveness in bladder cancer, supporting VI-RADS implementation in routine clinical practice. The standard MRI sequences for VI-RADS scoring are anatomical imaging, including T2w images, and physiological imaging with diffusion-weighted MRI (DW-MRI) and dynamic contrast-enhanced MRI (DCE-MRI). Physiological QIBs derived from analysis of DW- and DCE-MRI data and radiomic image features extracted from mpMRI images play an important role in bladder cancer. The current development of AI tools for analyzing mpMRI data and their potential impact on bladder imaging are surveyed. AI architectures are often implemented based on convolutional neural networks (CNNs), focusing on narrow/specific tasks. The application of AI can substantially impact bladder imaging clinical workflows; for example, manual tumor segmentation, which demands high time commitment and has inter-reader variability, can be replaced by an autosegmentation tool. The use of mpMRI and AI is projected to drive the field toward the personalized management of bladder cancer patients.

Proof-of-concept and Randomized, Placebo-controlled Trials of an FcRn Inhibitor, Batoclimab, for Thyroid Eye Disease

CONTEXT

Inhibition of the neonatal fragment crystallizable receptor (FcRn) reduces pathogenic thyrotropin receptor antibodies (TSH-R-Ab) that drive pathology in thyroid eye disease (TED).

OBJECTIVE

We report the first clinical studies of an FcRn inhibitor, batoclimab, in TED.

DESIGN

Proof-of-concept (POC) and randomized, double-blind placebo-controlled trials.

SETTING

Multicenter.

PARTICIPANTS

Patients with moderate-to-severe, active TED.

INTERVENTION

In the POC trial, patients received weekly subcutaneous injections of batoclimab 680 mg for 2 weeks, followed by 340 mg for 4 weeks. In the double-blind trial, patients were randomized 2:2:1:2 to weekly batoclimab (680 mg, 340 mg, 255 mg) or placebo for 12 weeks.

MAIN OUTCOME

Change from baseline in serum anti-TSH-R-Ab and total IgG (POC); 12-week proptosis response (randomized trial).

RESULTS

The randomized trial was terminated because of an unanticipated increase in serum cholesterol; therefore, data from 65 of the planned 77 patients were analyzed. Both trials showed marked decreases in pathogenic anti-TSH-R-Ab and total IgG serum levels (P < .001) with batoclimab. In the randomized trial, there was no statistically significant difference with batoclimab vs placebo in proptosis response at 12 weeks, although significant differences were observed at several earlier timepoints. In addition, orbital muscle volume decreased (P < .03) at 12 weeks, whereas quality of life (appearance subscale) improved (P < .03) at 19 weeks in the 680-mg group. Batoclimab was generally well tolerated, with albumin reductions and increases in lipids that reversed upon discontinuation.

CONCLUSIONS

These results provide insight into the efficacy and safety of batoclimab and support its further investigation as a potential therapy for TED.

Deep learning denoising of digital breast tomosynthesis: Observer performance study of the effect on detection of microcalcifications in breast phantom images

BACKGROUND

The noise in digital breast tomosynthesis (DBT) includes x-ray quantum noise and detector readout noise. The total radiation dose of a DBT scan is kept at about the level of a digital mammogram but the detector noise is increased due to acquisition of multiple projections. The high noise can degrade the detectability of subtle lesions, specifically microcalcifications (MCs).

PURPOSE

We previously developed a deep-learning-based denoiser to improve the image quality of DBT. In the current study, we conducted an observer performance study with breast radiologists to investigate the feasibility of using deep-learning-based denoising to improve the detection of MCs in DBT.

METHODS

We have a modular breast phantom set containing seven 1-cm-thick heterogeneous 50% adipose/50% fibroglandular slabs custom-made by CIRS, Inc. (Norfolk, VA). We made six 5-cm-thick breast phantoms embedded with 144 simulated MC clusters of four nominal speck sizes (0.125-0.150, 0.150-0.180, 0.180-0.212, 0.212-0.250 mm) at random locations. The phantoms were imaged with a GE Pristina DBT system using the automatic standard (STD) mode. The phantoms were also imaged with the STD+ mode that increased the average glandular dose by 54% to be used as a reference condition for comparison of radiologists' reading. Our previously trained and validated denoiser was deployed to the STD images to obtain a denoised DBT set (dnSTD). Seven breast radiologists participated as readers to detect the MCs in the DBT volumes of the six phantoms under the three conditions (STD, STD+, dnSTD), totaling 18 DBT volumes. Each radiologist read all the 18 DBT volumes sequentially, which were arranged in a different order for each reader in a counter-balanced manner to minimize any potential reading order effects. They marked the location of each detected MC cluster and provided a conspicuity rating and their confidence level for the perceived cluster. The visual grading characteristics (VGC) analysis was used to compare the conspicuity ratings and the confidence levels of the radiologists for the detection of MCs.

RESULTS

The average sensitivities over all MC speck sizes were 65.3%, 73.2%, and 72.3%, respectively, for the radiologists reading the STD, dnSTD, and STD+ volumes. The sensitivity for dnSTD was significantly higher than that for STD (p < 0.005, two-tailed Wilcoxon signed rank test) and comparable to that for STD+. The average false positive rates were 3.9 ± 4.6, 2.8 ± 3.7, and 2.7 ± 3.9 marks per DBT volume, respectively, for reading the STD, dnSTD, and STD+ images but the difference between dnSTD and STD or STD+ did not reach statistical significance. The overall conspicuity ratings and confidence levels by VGC analysis for dnSTD were significantly higher than those for both STD and STD+ (p ≤ 0.001). The critical alpha value for significance was adjusted to be 0.025 with Bonferroni correction.

CONCLUSIONS

This observer study using breast phantom images showed that deep-learning-based denoising has the potential to improve the detection of MCs in noisy DBT images and increase radiologists' confidence in differentiating noise from MCs without increasing radiation dose. Further studies are needed to evaluate the generalizability of these results to the wide range of DBTs from human subjects and patient populations in clinical settings.

AI in medical imaging grand challenges: translation from competition to research benefit and patient care

Artificial intelligence (AI), in one form or another, has been a part of medical imaging for decades. The recent evolution of AI into approaches such as deep learning has dramatically accelerated the application of AI across a wide range of radiologic settings. Despite the promises of AI, developers and users of AI technology must be fully aware of its potential biases and pitfalls, and this knowledge must be incorporated throughout the AI system development pipeline that involves training, validation, and testing. Grand challenges offer an opportunity to advance the development of AI methods for targeted applications and provide a mechanism for both directing and facilitating the development of AI systems. In the process, a grand challenge centralizes (with the challenge organizers) the burden of providing a valid benchmark test set to assess performance and generalizability of participants' models and the collection and curation of image metadata, clinical/demographic information, and the required reference standard. The most relevant grand challenges are those designed to maximize the open-science nature of the competition, with code and trained models deposited for future public access. The ultimate goal of AI grand challenges is to foster the translation of AI systems from competition to research benefit and patient care. Rather than reference the many medical imaging grand challenges that have been organized by groups such as MICCAI, RSNA, AAPM, and grand-challenge.org, this review assesses the role of grand challenges in promoting AI technologies for research advancement and for eventual clinical implementation, including their promises and limitations.

Survival Prediction of Patients with Bladder Cancer after Cystectomy Based on Clinical, Radiomics, and Deep-Learning Descriptors

Accurate survival prediction for bladder cancer patients who have undergone radical cystectomy can improve their treatment management. However, the existing predictive models do not take advantage of both clinical and radiological imaging data. This study aimed to fill this gap by developing an approach that leverages the strengths of clinical (C), radiomics (R), and deep-learning (D) descriptors to improve survival prediction. The dataset comprised 163 patients, including clinical, histopathological information, and CT urography scans. The data were divided by patient into training, validation, and test sets. We analyzed the clinical data by a nomogram and the image data by radiomics and deep-learning models. The descriptors were input into a BPNN model for survival prediction. The AUCs on the test set were (C): 0.82 ± 0.06, (R): 0.73 ± 0.07, (D): 0.71 ± 0.07, (CR): 0.86 ± 0.05, (CD): 0.86 ± 0.05, and (CRD): 0.87 ± 0.05. The predictions based on D and CRD descriptors showed a significant difference (p = 0.007). For Kaplan-Meier survival analysis, the deceased and alive groups were stratified successfully by C (p < 0.001) and CRD (p < 0.001), with CRD predicting the alive group more accurately. The results highlight the potential of combining C, R, and D descriptors to accurately predict the survival of bladder cancer patients after cystectomy.

A Competition, Benchmark, Code, and Data for Using Artificial Intelligence to Detect Lesions in Digital Breast Tomosynthesis

IMPORTANCE

An accurate and robust artificial intelligence (AI) algorithm for detecting cancer in digital breast tomosynthesis (DBT) could significantly improve detection accuracy and reduce health care costs worldwide.

OBJECTIVES

To make training and evaluation data for the development of AI algorithms for DBT analysis available, to develop well-defined benchmarks, and to create publicly available code for existing methods.

DESIGN, SETTING, AND PARTICIPANTS

This diagnostic study is based on a multi-institutional international grand challenge in which research teams developed algorithms to detect lesions in DBT. A data set of 22 032 reconstructed DBT volumes was made available to research teams. Phase 1, in which teams were provided 700 scans from the training set, 120 from the validation set, and 180 from the test set, took place from December 2020 to January 2021, and phase 2, in which teams were given the full data set, took place from May to July 2021.

MAIN OUTCOMES AND MEASURES

The overall performance was evaluated by mean sensitivity for biopsied lesions using only DBT volumes with biopsied lesions; ties were broken by including all DBT volumes.

