Multiview-based computer-aided detection scheme for breast masses

A multi-stage fusion framework to classify breast lesions using deep learning and radiomics features computed from four-view mammograms

BACKGROUND

Developing computer aided diagnosis (CAD) schemes of mammograms to classify between malignant and benign breast lesions has attracted a lot of research attention over the last several decades. However, unlike radiologists who make diagnostic decisions based on the fusion of image features extracted from multi-view mammograms, most CAD schemes are single-view-based schemes, which limit CAD performance and clinical utility.

PURPOSE

This study aims to develop and test a novel CAD framework that optimally fuses information extracted from ipsilateral views of bilateral mammograms using both deep transfer learning (DTL) and radiomics feature extraction methods.

METHODS

An image dataset containing 353 benign and 611 malignant cases is assembled. Each case contains four images: the craniocaudal (CC) and mediolateral oblique (MLO) view of the left and right breast. First, we extract four matching regions of interest (ROIs) from images that surround centers of two suspicious lesion regions seen in CC and MLO views, as well as matching ROIs in the contralateral breasts. Next, the handcrafted radiomics (HCRs) features and VGG16 model-generated automated features are extracted from each ROI resulting in eight feature vectors. Then, after reducing feature dimensionality and quantifying the bilateral and ipsilateral asymmetry of four ROIs to yield four new feature vectors, we test four fusion methods to build three support vector machine (SVM) classifiers by an optimal fusion of asymmetrical image features extracted from four view images.

RESULTS

Using a 10-fold cross-validation method, results show that a SVM classifier trained using an optimal fusion of four view images yields the highest classification performance (AUC = 0.876 ± 0.031), which significantly outperforms SVM classifiers trained using one projection view alone, AUC = 0.817 ± 0.026 and 0.792 ± 0.026 for the CC and MLO view of bilateral mammograms, respectively (p < 0.001).

CONCLUSIONS

The study demonstrates that the shift from single-view CAD to four-view CAD and the inclusion of both DTL and radiomics features significantly increases CAD performance in distinguishing between malignant and benign breast lesions.

Transformers Improve Breast Cancer Diagnosis from Unregistered Multi-View Mammograms

Deep convolutional neural networks (CNNs) have been widely used in various medical imaging tasks. However, due to the intrinsic locality of convolution operations, CNNs generally cannot model long-range dependencies well, which are important for accurately identifying or mapping corresponding breast lesion features computed from unregistered multiple mammograms. This motivated us to leverage the architecture of Multi-view Vision Transformers to capture long-range relationships of multiple mammograms from the same patient in one examination. For this purpose, we employed local transformer blocks to separately learn patch relationships within four mammograms acquired from two-view (CC/MLO) of two-side (right/left) breasts. The outputs from different views and sides were concatenated and fed into global transformer blocks, to jointly learn patch relationships between four images representing two different views of the left and right breasts. To evaluate the proposed model, we retrospectively assembled a dataset involving 949 sets of mammograms, which included 470 malignant cases and 479 normal or benign cases. We trained and evaluated the model using a five-fold cross-validation method. Without any arduous preprocessing steps (e.g., optimal window cropping, chest wall or pectoral muscle removal, two-view image registration, etc.), our four-image (two-view-two-side) transformer-based model achieves case classification performance with an area under ROC curve (AUC = 0.818 ± 0.039), which significantly outperforms AUC = 0.784 ± 0.016 achieved by the state-of-the-art multi-view CNNs (p = 0.009). It also outperforms two one-view-two-side models that achieve AUC of 0.724 ± 0.013 (CC view) and 0.769 ± 0.036 (MLO view), respectively. The study demonstrates the potential of using transformers to develop high-performing computer-aided diagnosis schemes that combine four mammograms.