RESULTS

A total of 8 teams participated in the challenge. The team with the highest mean sensitivity for biopsied lesions was the NYU B-Team, with 0.957 (95% CI, 0.924-0.984), and the second-place team, ZeDuS, had a mean sensitivity of 0.926 (95% CI, 0.881-0.964). When the results were aggregated, the mean sensitivity for all submitted algorithms was 0.879; for only those who participated in phase 2, it was 0.926.

CONCLUSIONS AND RELEVANCE

In this diagnostic study, an international competition produced algorithms with high sensitivity for using AI to detect lesions on DBT images. A standardized performance benchmark for the detection task using publicly available clinical imaging data was released, with detailed descriptions and analyses of submitted algorithms accompanied by a public release of their predictions and code for selected methods. These resources will serve as a foundation for future research on computer-assisted diagnosis methods for DBT, significantly lowering the barrier of entry for new researchers.

AAPM task group report 273: Recommendations on best practices for AI and machine learning for computer-aided diagnosis in medical imaging

Rapid advances in artificial intelligence (AI) and machine learning, and specifically in deep learning (DL) techniques, have enabled broad application of these methods in health care. The promise of the DL approach has spurred further interest in computer-aided diagnosis (CAD) development and applications using both "traditional" machine learning methods and newer DL-based methods. We use the term CAD-AI to refer to this expanded clinical decision support environment that uses traditional and DL-based AI methods. Numerous studies have been published to date on the development of machine learning tools for computer-aided, or AI-assisted, clinical tasks. However, most of these machine learning models are not ready for clinical deployment. It is of paramount importance to ensure that a clinical decision support tool undergoes proper training and rigorous validation of its generalizability and robustness before adoption for patient care in the clinic. To address these important issues, the American Association of Physicists in Medicine (AAPM) Computer-Aided Image Analysis Subcommittee (CADSC) is charged, in part, to develop recommendations on practices and standards for the development and performance assessment of computer-aided decision support systems. The committee has previously published two opinion papers on the evaluation of CAD systems and issues associated with user training and quality assurance of these systems in the clinic. With machine learning techniques continuing to evolve and CAD applications expanding to new stages of the patient care process, the current task group report considers the broader issues common to the development of most, if not all, CAD-AI applications and their translation from the bench to the clinic. The goal is to bring attention to the proper training and validation of machine learning algorithms that may improve their generalizability and reliability and accelerate the adoption of CAD-AI systems for clinical decision support.

Computerized Decision Support for Bladder Cancer Treatment Response Assessment in CT Urography: Effect on Diagnostic Accuracy in Multi-Institution Multi-Specialty Study

This observer study investigates the effect of computerized artificial intelligence (AI)-based decision support system (CDSS-T) on physicians’ diagnostic accuracy in assessing bladder cancer treatment response. The performance of 17 observers was evaluated when assessing bladder cancer treatment response without and with CDSS-T using pre- and post-chemotherapy CTU scans in 123 patients having 157 pre- and post-treatment cancer pairs. The impact of cancer case difficulty, observers’ clinical experience, institution affiliation, specialty, and the assessment times on the observers’ diagnostic performance with and without using CDSS-T were analyzed. It was found that the average performance of the 17 observers was significantly improved (p = 0.002) when aided by the CDSS-T. The cancer case difficulty, institution affiliation, specialty, and the assessment times influenced the observers’ performance without CDSS-T. The AI-based decision support system has the potential to improve the diagnostic accuracy in assessing bladder cancer treatment response and result in more consistent performance among all physicians.

Lung Nodule Malignancy Prediction in Sequential CT Scans: Summary of ISBI 2018 Challenge

Lung cancer is by far the leading cause of cancer death in the US. Recent studies have demonstrated the effectiveness of screening using low dose CT (LDCT) in reducing lung cancer related mortality. While lung nodules are detected with a high rate of sensitivity, this exam has a low specificity rate and it is still difficult to separate benign and malignant lesions. The ISBI 2018 Lung Nodule Malignancy Prediction Challenge, developed by a team from the Quantitative Imaging Network of the National Cancer Institute, was focused on the prediction of lung nodule malignancy from two sequential LDCT screening exams using automated (non-manual) algorithms. We curated a cohort of 100 subjects who participated in the National Lung Screening Trial and had established pathological diagnoses. Data from 30 subjects were randomly selected for training and the remaining was used for testing. Participants were evaluated based on the area under the receiver operating characteristic curve (AUC) of nodule-wise malignancy scores generated by their algorithms on the test set. The challenge had 17 participants, with 11 teams submitting reports with method description, mandated by the challenge rules. Participants used quantitative methods, resulting in a reporting test AUC ranging from 0.698 to 0.913. The top five contestants used deep learning approaches, reporting an AUC between 0.87 - 0.91. The team's predictor did not achieve significant differences from each other nor from a volume change estimate (p =.05 with Bonferroni-Holm's correction).

AI in medical physics: guidelines for publication

The Abstract is intended to provide a concise summary of the study and its scientific findings. For AI/ML applications in medical physics, a problem statement and rationale for utilizing these algorithms are necessary while highlighting the novelty of the approach. A brief numerical description of how the data are partitioned into subsets for training of the AI/ML algorithm, validation (including tuning of parameters), and independent testing of algorithm performance is required. This is to be followed by a summary of the results and statistical metrics that quantify the performance of the AI/ML algorithm.

Standardization in Quantitative Imaging: A Multicenter Comparison of Radiomic Features from Different Software Packages on Digital Reference Objects and Patient Data Sets

Radiomic features are being increasingly studied for clinical applications. We aimed to assess the agreement among radiomic features when computed by several groups by using different software packages under very tightly controlled conditions, which included standardized feature definitions and common image data sets. Ten sites (9 from the NCI's Quantitative Imaging Network] positron emission tomography–computed tomography working group plus one site from outside that group) participated in this project. Nine common quantitative imaging features were selected for comparison including features that describe morphology, intensity, shape, and texture. The common image data sets were: three 3D digital reference objects (DROs) and 10 patient image scans from the Lung Image Database Consortium data set using a specific lesion in each scan. Each object (DRO or lesion) was accompanied by an already-defined volume of interest, from which the features were calculated. Feature values for each object (DRO or lesion) were reported. The coefficient of variation (CV), expressed as a percentage, was calculated across software packages for each feature on each object. Thirteen sets of results were obtained for the DROs and patient data sets. Five of the 9 features showed excellent agreement with CV < 1%; 1 feature had moderate agreement (CV < 10%), and 3 features had larger variations (CV ≥ 10%) even after attempts at harmonization of feature calculations. This work highlights the value of feature definition standardization as well as the need to further clarify definitions for some features.

Risks of feature leakage and sample size dependencies in deep feature extraction for breast mass classification

PURPOSE

Transfer learning is commonly used in deep learning for medical imaging to alleviate the problem of limited available data. In this work, we studied the risk of feature leakage and its dependence on sample size when using pretrained deep convolutional neural network (DCNN) as feature extractor for classification breast masses in mammography.

METHODS

Feature leakage occurs when the training set is used for feature selection and classifier modeling while the cost function is guided by the validation performance or informed by the test performance. The high-dimensional feature space extracted from pretrained DCNN suffers from the curse of dimensionality; feature subsets that can provide excessively optimistic performance can be found for the validation set or test set if the latter is allowed for unlimited reuse during algorithm development. We designed a simulation study to examine feature leakage when using DCNN as feature extractor for mass classification in mammography. Four thousand five hundred and seventy-seven unique mass lesions were partitioned by patient into three sets: 3222 for training, 508 for validation, and 847 for independent testing. Three pretrained DCNNs, AlexNet, GoogLeNet, and VGG16, were first compared using a training set in fourfold cross validation and one was selected as the feature extractor. To assess generalization errors, the independent test set was sequestered as truly unseen cases. A training set of a range of sizes from 10% to 75% was simulated by random drawing from the available training set in addition to 100% of the training set. Three commonly used feature classifiers, the linear discriminant, the support vector machine, and the random forest were evaluated. A sequential feature selection method was used to find feature subsets that could achieve high classification performance in terms of the area under the receiver operating characteristic curve (AUC) in the validation set. The extent of feature leakage and the impact of training set size were analyzed by comparison to the performance in the unseen test set.

RESULTS

All three classifiers showed large generalization error between the validation set and the independent sequestered test set at all sample sizes. The generalization error decreased as the sample size increased. At 100% of the sample size, one classifier achieved an AUC as high as 0.91 on the validation set while the corresponding performance on the unseen test set only reached an AUC of 0.72.

CONCLUSIONS

Our results demonstrate that large generalization errors can occur in AI tools due to feature leakage. Without evaluation on unseen test cases, optimistically biased performance may be reported inadvertently, and can lead to unrealistic expectations and reduce confidence for clinical implementation.

Prediction of Disease Free Survival in Laryngeal and Hypopharyngeal Cancers Using CT Perfusion and Radiomic Features: A Pilot Study

(1) Purpose: The objective was to evaluate CT perfusion and radiomic features for prediction of one year disease free survival in laryngeal and hypopharyngeal cancer. (2) Method and Materials: This retrospective study included pre and post therapy CT neck studies in 36 patients with laryngeal/hypopharyngeal cancer. Tumor contouring was performed semi-autonomously by the computer and manually by two radiologists. Twenty-six radiomic features including morphological and gray-level features were extracted by an internally developed and validated computer-aided image analysis system. The five perfusion features analyzed included permeability surface area product (PS), blood flow (flow), blood volume (BV), mean transit time (MTT), and time-to-maximum (Tmax). One year persistent/recurrent disease data were obtained following the final treatment of definitive chemoradiation or after total laryngectomy. We performed a two-loop leave-one-out feature selection and linear discriminant analysis classifier with generation of receiver operating characteristic (ROC) curves and confidence intervals (CI). (3) Results: 10 patients (28%) had recurrence/persistent disease at 1 year. For prediction, the change in blood flow demonstrated a training AUC of 0.68 (CI 0.47-0.85) and testing AUC of 0.66 (CI 0.47-0.85). The best features selected were a combination of perfusion and radiomic features including blood flow and computer-estimated percent volume changes-training AUC of 0.68 (CI 0.5-0.85) and testing AUC of 0.69 (CI 0.5-0.85). The laryngoscopic percent change in volume was a poor predictor with a testing AUC of 0.4 (CI 0.16-0.57). (4) Conclusions: A combination of CT perfusion and radiomic features are potential predictors of one-year disease free survival in laryngeal and hypopharyngeal cancer patients.

Controlled Drug Delivery Systems for Oral Cancer Treatment-Current Status and Future Perspectives

Oral squamous cell carcinoma (OSCC), which encompasses the oral cavity-derived malignancies, is a devastating disease causing substantial morbidity and mortality in both men and women. It is the most common subtype of the head and neck squamous cell carcinoma (HNSCC), which is ranked the sixth most common malignancy worldwide. Despite promising advancements in the conventional therapeutic approaches currently available for patients with oral cancer, many drawbacks are still to be addressed; surgical resection leads to permanent disfigurement, altered sense of self and debilitating physiological consequences, while chemo- and radio-therapies result in significant toxicities, all affecting patient wellbeing and quality of life. Thus, the development of novel therapeutic approaches or modifications of current strategies is paramount to improve individual health outcomes and survival, while early tumour detection remains a priority and significant challenge. In recent years, drug delivery systems and chronotherapy have been developed as alternative methods aiming to enhance the benefits of the current anticancer therapies, while minimizing their undesirable toxic effects on the healthy non-cancerous cells. Targeted drug delivery systems have the potential to increase drug bioavailability and bio-distribution at the site of the primary tumour. This review confers current knowledge on the diverse drug delivery methods, potential carriers (e.g., polymeric, inorganic, and combinational nanoparticles; nanolipids; hydrogels; exosomes) and anticancer targeted approaches for oral squamous cell carcinoma treatment, with an emphasis on their clinical relevance in the era of precision medicine, circadian chronobiology and patient-centred health care.

Breast Cancer Diagnosis in Digital Breast Tomosynthesis: Effects of Training Sample Size on Multi-Stage Transfer Learning Using Deep Neural Nets

In this paper, we developed a deep convolutional neural network (CNN) for the classification of malignant and benign masses in digital breast tomosynthesis (DBT) using a multi-stage transfer learning approach that utilized data from similar auxiliary domains for intermediate-stage fine-tuning. Breast imaging data from DBT, digitized screen-film mammography, and digital mammography totaling 4039 unique regions of interest (1797 malignant and 2242 benign) were collected. Using cross validation, we selected the best transfer network from six transfer networks by varying the level up to which the convolutional layers were frozen. In a single-stage transfer learning approach, knowledge from CNN trained on the ImageNet data was fine-tuned directly with the DBT data. In a multi-stage transfer learning approach, knowledge learned from ImageNet was first fine-tuned with the mammography data and then fine-tuned with the DBT data. Two transfer networks were compared for the second-stage transfer learning by freezing most of the CNN structures versus freezing only the first convolutional layer. We studied the dependence of the classification performance on training sample size for various transfer learning and fine-tuning schemes by varying the training data from 1% to 100% of the available sets. The area under the receiver operating characteristic curve (AUC) was used as a performance measure. The view-based AUC on the test set for single-stage transfer learning was 0.85 ± 0.05 and improved significantly (p <; 0.05$ ) to 0.91 ± 0.03 for multi-stage learning. This paper demonstrated that, when the training sample size from the target domain is limited, an additional stage of transfer learning using data from a similar auxiliary domain is advantageous.

The Use of Quantitative Imaging in Radiation Oncology: A Quantitative Imaging Network (QIN) Perspective

Modern radiation therapy is delivered with great precision, in part by relying on high-resolution multidimensional anatomic imaging to define targets in space and time. The development of quantitative imaging (QI) modalities capable of monitoring biologic parameters could provide deeper insight into tumor biology and facilitate more personalized clinical decision-making. The Quantitative Imaging Network (QIN) was established by the National Cancer Institute to advance and validate these QI modalities in the context of oncology clinical trials. In particular, the QIN has significant interest in the application of QI to widen the therapeutic window of radiation therapy. QI modalities have great promise in radiation oncology and will help address significant clinical needs, including finer prognostication, more specific target delineation, reduction of normal tissue toxicity, identification of radioresistant disease, and clearer interpretation of treatment response. Patient-specific QI is being incorporated into radiation treatment design in ways such as dose escalation and adaptive replanning, with the intent of improving outcomes while lessening treatment morbidities. This review discusses the current vision of the QIN, current areas of investigation, and how the QIN hopes to enhance the integration of QI into the practice of radiation oncology.

PROSTATEx Challenges for computerized classification of prostate lesions from multiparametric magnetic resonance images

Grand challenges stimulate advances within the medical imaging research community; within a competitive yet friendly environment, they allow for a direct comparison of algorithms through a well-defined, centralized infrastructure. The tasks of the two-part PROSTATEx Challenges (the PROSTATEx Challenge and the PROSTATEx-2 Challenge) are (1) the computerized classification of clinically significant prostate lesions and (2) the computerized determination of Gleason Grade Group in prostate cancer, both based on multiparametric magnetic resonance images. The challenges incorporate well-vetted cases for training and testing, a centralized performance assessment process to evaluate results, and an established infrastructure for case dissemination, communication, and result submission. In the PROSTATEx Challenge, 32 groups apply their computerized methods (71 methods total) to 208 prostate lesions in the test set. The area under the receiver operating characteristic curve for these methods in the task of differentiating between lesions that are and are not clinically significant ranged from 0.45 to 0.87; statistically significant differences in performance among the top-performing methods, however, are not observed. In the PROSTATEx-2 Challenge, 21 groups apply their computerized methods (43 methods total) to 70 prostate lesions in the test set. When compared with the reference standard, the quadratic-weighted kappa values for these methods in the task of assigning a five-point Gleason Grade Group to each lesion range from - 0.24 to 0.27; superiority to random guessing can be established for only two methods. When approached with a sense of commitment and scientific rigor, challenges foster interest in the designated task and encourage innovation in the field.

A pictorial review of bladder cancer nodal metastases

Lymph node involvement in bladder cancer is common and has prognostic implications. Early and accurate identification of metastatic lymph nodes is, therefore, important in ensuring appropriate patient triage and management. The purpose of this review is to provide a pictorial and educational overview of the staging and imaging appearance of metastatic lymph nodes in bladder cancer. Additionally, a secondary aim of this manuscript is to provide a review of the diagnostic accuracy of common imaging modalities available for detecting metastatic lymph nodes in affected patients.

Semi-automated pulmonary nodule interval segmentation using the NLST data

PURPOSE

To study the variability in volume change estimates of pulmonary nodules due to segmentation approaches used across several algorithms and to evaluate these effects on the ability to predict nodule malignancy.

METHODS

We obtained 100 patient image datasets from the National Lung Screening Trial (NLST) that had a nodule detected on each of two consecutive low dose computed tomography (LDCT) scans, with an equal proportion of malignant and benign cases (50 malignant, 50 benign). Information about the nodule location for the cases was provided by a screen capture with a bounding box and its axial location was indicated. Five participating quantitative imaging network (QIN) institutions performed nodule segmentation using their preferred semi-automated algorithms with no manual correction; teams were allowed to provide additional manually corrected segmentations (analyzed separately). The teams were asked to provide segmentation masks for each nodule at both time points. From these masks, the volume was estimated for the nodule at each time point; the change in volume (absolute and percent change) across time points was estimated as well. We used the concordance correlation coefficient (CCC) to compare the similarity of computed nodule volumes (absolute and percent change) across algorithms. We used Logistic regression model on the change in volume (absolute change and percent change) of the nodules to predict the malignancy status, the area under the receiver operating characteristic curve (AUROC) and confidence intervals were reported. Because the size of nodules was expected to have a substantial effect on segmentation variability, analysis of change in volumes was stratified by lesion size, where lesions were grouped into those with a longest diameter of <8 mm and those with longest diameter ≥ 8 mm.

RESULTS

We find that segmentation of the nodules shows substantial variability across algorithms, with the CCC ranging from 0.56 to 0.95 for change in volume (percent change in volume range was [0.15 to 0.86]) across the nodules. When examining nodules based on their longest diameter, we find the CCC had higher values for large nodules with a range of [0.54 to 0.93] among the algorithms, while percent change in volume was [0.3 to 0.95]. Compared to that of smaller nodules which had a range of [-0.0038 to 0.69] and percent change in volume was [-0.039 to 0.92]. The malignancy prediction results showed fairly consistent results across the institutions, the AUC using change in volume ranged from 0.65 to 0.89 (Percent change in volume was 0.64 to 0.86) for entire nodule range. Prediction improves for large nodule range (≥ 8 mm) with AUC range 0.75 to 0.90 (percent change in volume was 0.74 to 0.92). Compared to smaller nodule range (<8 mm) with AUC range 0.57 to 0.78 (percent change in volume was 0.59 to 0.77).

CONCLUSIONS

We find there is a fairly high concordance in the size measurements for larger nodules (≥8 mm) than the lower sizes (<8 mm) across algorithms. We find the change in nodule volume (absolute and percent change) were consistent predictors of malignancy across institutions, despite using different segmentation algorithms. Using volume change estimates without corrections shows slightly lower predictability (for two teams).

Urinary bladder cancer staging in CT urography using machine learning

PURPOSE

To evaluate the feasibility of using an objective computer-aided system to assess bladder cancer stage in CT Urography (CTU).