Development and Assessment of a New Global Mammographic Image Feature Analysis Scheme to Predict Likelihood of Malignant Cases

This study aims to develop and evaluate a new computer-aided diagnosis (CADx) scheme based on analysis of global mammographic image features to predict likelihood of cases being malignant. An image dataset involving 1,959 cases was retrospectively assembled. Suspicious lesions were detected and biopsied in each case. Among them, 737 cases are malignant and 1,222 are benign. Each case includes four mammograms of craniocaudal and mediolateral oblique view of left and right breasts. CADx scheme is applied to pre-process mammograms, generate two image maps in frequency domain using discrete cosine transform and fast Fourier transform, compute bilateral image feature differences from left and right breasts, and apply a support vector machine (SVM) to predict likelihood of the case being malignant. Three sub-groups of image features were computed from the original mammograms and two transformation maps. Four SVMs using three sub-groups of image features and fusion of all features were trained and tested using a 10-fold cross-validation method. The computed areas under receiver operating characteristic curves (AUCs) range from 0.85 to 0.91 using image features computed from one of three sub-groups, respectively. By fusion of all image features computed in three sub-groups, the fourth SVM yields a significantly higher performance with AUC = 0.96±0.01 (p<0.01). This study demonstrates feasibility of developing a new global image feature analysis based CADx scheme of mammograms with high performance. By avoiding difficulty and possible errors in breast lesion segmentation, this new CADx approach is more efficient in development and potentially more robust in future application.

Developing global image feature analysis models to predict cancer risk and prognosis

In order to develop precision or personalized medicine, identifying new quantitative imaging markers and building machine learning models to predict cancer risk and prognosis has been attracting broad research interest recently. Most of these research approaches use the similar concepts of the conventional computer-aided detection schemes of medical images, which include steps in detecting and segmenting suspicious regions or tumors, followed by training machine learning models based on the fusion of multiple image features computed from the segmented regions or tumors. However, due to the heterogeneity and boundary fuzziness of the suspicious regions or tumors, segmenting subtle regions is often difficult and unreliable. Additionally, ignoring global and/or background parenchymal tissue characteristics may also be a limitation of the conventional approaches. In our recent studies, we investigated the feasibility of developing new computer-aided schemes implemented with the machine learning models that are trained by global image features to predict cancer risk and prognosis. We trained and tested several models using images obtained from full-field digital mammography, magnetic resonance imaging, and computed tomography of breast, lung, and ovarian cancers. Study results showed that many of these new models yielded higher performance than other approaches used in current clinical practice. Furthermore, the computed global image features also contain complementary information from the features computed from the segmented regions or tumors in predicting cancer prognosis. Therefore, the global image features can be used alone to develop new case-based prediction models or can be added to current tumor-based models to increase their discriminatory power.

Developing a new case based computer-aided detection scheme and an adaptive cueing method to improve performance in detecting mammographic lesions

The purpose of this study is to evaluate a new method to improve performance of computer-aided detection (CAD) schemes of screening mammograms with two approaches. In the first approach, we developed a new case based CAD scheme using a set of optimally selected global mammographic density, texture, spiculation, and structural similarity features computed from all four full-field digital mammography images of the craniocaudal (CC) and mediolateral oblique (MLO) views by using a modified fast and accurate sequential floating forward selection feature selection algorithm. Selected features were then applied to a 'scoring fusion' artificial neural network classification scheme to produce a final case based risk score. In the second approach, we combined the case based risk score with the conventional lesion based scores of a conventional lesion based CAD scheme using a new adaptive cueing method that is integrated with the case based risk scores. We evaluated our methods using a ten-fold cross-validation scheme on 924 cases (476 cancer and 448 recalled or negative), whereby each case had all four images from the CC and MLO views. The area under the receiver operating characteristic curve was AUC  =  0.793  ±  0.015 and the odds ratio monotonically increased from 1 to 37.21 as CAD-generated case based detection scores increased. Using the new adaptive cueing method, the region based and case based sensitivities of the conventional CAD scheme at a false positive rate of 0.71 per image increased by 2.4% and 0.8%, respectively. The study demonstrated that supplementary information can be derived by computing global mammographic density image features to improve CAD-cueing performance on the suspicious mammographic lesions.

A new and fast image feature selection method for developing an optimal mammographic mass detection scheme

PURPOSE

Selecting optimal features from a large image feature pool remains a major challenge in developing computer-aided detection (CAD) schemes of medical images. The objective of this study is to investigate a new approach to significantly improve efficacy of image feature selection and classifier optimization in developing a CAD scheme of mammographic masses.