MATERIALS AND METHODS

A dataset consisting of 84 bladder cancer lesions from 76 CTU cases was used to develop the computerized system for bladder cancer staging based on machine learning approaches. The cases were grouped into two classes based on pathological stage ≥ T2 or below T2, which is the decision threshold for neoadjuvant chemotherapy treatment clinically. There were 43 cancers below stage T2 and 41 cancers at stage T2 or above. All 84 lesions were automatically segmented using our previously developed auto-initialized cascaded level sets (AI-CALS) method. Morphological and texture features were extracted. The features were divided into subspaces of morphological features only, texture features only, and a combined set of both morphological and texture features. The dataset was split into Set 1 and Set 2 for two-fold cross-validation. Stepwise feature selection was used to select the most effective features. A linear discriminant analysis (LDA), a neural network (NN), a support vector machine (SVM), and a random forest (RAF) classifier were used to combine the features into a single score. The classification accuracy of the four classifiers was compared using the area under the receiver operating characteristic (ROC) curve (A_z ).

RESULTS

Based on the texture features only, the LDA classifier achieved a test A_z of 0.91 on Set 1 and a test A_z of 0.88 on Set 2. The test A_z of the NN classifier for Set 1 and Set 2 were 0.89 and 0.92, respectively. The SVM classifier achieved test A_z of 0.91 on Set 1 and test A_z of 0.89 on Set 2. The test A_z of the RAF classifier for Set 1 and Set 2 was 0.89 and 0.97, respectively. The morphological features alone, the texture features alone, and the combined feature set achieved comparable classification performance.

CONCLUSION

The predictive model developed in this study shows promise as a classification tool for stratifying bladder cancer into two staging categories: greater than or equal to stage T2 and below stage T2.

Urinary bladder cancer staging in CT urography using machine learning

PURPOSE

To evaluate the feasibility of using an objective computer-aided system to assess bladder cancer stage in CT Urography (CTU).

MATERIALS AND METHODS

A dataset consisting of 84 bladder cancer lesions from 76 CTU cases was used to develop the computerized system for bladder cancer staging based on machine learning approaches. The cases were grouped into two classes based on pathological stage ≥ T2 or below T2, which is the decision threshold for neoadjuvant chemotherapy treatment clinically. There were 43 cancers below stage T2 and 41 cancers at stage T2 or above. All 84 lesions were automatically segmented using our previously developed auto-initialized cascaded level sets (AI-CALS) method. Morphological and texture features were extracted. The features were divided into subspaces of morphological features only, texture features only, and a combined set of both morphological and texture features. The dataset was split into Set 1 and Set 2 for two-fold cross-validation. Stepwise feature selection was used to select the most effective features. A linear discriminant analysis (LDA), a neural network (NN), a support vector machine (SVM), and a random forest (RAF) classifier were used to combine the features into a single score. The classification accuracy of the four classifiers was compared using the area under the receiver operating characteristic (ROC) curve (A_z ).

RESULTS

Based on the texture features only, the LDA classifier achieved a test A_z of 0.91 on Set 1 and a test A_z of 0.88 on Set 2. The test A_z of the NN classifier for Set 1 and Set 2 were 0.89 and 0.92, respectively. The SVM classifier achieved test A_z of 0.91 on Set 1 and test A_z of 0.89 on Set 2. The test A_z of the RAF classifier for Set 1 and Set 2 was 0.89 and 0.97, respectively. The morphological features alone, the texture features alone, and the combined feature set achieved comparable classification performance.

CONCLUSION

The predictive model developed in this study shows promise as a classification tool for stratifying bladder cancer into two staging categories: greater than or equal to stage T2 and below stage T2.

Mass detection in digital breast tomosynthesis: Deep convolutional neural network with transfer learning from mammography

PURPOSE

Develop a computer-aided detection (CAD) system for masses in digital breast tomosynthesis (DBT) volume using a deep convolutional neural network (DCNN) with transfer learning from mammograms.

METHODS

A data set containing 2282 digitized film and digital mammograms and 324 DBT volumes were collected with IRB approval. The mass of interest on the images was marked by an experienced breast radiologist as reference standard. The data set was partitioned into a training set (2282 mammograms with 2461 masses and 230 DBT views with 228 masses) and an independent test set (94 DBT views with 89 masses). For DCNN training, the region of interest (ROI) containing the mass (true positive) was extracted from each image. False positive (FP) ROIs were identified at prescreening by their previously developed CAD systems. After data augmentation, a total of 45 072 mammographic ROIs and 37 450 DBT ROIs were obtained. Data normalization and reduction of non-uniformity in the ROIs across heterogeneous data was achieved using a background correction method applied to each ROI. A DCNN with four convolutional layers and three fully connected (FC) layers was first trained on the mammography data. Jittering and dropout techniques were used to reduce overfitting. After training with the mammographic ROIs, all weights in the first three convolutional layers were frozen, and only the last convolution layer and the FC layers were randomly initialized again and trained using the DBT training ROIs. The authors compared the performances of two CAD systems for mass detection in DBT: one used the DCNN-based approach and the other used their previously developed feature-based approach for FP reduction. The prescreening stage was identical in both systems, passing the same set of mass candidates to the FP reduction stage. For the feature-based CAD system, 3D clustering and active contour method was used for segmentation; morphological, gray level, and texture features were extracted and merged with a linear discriminant classifier to score the detected masses. For the DCNN-based CAD system, ROIs from five consecutive slices centered at each candidate were passed through the trained DCNN and a mass likelihood score was generated. The performances of the CAD systems were evaluated using free-response ROC curves and the performance difference was analyzed using a non-parametric method.

RESULTS

Before transfer learning, the DCNN trained only on mammograms with an AUC of 0.99 classified DBT masses with an AUC of 0.81 in the DBT training set. After transfer learning with DBT, the AUC improved to 0.90. For breast-based CAD detection in the test set, the sensitivity for the feature-based and the DCNN-based CAD systems was 83% and 91%, respectively, at 1 FP/DBT volume. The difference between the performances for the two systems was statistically significant (p-value < 0.05).

CONCLUSIONS

The image patterns learned from the mammograms were transferred to the mass detection on DBT slices through the DCNN. This study demonstrated that large data sets collected from mammography are useful for developing new CAD systems for DBT, alleviating the problem and effort of collecting entirely new large data sets for the new modality.

Computer-aided detection of retained surgical needles from postoperative radiographs

PURPOSE

Foreign objects, such as surgical sponges, needles, sutures, and other surgical instruments, retained in the patient's body can have dire consequences in terms of patient mortality as well as legal and financial penalties. We propose computer-aided detection (CAD) on postoperative radiographs as a potential solution to reduce the chance of retained foreign objects (RFOs) after surgery, thus alleviating one of the major concerns for patient safety in the operation room. A CAD system can function as a second pair of eyes or a prescreener for the surgeon and radiologist, depending on the CAD system design and the workflow. In this work, we focus on the detection of surgical needles on postoperative radiographs. As needles are frequently observed RFOs, a CAD system that can offer high sensitivity and specificity toward detecting surgical needles will be useful.

METHODS

Our CAD system incorporates techniques such as image segmentation, image enhancement, feature analysis, and curve fitting to detect surgical needles on radiographs. A dataset consisting of 108 cadaver images with a total of 116 needles and 100 cadaver "normal" images without needles was acquired with a portable digital x-ray system. A reference standard was obtained by marking the needle locations using an in-house developed graphical user interface. The 108 cadaver images with the needles were partitioned into a training set containing 53 cadaver images with 59 needles and a test set containing 55 cadaver images with 57 needles. All of the 100 cadaver normal images were reserved as a part of the test set and used to estimate the false-positive detection rate. Two operating points were chosen from the CAD system such that it can be operated in two modes, one with higher specificity (mode I) and the other with higher sensitivity (mode II).

RESULTS

For the training set, the CAD system with the rule-based classifier achieved a sensitivity of 74.6% with 0.15 false positives per image (FPs/image) in mode I and a sensitivity of 89.8% with 0.36 FPs/image in mode II. For the test set, the CAD system achieved a sensitivity of 77.2% with 0.26 FPs/image in mode I and a sensitivity of 84.2% with 0.6 FPs/image in mode II. For comparison, the CAD system with the neural network classifier achieved a sensitivity of 74.6% with 0.08 FPs/image in mode I and a sensitivity of 88.1% with 0.28 FPs/image in mode II for the training set, and a sensitivity of 75.4% with 0.23 FPs/image in mode I and a sensitivity of 86.0% with 0.57 FPs/image in mode II for the test set.

CONCLUSION

A novel CAD system has been developed for automated detection of needles inadvertently left behind in a patient's body from postsurgery radiographs. The pilot system offers reasonable performance in both the high sensitivity and high specificity modes. This preliminary study shows the promise of CAD as a low-cost and efficient aid for reducing retained surgical needles in patients.

Urinary bladder segmentation in CT urography using deep-learning convolutional neural network and level sets

PURPOSE

The authors are developing a computerized system for bladder segmentation in CT urography (CTU) as a critical component for computer-aided detection of bladder cancer.

METHODS

A deep-learning convolutional neural network (DL-CNN) was trained to distinguish between the inside and the outside of the bladder using 160 000 regions of interest (ROI) from CTU images. The trained DL-CNN was used to estimate the likelihood of an ROI being inside the bladder for ROIs centered at each voxel in a CTU case, resulting in a likelihood map. Thresholding and hole-filling were applied to the map to generate the initial contour for the bladder, which was then refined by 3D and 2D level sets. The segmentation performance was evaluated using 173 cases: 81 cases in the training set (42 lesions, 21 wall thickenings, and 18 normal bladders) and 92 cases in the test set (43 lesions, 36 wall thickenings, and 13 normal bladders). The computerized segmentation accuracy using the DL likelihood map was compared to that using a likelihood map generated by Haar features and a random forest classifier, and that using our previous conjoint level set analysis and segmentation system (CLASS) without using a likelihood map. All methods were evaluated relative to the 3D hand-segmented reference contours.

RESULTS

With DL-CNN-based likelihood map and level sets, the average volume intersection ratio, average percent volume error, average absolute volume error, average minimum distance, and the Jaccard index for the test set were 81.9% ± 12.1%, 10.2% ± 16.2%, 14.0% ± 13.0%, 3.6 ± 2.0 mm, and 76.2% ± 11.8%, respectively. With the Haar-feature-based likelihood map and level sets, the corresponding values were 74.3% ± 12.7%, 13.0% ± 22.3%, 20.5% ± 15.7%, 5.7 ± 2.6 mm, and 66.7% ± 12.6%, respectively. With our previous CLASS with local contour refinement (LCR) method, the corresponding values were 78.0% ± 14.7%, 16.5% ± 16.8%, 18.2% ± 15.0%, 3.8 ± 2.3 mm, and 73.9% ± 13.5%, respectively.

CONCLUSIONS

The authors demonstrated that the DL-CNN can overcome the strong boundary between two regions that have large difference in gray levels and provides a seamless mask to guide level set segmentation, which has been a problem for many gradient-based segmentation methods. Compared to our previous CLASS with LCR method, which required two user inputs to initialize the segmentation, DL-CNN with level sets achieved better segmentation performance while using a single user input. Compared to the Haar-feature-based likelihood map, the DL-CNN-based likelihood map could guide the level sets to achieve better segmentation. The results demonstrate the feasibility of our new approach of using DL-CNN in combination with level sets for segmentation of the bladder.

Radiomics of Lung Nodules: A Multi-Institutional Study of Robustness and Agreement of Quantitative Imaging Features

Radiomics is to provide quantitative descriptors of normal and abnormal tissues during classification and prediction tasks in radiology and oncology. Quantitative Imaging Network members are developing radiomic “feature” sets to characterize tumors, in general, the size, shape, texture, intensity, margin, and other aspects of the imaging features of nodules and lesions. Efforts are ongoing for developing an ontology to describe radiomic features for lung nodules, with the main classes consisting of size, local and global shape descriptors, margin, intensity, and texture-based features, which are based on wavelets, Laplacian of Gaussians, Law's features, gray-level co-occurrence matrices, and run-length features. The purpose of this study is to investigate the sensitivity of quantitative descriptors of pulmonary nodules to segmentations and to illustrate comparisons across different feature types and features computed by different implementations of feature extraction algorithms. We calculated the concordance correlation coefficients of the features as a measure of their stability with the underlying segmentation; 68% of the 830 features in this study had a concordance CC of ≥0.75. Pairwise correlation coefficients between pairs of features were used to uncover associations between features, particularly as measured by different participants. A graphical model approach was used to enumerate the number of uncorrelated feature groups at given thresholds of correlation. At a threshold of 0.75 and 0.95, there were 75 and 246 subgroups, respectively, providing a measure for the features' redundancy.

LUNGx Challenge for computerized lung nodule classification

The purpose of this work is to describe the LUNGx Challenge for the computerized classification of lung nodules on diagnostic computed tomography (CT) scans as benign or malignant and report the performance of participants' computerized methods along with that of six radiologists who participated in an observer study performing the same Challenge task on the same dataset. The Challenge provided sets of calibration and testing scans, established a performance assessment process, and created an infrastructure for case dissemination and result submission. Ten groups applied their own methods to 73 lung nodules (37 benign and 36 malignant) that were selected to achieve approximate size matching between the two cohorts. Area under the receiver operating characteristic curve (AUC) values for these methods ranged from 0.50 to 0.68; only three methods performed statistically better than random guessing. The radiologists' AUC values ranged from 0.70 to 0.85; three radiologists performed statistically better than the best-performing computer method. The LUNGx Challenge compared the performance of computerized methods in the task of differentiating benign from malignant lung nodules on CT scans, placed in the context of the performance of radiologists on the same task. The continued public availability of the Challenge cases will provide a valuable resource for the medical imaging research community.

Computer aided detection of surgical retained foreign object for prevention

PURPOSE

Surgical retained foreign objects (RFOs) have significant morbidity and mortality. They are associated with approximately $1.5 × 10(9) annually in preventable medical costs. The detection accuracy of radiographs for RFOs is a mediocre 59%. The authors address the RFO problem with two complementary technologies: a three-dimensional (3D) gossypiboma micro tag, the μTag that improves the visibility of RFOs on radiographs, and a computer aided detection (CAD) system that detects the μTag. It is desirable for the CAD system to operate in a high specificity mode in the operating room (OR) and function as a first reader for the surgeon. This allows for fast point of care results and seamless workflow integration. The CAD system can also operate in a high sensitivity mode as a second reader for the radiologist to ensure the highest possible detection accuracy.

METHODS

The 3D geometry of the μTag produces a similar two dimensional (2D) depiction on radiographs regardless of its orientation in the human body and ensures accurate detection by a radiologist and the CAD. The authors created a data set of 1800 cadaver images with the 3D μTag and other common man-made surgical objects positioned randomly. A total of 1061 cadaver images contained a single μTag and the remaining 739 were without μTag. A radiologist marked the location of the μTag using an in-house developed graphical user interface. The data set was partitioned into three independent subsets: a training set, a validation set, and a test set, consisting of 540, 560, and 700 images, respectively. A CAD system with modules that included preprocessing μTag enhancement, labeling, segmentation, feature analysis, classification, and detection was developed. The CAD system was developed using the training and the validation sets.

RESULTS

On the training set, the CAD achieved 81.5% sensitivity with 0.014 false positives (FPs) per image in a high specificity mode for the surgeons in the OR and 96.1% sensitivity with 0.81 FPs per image in a high sensitivity mode for the radiologists. On the independent test set, the CAD achieved 79.5% sensitivity with 0.003 FPs per image in a high specificity mode for the surgeons and 90.2% sensitivity with 0.23 FPs per image in a high sensitivity mode for the radiologists.

CONCLUSIONS

To the best of the authors' knowledge, this is the first time a 3D μTag is used to produce a recognizable, substantially similar 2D projection on radiographs regardless of orientation in space. It is the first time a CAD system is used to search for man-made objects over anatomic background. The CAD system for the μTags achieved reasonable performance in both the high specificity and the high sensitivity modes.

Detection of urinary bladder mass in CT urography with SPAN

PURPOSE

The authors are developing a computer-aided detection system for bladder cancer on CT urography (CTU). In this study, the authors focused on developing a system for detecting masses fully or partially within the contrast-enhanced (C) region of the bladder.

METHODS

With IRB approval, a data set of 70 patients with biopsy-proven bladder lesions fully or partially immersed within the contrast-enhanced region (C region) of the bladder was collected for this study: 35 patients for the training set (39 malignant, 7 benign lesions) and 35 patients for the test set (49 malignant, 4 benign lesions). The bladder in the CTU images was automatically segmented using the authors' conjoint level set analysis and segmentation system, which they developed specifically to segment the bladder. A closed contour of the C region of the bladder was generated by maximum intensity projection using the property that the dependently layering contrast material in the bladder will be filled consistently to the same level along all CTU slices due to gravity. Potential lesion candidates within the C region contour were found using the authors' Straightened Periphery ANalysis (SPAN) method. SPAN transforms a bladder wall to a straightened thickness profile, marks suspicious pixels on the profile, and clusters them into regions of interest to identify potential lesion candidates. The candidate regions were automatically segmented using the authors' autoinitialized cascaded level set segmentation method. Twenty-three morphological features were automatically extracted from the segmented lesions. The training set was used to determine the best subset of these features using simplex optimization with the leave-one-out case method. A linear discriminant classifier was designed for the classification of bladder lesions and false positives. The detection performance was evaluated on the independent test set by free-response receiver operating characteristic analysis.

RESULTS

At the prescreening step, the authors' system achieved 84.4% sensitivity with an average of 4.3 false positives per case (FPs/case) for the training set, and 84.9% sensitivity with 5.4 FPs/case for the test set. After linear discriminant analysis (LDA) classification with the selected features, the FP rate improved to 2.5 FPs/case for the training set, and 4.3 FPs/case for the test set without missing additional true lesions. By varying the threshold for the LDA scores, at 2.5 FPs/case, the sensitivities were 84.4% and 81.1% for the training and test sets, respectively. At 1.7 FPs/case, the sensitivities decreased to 77.8% and 75.5%, respectively.

CONCLUSIONS

The results demonstrate the feasibility of the authors' method for detection of bladder lesions fully or partially immersed in the contrast-enhanced region of CTU.

Ureter tracking and segmentation in CT urography (CTU) using COMPASS

PURPOSE

The authors are developing a computerized system for automated segmentation of ureters in CTU, referred to as combined model-guided path-finding analysis and segmentation system (COMPASS). Ureter segmentation is a critical component for computer-aided diagnosis of ureter cancer.

METHODS

COMPASS consists of three stages: (1) rule-based adaptive thresholding and region growing, (2) path-finding and propagation, and (3) edge profile extraction and feature analysis. With institutional review board approval, 79 CTU scans performed with intravenous (IV) contrast material enhancement were collected retrospectively from 79 patient files. One hundred twenty-four ureters were selected from the 79 CTU volumes. On average, the ureters spanned 283 computed tomography slices (range: 116-399, median: 301). More than half of the ureters contained malignant or benign lesions and some had ureter wall thickening due to malignancy. A starting point for each of the 124 ureters was identified manually to initialize the tracking by COMPASS. In addition, the centerline of each ureter was manually marked and used as reference standard for evaluation of tracking performance. The performance of COMPASS was quantitatively assessed by estimating the percentage of the length that was successfully tracked and segmented for each ureter and by estimating the average distance and the average maximum distance between the computer and the manually tracked centerlines.