METHODS

An image dataset including 1600 regions of interest (ROIs) in which 800 are positive (depicting malignant masses) and 800 are negative (depicting CAD-generated false positive regions) was used in this study. After segmentation of each suspicious lesion by a multilayer topographic region growth algorithm, 271 features were computed in different feature categories including shape, texture, contrast, isodensity, spiculation, local topological features, as well as the features related to the presence and location of fat and calcifications. Besides computing features from the original images, the authors also computed new texture features from the dilated lesion segments. In order to select optimal features from this initial feature pool and build a highly performing classifier, the authors examined and compared four feature selection methods to optimize an artificial neural network (ANN) based classifier, namely: (1) Phased Searching with NEAT in a Time-Scaled Framework, (2) A sequential floating forward selection (SFFS) method, (3) A genetic algorithm (GA), and (4) A sequential forward selection (SFS) method. Performances of the four approaches were assessed using a tenfold cross validation method.

RESULTS

Among these four methods, SFFS has highest efficacy, which takes 3%-5% of computational time as compared to GA approach, and yields the highest performance level with the area under a receiver operating characteristic curve (AUC) = 0.864 ± 0.034. The results also demonstrated that except using GA, including the new texture features computed from the dilated mass segments improved the AUC results of the ANNs optimized using other three feature selection methods. In addition, among 271 features, the shape, local morphological features, fat and calcification based features were the most frequently selected features to build ANNs.

CONCLUSIONS

Although conventional GA is a powerful tool in optimizing classifiers used in CAD schemes of medical images, it is very computationally intensive. This study demonstrated that using a new SFFS based approach enabled to significantly improve efficacy of image feature selection for developing CAD schemes.

Computer-aided detection system for clustered microcalcifications in digital breast tomosynthesis using joint information from volumetric and planar projection images

We propose a novel approach for the detection of microcalcification clusters (MCs) using joint information from digital breast tomosynthesis (DBT) volume and planar projection (PPJ) image. A data set of 307 DBT views was collected with IRB approval using a prototype DBT system. The system acquires 21 projection views (PVs) from a wide tomographic angle of 60° (60°-21PV) at about twice the dose of a digital mammography (DM) system, which allows us the flexibility of simulating other DBT acquisition geometries using a subset of the PVs. In this study, we simulated a 30° DBT geometry using the central 11 PVs (30°-11PV). The narrower tomographic angle is closer to DBT geometries commercially available or under development and the dose is matched approximately to that of a DM. We developed a new joint-CAD system for detection of clustered microcalcifications. The DBT volume was reconstructed with a multiscale bilateral filtering regularized method and a PPJ image was generated from the reconstructed volume. Task-specific detection strategies were designed to combine information from the DBT volume and the PPJ image. The data set was divided into a training set (127 views with MCs) and an independent test set (104 views with MCs and 76 views without MCs). The joint-CAD system outperformed the individual CAD systems for DBT volume or PPJ image alone; the differences in the test performances were statistically significant (p  <  0.05) using JAFROC analysis.

A new approach to develop computer-aided detection schemes of digital mammograms

The purpose of this study is to develop a new global mammographic image feature analysis based computer-aided detection (CAD) scheme and evaluate its performance in detecting positive screening mammography examinations. A dataset that includes images acquired from 1896 full-field digital mammography (FFDM) screening examinations was used in this study. Among them, 812 cases were positive for cancer and 1084 were negative or benign. After segmenting the breast area, a computerized scheme was applied to compute 92 global mammographic tissue density based features on each of four mammograms of the craniocaudal (CC) and mediolateral oblique (MLO) views. After adding three existing popular risk factors (woman's age, subjectively rated mammographic density, and family breast cancer history) into the initial feature pool, we applied a sequential forward floating selection feature selection algorithm to select relevant features from the bilateral CC and MLO view images separately. The selected CC and MLO view image features were used to train two artificial neural networks (ANNs). The results were then fused by a third ANN to build a two-stage classifier to predict the likelihood of the FFDM screening examination being positive. CAD performance was tested using a ten-fold cross-validation method. The computed area under the receiver operating characteristic curve was AUC = 0.779   ±   0.025 and the odds ratio monotonically increased from 1 to 31.55 as CAD-generated detection scores increased. The study demonstrated that this new global image feature based CAD scheme had a relatively higher discriminatory power to cue the FFDM examinations with high risk of being positive, which may provide a new CAD-cueing method to assist radiologists in reading and interpreting screening mammograms.

Optimization of Network Topology in Computer-Aided Detection Schemes Using Phased Searching with NEAT in a Time-Scaled Framework

In the field of computer-aided mammographic mass detection, many different features and classifiers have been tested. Frequently, the relevant features and optimal topology for the artificial neural network (ANN)-based approaches at the classification stage are unknown, and thus determined by trial-and-error experiments. In this study, we analyzed a classifier that evolves ANNs using genetic algorithms (GAs), which combines feature selection with the learning task. The classifier named "Phased Searching with NEAT in a Time-Scaled Framework" was analyzed using a dataset with 800 malignant and 800 normal tissue regions in a 10-fold cross-validation framework. The classification performance measured by the area under a receiver operating characteristic (ROC) curve was 0.856 ± 0.029. The result was also compared with four other well-established classifiers that include fixed-topology ANNs, support vector machines (SVMs), linear discriminant analysis (LDA), and bagged decision trees. The results show that Phased Searching outperformed the LDA and bagged decision tree classifiers, and was only significantly outperformed by SVM. Furthermore, the Phased Searching method required fewer features and discarded superfluous structure or topology, thus incurring a lower feature computational and training and validation time requirement. Analyses performed on the network complexities evolved by Phased Searching indicate that it can evolve optimal network topologies based on its complexification and simplification parameter selection process. From the results, the study also concluded that the three classifiers - SVM, fixed-topology ANN, and Phased Searching with NeuroEvolution of Augmenting Topologies (NEAT) in a Time-Scaled Framework - are performing comparably well in our mammographic mass detection scheme.

Reduction of false-positive recalls using a computerized mammographic image feature analysis scheme

The high false-positive recall rate is one of the major dilemmas that significantly reduce the efficacy of screening mammography, which harms a large fraction of women and increases healthcare cost. This study aims to investigate the feasibility of helping reduce false-positive recalls by developing a new computer-aided diagnosis (CAD) scheme based on the analysis of global mammographic texture and density features computed from four-view images. Our database includes full-field digital mammography (FFDM) images acquired from 1052 recalled women (669 positive for cancer and 383 benign). Each case has four images: two craniocaudal (CC) and two mediolateral oblique (MLO) views. Our CAD scheme first computed global texture features related to the mammographic density distribution on the segmented breast regions of four images. Second, the computed features were given to two artificial neural network (ANN) classifiers that were separately trained and tested in a ten-fold cross-validation scheme on CC and MLO view images, respectively. Finally, two ANN classification scores were combined using a new adaptive scoring fusion method that automatically determined the optimal weights to assign to both views. CAD performance was tested using the area under a receiver operating characteristic curve (AUC). The AUC = 0.793  ±  0.026 was obtained for this four-view CAD scheme, which was significantly higher at the 5% significance level than the AUCs achieved when using only CC (p = 0.025) or MLO (p = 0.0004) view images, respectively. This study demonstrates that a quantitative assessment of global mammographic image texture and density features could provide useful and/or supplementary information to classify between malignant and benign cases among the recalled cases, which may eventually help reduce the false-positive recall rate in screening mammography.

Association between computed tissue density asymmetry in bilateral mammograms and near-term breast cancer risk

This study investigated association between bilateral mammographic density asymmetry and near-term breast cancer risk. A data base of digital mammograms acquired from 690 women was retrospectively collected. All images were originally interpreted as negative by radiologists. During the next subsequent screening examinations (between 12 and 36 months later), 230 women were diagnosed positive for cancer, 230 were recalled for additional diagnostic workups and proved to be benign, and 230 remained negative (not recalled). We applied a computerized scheme to compute the differences of five image features between the left and right mammograms, and trained an artificial neural network (ANN) to compute a bilateral mammographic density asymmetry score. Odds ratios (ORs) were used to assess associations between the ANN-generated scores and risk of women having detectable cancers during the next screening examinations. A logistic regression method was applied to test for trend as a function of the increase in ANN-generated scores. The results were also compared with ORs computed using other existing cancer risk factors. The ORs showed an increasing risk trend with the increase in ANN-generated scores (from 1.00 to 9.07 between positive and negative case groups). The regression analysis also showed a significant increase trend in slope (p < 0.05). No significant increase trends of the ORs were found when using woman's age, subjectively rated breast density, or family history of breast cancer. This study demonstrated that the computed bilateral mammographic density asymmetry had potential to be used as a new risk factor to improve discriminatory power in predicting near-term risk of women developing breast cancer.