RESULTS

Of the 124 ureters, 120 (97%) were segmented completely (100%), 121 (98%) were segmented through at least 70%, and 123 (99%) were segmented through at least 50% of its length. In comparison, using our previous method, 85 (69%) ureters were segmented completely (100%), 100 (81%) were segmented through at least 70%, and 107 (86%) were segmented at least 50% of its length. With COMPASS, the average distance between the computer and the manually generated centerlines is 0.54 mm, and the average maximum distance is 2.02 mm. With our previous method, the average distance between the centerlines was 0.80 mm, and the average maximum distance was 3.38 mm. The improvements in the ureteral tracking length and both distance measures were statistically significant (p < 0.0001).

CONCLUSIONS

COMPASS improved significantly the ureter tracking, including regions across ureter lesions, wall thickening, and the narrowing of the lumen.

Inter-Method Performance Study of Tumor Volumetry Assessment on Computed Tomography Test-Retest Data

RATIONALE AND OBJECTIVES

Tumor volume change has potential as a biomarker for diagnosis, therapy planning, and treatment response. Precision was evaluated and compared among semiautomated lung tumor volume measurement algorithms from clinical thoracic computed tomography data sets. The results inform approaches and testing requirements for establishing conformance with the Quantitative Imaging Biomarker Alliance (QIBA) Computed Tomography Volumetry Profile.

MATERIALS AND METHODS

Industry and academic groups participated in a challenge study. Intra-algorithm repeatability and inter-algorithm reproducibility were estimated. Relative magnitudes of various sources of variability were estimated using a linear mixed effects model. Segmentation boundaries were compared to provide a basis on which to optimize algorithm performance for developers.

RESULTS

Intra-algorithm repeatability ranged from 13% (best performing) to 100% (least performing), with most algorithms demonstrating improved repeatability as the tumor size increased. Inter-algorithm reproducibility was determined in three partitions and was found to be 58% for the four best performing groups, 70% for the set of groups meeting repeatability requirements, and 84% when all groups but the least performer were included. The best performing partition performed markedly better on tumors with equivalent diameters greater than 40 mm. Larger tumors benefitted by human editing but smaller tumors did not. One-fifth to one-half of the total variability came from sources independent of the algorithms. Segmentation boundaries differed substantially, not ony in overall volume but also in detail.

CONCLUSIONS

Nine of the 12 participating algorithms pass precision requirements similar to what is indicated in the QIBA Profile, with the caveat that the present study was not designed to explicitly evaluate algorithm profile conformance. Change in tumor volume can be measured with confidence to within ±14% using any of these nine algorithms on tumor sizes greater than 10 mm. No partition of the algorithms was able to meet the QIBA requirements for interchangeability down to 10 mm, although the partition comprising best performing algorithms did meet this requirement for a tumor size of greater than approximately 40 mm.

Treatment Response Assessment for Bladder Cancer on CT Based on Computerized Volume Analysis, World Health Organization Criteria, and RECIST

OBJECTIVE

The purpose of this study was to evaluate the accuracy of our autoinitialized cascaded level set 3D segmentation system as compared with the World Health Organization (WHO) criteria and the Response Evaluation Criteria In Solid Tumors (RECIST) for estimation of treatment response of bladder cancer in CT urography.

MATERIALS AND METHODS

CT urograms before and after neoadjuvant chemo-therapy treatment were collected from 18 patients with muscle-invasive localized or locally advanced bladder cancers. The disease stage as determined on pathologic samples at cystectomy after chemotherapy was considered as reference standard of treatment response. Two radiologists measured the longest diameter and its perpendicular on the pre- and posttreatment scans. Full 3D contours for all tumors were manually outlined by one radiologist. The autoinitialized cascaded level set method was used to automatically extract 3D tumor boundary. The prediction accuracy of pT0 disease (complete response) at cystectomy was estimated by the manual, autoinitialized cascaded level set, WHO, and RECIST methods on the basis of the AUC.

RESULTS

The AUC for prediction of pT0 disease at cystectomy was 0.78 ± 0.11 for autoinitialized cascaded level set compared with 0.82 ± 0.10 for manual segmentation. The difference did not reach statistical significance (p = 0.67). The AUCs using RECIST criteria were 0.62 ± 0.16 and 0.71 ± 0.12 for the two radiologists, both lower than those of the two 3D methods. The AUCs using WHO criteria were 0.56 ± 0.15 and 0.60 ± 0.13 and thus were lower than all other methods.

CONCLUSION

The pre- and posttreatment 3D volume change estimates obtained by the radiologist's manual outlines and the autoinitialized cascaded level set segmentation were more accurate for irregularly shaped tumors than were those based on RECIST and WHO criteria.

Digital breast tomosynthesis: studies of the effects of acquisition geometry on contrast-to-noise ratio and observer preference of low-contrast objects in breast phantom images

The effect of acquisition geometry in digital breast tomosynthesis was evaluated with studies of contrast-to-noise ratios (CNRs) and observer preference. Contrast-detail (CD) test objects in 5 cm thick phantoms with breast-like backgrounds were imaged. Twelve different angular acquisitions (average glandular dose for each ~1.1 mGy) were performed ranging from narrow angle 16° with 17 projection views (16d17p) to wide angle 64d17p. Focal slices of SART-reconstructed images of the CD arrays were selected for CNR computations and the reader preference study. For the latter, pairs of images obtained with different acquisition geometries were randomized and scored by 7 trained readers. The total scores for all images and readings for each acquisition geometry were compared as were the CNRs. In general, readers preferred images acquired with wide angle as opposed to narrow angle geometries. The mean percent preferred was highly correlated with tomosynthesis angle (R = 0.91). The highest scoring geometries were 60d21p (95%), 64d17p (80%), and 48d17p (72%); the lowest scoring were 16d17p (4%), 24d9p (17%) and 24d13p (33%). The measured CNRs for the various acquisitions showed much overlap but were overall highest for wide-angle acquisitions. Finally, the mean reader scores were well correlated with the mean CNRs (R = 0.83).

Urinary bladder segmentation in CT urography (CTU) using CLASS

PURPOSE

The authors are developing a computerized system for bladder segmentation on CTU, as a critical component for computer aided diagnosis of bladder cancer.

METHODS

A challenge for bladder segmentation is the presence of regions without contrast (NC) and filled with intravenous contrast (C). The authors have designed a Conjoint Level set Analysis and Segmentation System (CLASS) specifically for this application. CLASS performs a series of image processing tasks: preprocessing, initial segmentation, 3D and 2D level set segmentation, and postprocessing, designed according to the characteristics of the bladder in CTU. The NC and the C regions of the bladder were segmented separately in CLASS. The final contour is obtained in the postprocessing stage by the union of the NC and C contours. With Institutional Review Board (IRB) approval, the authors retrospectively collected 81 CTU scans, in which 40 bladders contained lesions, 26 contained diffuse wall thickening, and 15 were considered to be normal. The bladders were segmented by CLASS and the performance was assessed by rating the quality of the contours on a 10-point scale (1 = "very poor," 5 = "fair," 10 = "perfect"). For 30 bladders, 3D hand-segmented contours were obtained and the segmentation accuracy of CLASS was evaluated and compared to that of a single level set method in terms of the average minimum distance, average volume intersection ratio, average volume error and Jaccard index.

RESULTS

Of the 81 bladders, the average quality rating for CLASS was 6.5 ± 1.3. Thirty nine bladders were given quality ratings of 7 or above. Only five bladders had ratings under 5. The average minimum distance, average volume intersection ratio, average volume error, and average Jaccard index for CLASS were 3.5 ± 1.3 mm, (79.0 ± 8.2)%, (16.1 ± 16.3)%, and (75.7 ± 8.4)%, respectively, and for the single level set method were 5.2 ± 2.6 mm, (78.8 ± 16.3)%, (8.3 ± 33.1)%, (71.0 ± 15.4)%, respectively.

CONCLUSIONS

The results demonstrate the potential of CLASS for segmentation of the bladder.

Computerized detection of noncalcified plaques in coronary CT angiography: evaluation of topological soft gradient prescreening method and luminal analysis

PURPOSE

The buildup of noncalcified plaques (NCPs) that are vulnerable to rupture in coronary arteries is a risk for myocardial infarction. Interpretation of coronary CT angiography (cCTA) to search for NCP is a challenging task for radiologists due to the low CT number of NCP, the large number of coronary arteries, and multiple phase CT acquisition. The authors conducted a preliminary study to develop machine learning method for automated detection of NCPs in cCTA.

METHODS

With IRB approval, a data set of 83 ECG-gated contrast enhanced cCTA scans with 120 NCPs was collected retrospectively from patient files. A multiscale coronary artery response and rolling balloon region growing (MSCAR-RBG) method was applied to each cCTA volume to extract the coronary arterial trees. Each extracted vessel was reformatted to a straightened volume composed of cCTA slices perpendicular to the vessel centerline. A topological soft-gradient (TSG) detection method was developed to prescreen for NCP candidates by analyzing the 2D topological features of the radial gradient field surface along the vessel wall. The NCP candidates were then characterized by a luminal analysis that used 3D geometric features to quantify the shape information and gray-level features to evaluate the density of the NCP candidates. With machine learning techniques, useful features were identified and combined into an NCP score to differentiate true NCPs from false positives (FPs). To evaluate the effectiveness of the image analysis methods, the authors performed tenfold cross-validation with the available data set. Receiver operating characteristic (ROC) analysis was used to assess the classification performance of individual features and the NCP score. The overall detection performance was estimated by free response ROC (FROC) analysis.

RESULTS

With our TSG prescreening method, a prescreening sensitivity of 92.5% (111/120) was achieved with a total of 1181 FPs (14.2 FPs/scan). On average, six features were selected during the tenfold cross-validation training. The average area under the ROC curve (AUC) value for training was 0.87 ± 0.01 and the AUC value for validation was 0.85 ± 0.01. Using the NCP score, FROC analysis of the validation set showed that the FP rates were reduced to 3.16, 1.90, and 1.39 FPs/scan at sensitivities of 90%, 80%, and 70%, respectively.

CONCLUSIONS

The topological soft-gradient prescreening method in combination with the luminal analysis for FP reduction was effective for detection of NCPs in cCTA, including NCPs causing positive or negative vessel remodeling. The accuracy of vessel segmentation, tracking, and centerline identification has a strong impact on NCP detection. Studies are underway to further improve these techniques and reduce the FPs of the CADe system.

Coronary CT angiography (cCTA): automated registration of coronary arterial trees from multiple phases

Coronary computed tomography angiography (cCTA) is a commonly used imaging modality for the evaluation of coronary artery disease. cCTA is generally reconstructed in multiple cardiac phases because different coronary arteries may be better visualized in some phases than in others due to the periodic cardiac motion. We are developing an automated registration method for coronary arterial trees from multiple-phase cCTA that has potential application in building a 'best-quality' tree to facilitate image analysis and detection of stenotic plaques. Given the segmented left or right coronary arterial (LCA or RCA) trees from the multiple phases as input, the adjacent phase pairs, where displacements are relatively small, are registered by a specifically designed method based on a cubic B-spline with fast localized optimization (CBSO). For the phase pairs with large displacements, a global registration using an affine transform with quadratic terms and nonlinear simplex optimization (AQSO) is followed by a local registration using CBSO to refine the AQSO registered volumes. 26 LCA and 26 RCA trees with six cCTA phases from 26 patients were used for registration evaluation. The average distances for the tree pairs between the adjacent phases with small displacements before and after CBSO registration were 0.96  ±  0.79 and 0.76  ±  0.61 mm respectively for LCA, and 0.93  ±  0.97 and 0.64  ±  0.43 mm, respectively for RCA. The average distance differences before and after registration were statistically significant (p < 0.001) for both LCA and RCA trees. The average distances for the distant phases with large displacements before registration, after AQSO registration, and finally after the CBSO registration were 2.85  ±  1.46, 1.62  ±  0.76, and 0.97  ±  0.43 mm, respectively for LCA, and 4.03  ±  2.36, 2.18  ±  1.11, and 0.97  ±  0.44 mm, respectively for RCA. The average distance differences between every two consecutive stages of registration were statistically significant. The corresponding phases of LCA and RCA trees were aligned to an average of less than 1 mm, providing a basis for a best-quality tree construction.

Digital breast tomosynthesis: observer performance of clustered microcalcification detection on breast phantom images acquired with an experimental system using variable scan angles, angular increments, and number of projection views

PURPOSE

To investigate the dependence of microcalcification cluster detectability on tomographic scan angle, angular increment, and number of projection views acquired at digital breast tomosynthesis ( DBT digital breast tomosynthesis ).

MATERIALS AND METHODS

A prototype DBT digital breast tomosynthesis system operated in step-and-shoot mode was used to image breast phantoms. Four 5-cm-thick phantoms embedded with 81 simulated microcalcification clusters of three speck sizes (subtle, medium, and obvious) were imaged by using a rhodium target and rhodium filter with 29 kV, 50 mAs, and seven acquisition protocols. Fixed angular increments were used in four protocols (denoted as scan angle, angular increment, and number of projection views, respectively: 16°, 1°, and 17; 24°, 3°, and nine; 30°, 3°, and 11; and 60°, 3°, and 21), and variable increments were used in three (40°, variable, and 13; 40°, variable, and 15; and 60°, variable, and 21). The reconstructed DBT digital breast tomosynthesis images were interpreted by six radiologists who located the microcalcification clusters and rated their conspicuity.

RESULTS

The mean sensitivity for detection of subtle clusters ranged from 80% (22.5 of 28) to 96% (26.8 of 28) for the seven DBT digital breast tomosynthesis protocols; the highest sensitivity was achieved with the 16°, 1°, and 17 protocol (96%), but the difference was significant only for the 60°, 3°, and 21 protocol (80%, P < .002) and did not reach significance for the other five protocols (P = .01-.15). The mean sensitivity for detection of medium and obvious clusters ranged from 97% (28.2 of 29) to 100% (24 of 24), but the differences fell short of significance (P = .08 to >.99). The conspicuity of subtle and medium clusters with the 16°, 1°, and 17 protocol was rated higher than those with other protocols; the differences were significant for subtle clusters with the 24°, 3°, and nine protocol and for medium clusters with 24°, 3°, and nine; 30°, 3°, and 11; 60°, 3° and 21; and 60°, variable, and 21 protocols (P < .002).

CONCLUSION

With imaging that did not include x-ray source motion or patient motion during acquisition of the projection views, narrow-angle DBT digital breast tomosynthesis provided higher sensitivity and conspicuity than wide-angle DBT digital breast tomosynthesis for subtle microcalcification clusters.

CT urography: segmentation of urinary bladder using CLASS with local contour refinement

We are developing a computerized system for bladder segmentation on CT urography (CTU), as a critical component for computer-aided detection of bladder cancer. The presence of regions filled with intravenous contrast and without contrast presents a challenge for bladder segmentation. Previously, we proposed a conjoint level set analysis and segmentation system (CLASS). In case the bladder is partially filled with contrast, CLASS segments the non-contrast (NC) region and the contrast-filled (C) region separately and automatically conjoins the NC and C region contours; however, inaccuracies in the NC and C region contours may cause the conjoint contour to exclude portions of the bladder. To alleviate this problem, we implemented a local contour refinement (LCR) method that exploits model-guided refinement (MGR) and energy-driven wavefront propagation (EDWP). MGR propagates the C region contours if the level set propagation in the C region stops prematurely due to substantial non-uniformity of the contrast. EDWP with regularized energies further propagates the conjoint contours to the correct bladder boundary. EDWP uses changes in energies, smoothness criteria of the contour, and previous slice contour to determine when to stop the propagation, following decision rules derived from training. A data set of 173 cases was collected for this study: 81 cases in the training set (42 lesions, 21 wall thickenings, 18 normal bladders) and 92 cases in the test set (43 lesions, 36 wall thickenings, 13 normal bladders). For all cases, 3D hand segmented contours were obtained as reference standard and used for the evaluation of the computerized segmentation accuracy. For CLASS with LCR, the average volume intersection ratio, average volume error, absolute average volume error, average minimum distance and Jaccard index were 84.2 ± 11.4%, 8.2 ± 17.4%, 13.0 ± 14.1%, 3.5 ± 1.9 mm, 78.8 ± 11.6%, respectively, for the training set and 78.0 ± 14.7%, 16.4 ± 16.9%, 18.2 ± 15.0%, 3.8 ± 2.3 mm, 73.8 ± 13.4% respectively, for the test set. With CLASS only, the corresponding values were 75.1 ± 13.2%, 18.7 ± 19.5%, 22.5 ± 14.9%, 4.3 ± 2.2 mm, 71.0 ± 12.6%, respectively, for the training set and 67.3 ± 14.3%, 29.3 ± 15.9%, 29.4 ± 15.6%, 4.9 ± 2.6 mm, 65.0 ± 13.3%, respectively, for the test set. The differences between the two methods for all five measures were statistically significant (p < 0.001) for both the training and test sets. The results demonstrate the potential of CLASS with LCR for segmentation of the bladder.

Evaluation of computer-aided detection and diagnosis systems

Computer-aided detection and diagnosis (CAD) systems are increasingly being used as an aid by clinicians for detection and interpretation of diseases. Computer-aided detection systems mark regions of an image that may reveal specific abnormalities and are used to alert clinicians to these regions during image interpretation. Computer-aided diagnosis systems provide an assessment of a disease using image-based information alone or in combination with other relevant diagnostic data and are used by clinicians as a decision support in developing their diagnoses. While CAD systems are commercially available, standardized approaches for evaluating and reporting their performance have not yet been fully formalized in the literature or in a standardization effort. This deficiency has led to difficulty in the comparison of CAD devices and in understanding how the reported performance might translate into clinical practice. To address these important issues, the American Association of Physicists in Medicine (AAPM) formed the Computer Aided Detection in Diagnostic Imaging Subcommittee (CADSC), in part, to develop recommendations on approaches for assessing CAD system performance. The purpose of this paper is to convey the opinions of the AAPM CADSC members and to stimulate the development of consensus approaches and "best practices" for evaluating CAD systems. Both the assessment of a standalone CAD system and the evaluation of the impact of CAD on end-users are discussed. It is hoped that awareness of these important evaluation elements and the CADSC recommendations will lead to further development of structured guidelines for CAD performance assessment. Proper assessment of CAD system performance is expected to increase the understanding of a CAD system's effectiveness and limitations, which is expected to stimulate further research and development efforts on CAD technologies, reduce problems due to improper use, and eventually improve the utility and efficacy of CAD in clinical practice.

Computer-aided detection of clustered microcalcifications in multiscale bilateral filtering regularized reconstructed digital breast tomosynthesis volume

PURPOSE

Develop a computer-aided detection (CADe) system for clustered microcalcifications in digital breast tomosynthesis (DBT) volume enhanced with multiscale bilateral filtering (MSBF) regularization.