Optimization of breast mass classification using sequential forward floating selection (SFFS) and a support vector machine (SVM) model

PURPOSE

Improving radiologists' performance in classification between malignant and benign breast lesions is important to increase cancer detection sensitivity and reduce false-positive recalls. For this purpose, developing computer-aided diagnosis schemes has been attracting research interest in recent years. In this study, we investigated a new feature selection method for the task of breast mass classification.

METHODS

We initially computed 181 image features based on mass shape, spiculation, contrast, presence of fat or calcifications, texture, isodensity, and other morphological features. From this large image feature pool, we used a sequential forward floating selection (SFFS)-based feature selection method to select relevant features and analyzed their performance using a support vector machine (SVM) model trained for the classification task. On a database of 600 benign and 600 malignant mass regions of interest, we performed the study using a tenfold cross-validation method. Feature selection and optimization of the SVM parameters were conducted on the training subsets only.

RESULTS

The area under the receiver operating characteristic curve [Formula: see text] was obtained for the classification task. The results also showed that the most frequently selected features by the SFFS-based algorithm in tenfold iterations were those related to mass shape, isodensity, and presence of fat, which are consistent with the image features frequently used by radiologists in the clinical environment for mass classification. The study also indicated that accurately computing mass spiculation features from the projection mammograms was difficult, and failed to perform well for the mass classification task due to tissue overlap within the benign mass regions.

CONCLUSION

In conclusion, this comprehensive feature analysis study provided new and valuable information for optimizing computerized mass classification schemes that may have potential to be useful as a "second reader" in future clinical practice.

Prediction of near-term breast cancer risk based on bilateral mammographic feature asymmetry

RATIONALE AND OBJECTIVES

The objective of this study is to investigate the feasibility of predicting near-term risk of breast cancer development in women after a negative mammography screening examination. It is based on a statistical learning model that combines computerized image features related to bilateral mammographic tissue asymmetry and other clinical factors.

MATERIALS AND METHODS

A database of negative digital mammograms acquired from 994 women was retrospectively collected. In the next sequential screening examination (12 to 36 months later), 283 women were diagnosed positive for cancer, 349 were recalled for additional diagnostic workups and later proved to be benign, and 362 remain negative (not recalled). From an initial pool of 183 features, we applied a Sequential Forward Floating Selection feature selection method to search for effective features. Using 10 selected features, we developed and trained a support vector machine classification model to compute a cancer risk or probability score for each case. The area under the receiver operating characteristic curve and odds ratios (ORs) were used as the two performance assessment indices.

RESULTS

The area under the receiver operating characteristic curve = 0.725 ± 0.018 was obtained for positive and negative/benign case classification. The ORs showed an increasing risk trend with increasing model-generated risk scores (from 1.00 to 12.34, between positive and negative/benign case groups). Regression analysis of ORs also indicated a significant increase trend in slope (P = .006).

CONCLUSIONS

This study demonstrates that the risk scores computed by a new support vector machine model involving bilateral mammographic feature asymmetry have potential to assist the prediction of near-term risk of women for developing breast cancer.

Breast image analysis for risk assessment, detection, diagnosis, and treatment of cancer

The role of breast image analysis in radiologists' interpretation tasks in cancer risk assessment, detection, diagnosis, and treatment continues to expand. Breast image analysis methods include segmentation, feature extraction techniques, classifier design, biomechanical modeling, image registration, motion correction, and rigorous methods of evaluation. We present a review of the current status of these task-based image analysis methods, which are being developed for the various image acquisition modalities of mammography, tomosynthesis, computed tomography, ultrasound, and magnetic resonance imaging. Depending on the task, image-based biomarkers from such quantitative image analysis may include morphological, textural, and kinetic characteristics and may depend on accurate modeling and registration of the breast images. We conclude with a discussion of future directions.

Improving performance of computer-aided detection of masses by incorporating bilateral mammographic density asymmetry: an assessment

RATIONALE AND OBJECTIVES

Bilateral mammographic density asymmetry is a promising indicator in assessing risk of having or developing breast cancer. This study aims to assess the performance improvement of a computer-aided detection (CAD) scheme in detecting masses by incorporating bilateral mammographic density asymmetrical information.