METHODS

With Institutional Review Board approval and written informed consent, two-view DBT of 154 breasts, of which 116 had biopsy-proven microcalcification (MC) clusters and 38 were free of MCs, was imaged with a General Electric GEN2 prototype DBT system. The DBT volumes were reconstructed with MSBF-regularized simultaneous algebraic reconstruction technique (SART) that was designed to enhance MCs and reduce background noise while preserving the quality of other tissue structures. The contrast-to-noise ratio (CNR) of MCs was further improved with enhancement-modulated calcification response (EMCR) preprocessing, which combined multiscale Hessian response to enhance MCs by shape and bandpass filtering to remove the low-frequency structured background. MC candidates were then located in the EMCR volume using iterative thresholding and segmented by adaptive region growing. Two sets of potential MC objects, cluster centroid objects and MC seed objects, were generated and the CNR of each object was calculated. The number of candidates in each set was controlled based on the breast volume. Dynamic clustering around the centroid objects grouped the MC candidates to form clusters. Adaptive criteria were designed to reduce false positive (FP) clusters based on the size, CNR values and the number of MCs in the cluster, cluster shape, and cluster based maximum intensity projection. Free-response receiver operating characteristic (FROC) and jackknife alternative FROC (JAFROC) analyses were used to assess the performance and compare with that of a previous study.

RESULTS

Unpaired two-tailed t-test showed a significant increase (p < 0.0001) in the ratio of CNRs for MCs with and without MSBF regularization compared to similar ratios for FPs. For view-based detection, a sensitivity of 85% was achieved at an FP rate of 2.16 per DBT volume. For case-based detection, a sensitivity of 85% was achieved at an FP rate of 0.85 per DBT volume. JAFROC analysis showed a significant improvement in the performance of the current CADe system compared to that of our previous system (p = 0.003).

CONCLUSIONS

MBSF regularized SART reconstruction enhances MCs. The enhancement in the signals, in combination with properly designed adaptive threshold criteria, effective MC feature analysis, and false positive reduction techniques, leads to a significant improvement in the detection of clustered MCs in DBT.

Segmentation of urinary bladder in CT urography

We are developing a Conjoint Level set Analysis and Segmentation System (CLASS) for bladder segmentation on CTU, which is a critical component for computer aided diagnosis of bladder cancer. A challenge for bladder segmentation is the presence of regions without contrast (NC) and filled with IV contrast (C). According to the characteristics of the bladder in CTU, CLASS is designed to perform number tasks such as preprocessing, initial segmentation, 3D and 2D level set segmentation and post-processing. CLASS segments separately the NC and the C regions of the bladder. In the post-processing stage the final contour is obtained based on the union of the NC and C contours. 70 bladders were segmented. Of the 70 bladders 31 contained lesions, 24 contained wall thickening, and 15 were normal. The performance of CLASS was assessed by rating the quality of the contours on a 5-point scale (1="very poor", 3="fair", 5="excellent"). The average quality ratings for the 12 completely no contrast (NC) and 5 completely contrast-filled (C) bladder contours were 3.3±1.0 and 3.4±0.5, respectively. The average quality ratings for the 53 NC and 53 C regions of the 53 partially contrast-filled bladders were 4.0±0.7 and 4.0±1.0, respectively. Quality ratings of 4 or above were given for 87% (46/53) NC and 77% (41/53) C regions. Only 4% (2/53) NC and 9% (5/53) C regions had ratings under 3. After combining the NC and C contours for each of the 70 bladders, 66% (46/70) had quality ratings of 4 or above. Only 6% (4/70) had ratings under 3. The average quality rating was 3.8±0.7. The results demonstrate the potential of CLASS for automated segmentation of the bladder.

Auto-initialized cascaded level set (AI-CALS) segmentation of bladder lesions on multidetector row CT urography

RATIONALE AND OBJECTIVES

To develop a computerized system for segmentation of bladder lesions on computed tomography urography (CTU) scans for detection and characterization of bladder cancer.

MATERIALS AND METHODS

We have developed an auto-initialized cascaded level set method to perform bladder lesion segmentation. The segmentation performance was evaluated on a preliminary dataset including 28 CTU scans from 28 patients collected retrospectively with institutional review board approval. The bladders were partially filled with intravenous contrast material. The lesions were located fully or partially within the contrast-enhanced area or in the non-contrast-enhanced area of the bladder. An experienced abdominal radiologist marked 28 lesions (14 malignant and 14 benign) with bounding boxes that served as input to the automated segmentation system and assigned a difficulty rating on a scale of 1 to 5 (5 = most subtle) to each lesion. The contours from automated segmentation were compared to three-dimensional contours manually drawn by the radiologist. Three performance metric measures were used for comparison. In addition, the automated segmentation quality was assessed by an expert panel of two experienced radiologists, who provided quality ratings of the contours on a scale from 1 to 10 (10 = excellent).

RESULTS

The average volume intersection ratio, the average absolute volume error, and the average distance measure were 67.2 ± 16.9%, 27.3 ± 26.9%, and 2.89 ± 1.69 mm, respectively. Of the 28 segmentations, 18 were given quality ratings of 8 or above. The average rating was 7.9 ± 1.5. The average quality ratings for lesions with difficulty ratings of 1, 2, 3, and 4 were 8.8 ± 0.9, 7.9 ± 1.8, 7.4 ± 0.9, and 6.6 ± 1.5, respectively.

CONCLUSION

Our preliminary study demonstrates the feasibility of using the three-dimensional level set method for segmenting bladder lesions in CTU scans.

A similarity study of content-based image retrieval system for breast cancer using decision tree

PURPOSE

We are developing a decision tree content-based image retrieval (DTCBIR) CADx system to assist radiologists in characterization of breast masses on ultrasound images.

METHODS

Three DTCBIR configurations, including decision tree with boosting (DTb), decision tree with full leaf features (DTL), and decision tree with selected leaf features (DTLs) were compared. For DTb, features of a query mass were combined first into a merged feature score and then masses with similar scores were retrieved. For DTL and DTLs, similar masses were retrieved based on the Euclidean distance between feature vectors of the query and those of selected references. For each DTCBIR configuration, we investigated the use of full feature set and subset of features selected by the stepwise linear discriminant analysis (LDA) and simplex optimization method, resulting in six retrieval methods and selected five, DTb-lda, DTL-lda, DTb-full, DTL-full, and DTLs-full, for the observer study. Three MQSA radiologists rated similarities between the query mass and computer-retrieved three most similar masses using nine-point similarity scale (9 = very similar).

RESULTS

For DTb-lda, DTL-lda, DTb-full, DTL-full, and DTLs-full, average A(z) values were 0.90 ± 0.03, 0.85 ± 0.04, 0.87 ± 0.04, 0.79 ± 0.05, and 0.71 ± 0.06, respectively, and average similarity ratings were 5.00, 5.41, 4.96, 5.33, and 5.13, respectively.

CONCLUSIONS

The DTL-lda is a promising DTCBIR CADx configuration which had simple tree structure, good classification performance, and highest similarity rating.

Image quality of microcalcifications in digital breast tomosynthesis: effects of projection-view distributions

PURPOSE

To analyze the effects of projection-view (PV) distribution on the contrast and spatial blurring of microcalcifications on the tomosynthesized slices (X-Y plane) and along the depth (Z) direction for the same radiation dose in digital breast tomosynthesis (DBT).

METHODS

A GE GEN2 prototype DBT system was used for acquisition of DBT scans. The system acquires PV images from 21 angles in 3° increments over a ±30° range. From these acquired PV images, the authors selected six subsets of PV images to simulate DBT of different angular ranges and angular increments. The number of PV images in each subset was fixed at 11 to simulate a constant total dose. These different PV distributions were subjectively divided into three categories: uniform group, nonuniform central group, and nonuniform extreme group with different angular ranges and angular increments. The simultaneous algebraic reconstruction technique (SART) was applied to each subset to reconstruct the DBT slices. A selective diffusion regularization method was employed to suppress noise. The image quality of microcalcifications in the reconstructed DBTs with different PV distributions was compared using the DBT scans of an American College of Radiology phantom and three human subjects. The contrast-to-noise ratio (CNR) and the full width at half maximum (FWHM) of the line profiles of microcalcifications within their in-focus DBT slices (parallel to detector plane) and the FWHMs of the interplane artifact spread function (ASF) in the Z-direction (perpendicular to detector plane) were used as image quality measures.

RESULTS

The results indicate that DBT acquired with a large angular range or, for an equal angular range,with a large fraction of PVs at large angles yielded superior ASF with smaller FWHM in the Z-direction. PV distributions with a narrow angular range or a large fraction of PVs at small angles had stronger interplane artifacts. In the X-Y focal planes, the effect of PV distributions on spatial blurring depended on the directions. In the X-direction (perpendicular to the chestwall), the normalized line profiles of the calcifications reconstructed with the different PV distributions were similar in terms of FWHM; the differences in the FWHMs between the different PV distributions were less than half a pixel. In the Y-direction (x-ray source motion), the normalized line profiles of the calcifications reconstructed with PVs acquired with a narrow angular range or a large fraction of PVs at small angles had smaller FWHMs and thus less blurring of the line profiles. In addition, PV distributions with a narrow angular range or a large fraction of PVs at small angles yielded slightly higher CNR than those with a wide angular range for small, subtle microcalcifications; however, PV distributions had no obvious effect on CNR for relatively large microcalcifications.

CONCLUSIONS

PV distributions affect the image quality of DBT. The relative importance of the impact depends on the characteristics of the signal and the direction (perpendicular or parallel) relative to the direction of x-ray source motion. For a given number of PVs, the angular range and the distribution of the PVs affect the degree of in-plane and interplane blurring in opposite ways. The design of the scan parameters of tomosynthesis systems would require proper consideration of the characteristics of the signals of interest and the potential trade-off of the image quality of different types of signals.