MATERIALS AND METHODS

A testing dataset containing 2400 full-field digital mammograms (FFDM) acquired from 600 examination cases was established. Among them, 300 were positive cases with verified cancer associated with malignant masses and 300 were negative cases. Two computerized schemes were applied to process images of each case. The first single-image based CAD scheme detected suspicious mass regions and the second scheme computed average and difference of mammographic tissue density depicted between the left and right breast. A fusion method based on rotation of the CAD scoring projection reference axis was then applied to combine CAD-generated mass detection scores and either the computed average or difference (asymmetry) of bilateral mammographic density scores. The CAD performance levels with and without incorporating mammographic density information were evaluated and compared using a free-response receiver operating characteristic type data analysis method.

RESULTS

CAD achieved a case-based mass detection sensitivity of 0.74 and a region-based sensitivity of 0.56 at a false-positive rate of 0.25 per image. By fusing the CAD and bilateral mammographic density asymmetry scores, the case-based and region-based sensitivity levels of the CAD scheme were increased to 0.84 and 0.69, respectively, at the same false-positive rate. Fusion with average mammographic density only slightly increased CAD sensitivity to 0.75 (case-based) and 0.57 (region-based).

CONCLUSIONS

This study indicated that 1) bilateral mammographic density asymmetry was a stronger indicator of the case depicting suspicious masses than the average density computed from two breasts and 2) fusion between the conventional CAD scores and bilateral mammographic density asymmetry information could substantially increase CAD performance in mass detection.

Improving the performance of computer-aided detection of subtle breast masses using an adaptive cueing method

Current computer-aided detection (CAD) schemes for detecting mammographic masses have several limitations including high correlation with radiologists' detection and cueing most subtle masses only on one view. To increase CAD sensitivity in cueing more subtle masses that are likely missed and/or overlooked by radiologists without increasing false-positive rates, we investigated a new case-dependent cueing method by combining the original CAD-generated detection scores with a computed bilateral mammographic density asymmetry index. Using the new method, we adaptively raise the CAD-generated scores of the regions detected on 'high-risk' cases to cue more subtle mass regions and reduce the CAD scores of the regions detected on 'low-risk' cases to discard more false-positive regions. A testing dataset involving 78 positive and 338 negative cases was used to test this adaptive cueing method. Each positive case involves two sequential examinations in which the mass was detected in 'current' examination and missed in 'prior' examination but detected in a retrospective review by radiologists. Applying to this dataset, a pre-optimized CAD scheme yielded 75% case-based and 55% region-based sensitivity on 'current' examinations at a false-positive rate of 0.25 per image. CAD sensitivity was reduced to 42% (case based) and 27% (region based) on 'prior' examinations. Using the new cueing method, case-based and region-based sensitivity could maximally increase 9% and 33% on the 'prior' examinations, respectively. The percentages of the masses cued on two views also increased from 27% to 65%. The study demonstrated that using this adaptive cueing method enabled us to help CAD cue more subtle cancers without increasing the false-positive cueing rate.

Correlating locations in ipsilateral breast tomosynthesis views using an analytical hemispherical compression model

To improve cancer detection in mammography, breast examinations usually consist of two views per breast. In order to combine information from both views, corresponding regions in the views need to be matched. In 3D digital breast tomosynthesis (DBT), this may be a difficult and time-consuming task for radiologists, because many slices have to be inspected individually. For multiview computer-aided detection (CAD) systems, matching corresponding regions is an essential step that needs to be automated. In this study, we developed an automatic method to quickly estimate corresponding locations in ipsilateral tomosynthesis views by applying a spatial transformation. First we match a model of a compressed breast to the tomosynthesis view containing a point of interest. Then we estimate the location of the corresponding point in the ipsilateral view by assuming that this model was decompressed, rotated and compressed again. In this study, we use a relatively simple, elastically deformable sphere model to obtain an analytical solution for the transformation in a given DBT case. We investigate three different methods to match the compression model to the data by using automatic segmentation of the pectoral muscle, breast tissue and nipple. For validation, we annotated 208 landmarks in both views of a total of 146 imaged breasts of 109 different patients and applied our method to each location. The best results are obtained by using the centre of gravity of the breast to define the central axis of the model, around which the breast is assumed to rotate between views. Results show a median 3D distance between the actual location and the estimated location of 14.6 mm, a good starting point for a registration method or a feature-based local search method to link suspicious regions in a multiview CAD system. Approximately half of the estimated locations are at most one slice away from the actual location, which makes the method useful as a mammographic workstation tool for radiologists to interactively find corresponding locations in ipsilateral tomosynthesis views.

Computer-aided detection of breast masses: four-view strategy for screening mammography

PURPOSE

To improve the performance of a computer-aided detection (CAD) system for mass detection by using four-view information in screening mammography.

METHODS

The authors developed a four-view CAD system that emulates radiologists' reading by using the craniocaudal and mediolateral oblique views of the ipsilateral breast to reduce false positives (FPs) and the corresponding views of the contralateral breast to detect asymmetry. The CAD system consists of four major components: (1) Initial detection of breast masses on individual views, (2) information fusion of the ipsilateral views of the breast (referred to as two-view analysis), (3) information fusion of the corresponding views of the contralateral breast (referred to as bilateral analysis), and (4) fusion of the four-view information with a decision tree. The authors collected two data sets for training and testing of the CAD system: A mass set containing 389 patients with 389 biopsy-proven masses and a normal set containing 200 normal subjects. All cases had four-view mammograms. The true locations of the masses on the mammograms were identified by an experienced MQSA radiologist. The authors randomly divided the mass set into two independent sets for cross validation training and testing. The overall test performance was assessed by averaging the free response receiver operating characteristic (FROC) curves of the two test subsets. The FP rates during the FROC analysis were estimated by using the normal set only. The jackknife free-response ROC (JAFROC) method was used to estimate the statistical significance of the difference between the test FROC curves obtained with the single-view and the four-view CAD systems.

RESULTS

Using the single-view CAD system, the breast-based test sensitivities were 58% and 77% at the FP rates of 0.5 and 1.0 per image, respectively. With the four-view CAD system, the breast-based test sensitivities were improved to 76% and 87% at the corresponding FP rates, respectively. The improvement was found to be statistically significant (p < 0.0001) by JAFROC analysis.

CONCLUSIONS

The four-view information fusion approach that emulates radiologists' reading strategy significantly improves the performance of breast mass detection of the CAD system in comparison with the single-view approach.

Computerized prediction of risk for developing breast cancer based on bilateral mammographic breast tissue asymmetry

This study developed and assessed a computerized scheme to detect breast abnormalities and predict the risk of developing cancer based on bilateral mammographic tissue asymmetry. A digital mammography database of 100 randomly selected negative cases and 100 positive cases for having high-risk of developing breast cancer was established. Each case includes four images of cranio-caudal (CC) and medio-lateral oblique (MLO) views of the left and right breast. To detect bilateral mammographic tissue asymmetry, a pool of 20 computed features was assembled. A genetic algorithm was applied to select optimal features and build an artificial neural network based classifier to predict the likelihood of a test case being positive. The leave-one-case-out validation method was used to evaluate the classifier performance. Several approaches were investigated to improve the classification performance including extracting asymmetrical tissue features from either selected regions of interests or the entire segmented breast area depicted on bilateral images in one view, and the fusion of classification results from two views. The results showed that (1) using the features computed from the entire breast area, the classifier yielded the higher performance than using ROIs, and (2) using a weighted average fusion method, the classifier achieved the highest performance with the area under ROC curve of 0.781±0.023. At 90% specificity, the scheme detected 58.3% of high-risk cases in which cancers developed and verified 6-18 months later. The study demonstrated the feasibility of applying a computerized scheme to detect cases with high risk of developing breast cancer based on computer-detected bilateral mammographic tissue asymmetry.

Optimizing Case-based detection performance in a multiview CAD system for mammography

When reading mammograms, radiologists combine information from multiple views to detect abnormalities. Most computer-aided detection (CAD) systems, however, use primitive methods for inclusion of multiview context or analyze each view independently. In previous research it was found that in mammography lesion-based detection performance of CAD systems can be improved when correspondences between MLO and CC views are taken into account. However, detection at case level detection did not improve. In this paper, we propose a new learning method for multiview CAD systems, which is aimed at optimizing case-based detection performance. The method builds on a single-view lesion detection system and a correspondence classifier. The latter provides class probabilities for the various types of region pairs and correspondence features. The correspondence classifier output is used to bias the selection of training patterns for a multiview CAD system. In this way training can be forced to focus on optimization of case-based detection performance. The method is applied to the problem of detecting malignant masses and architectural distortions. Experiments involve 454 mammograms consisting of four views with a malignant region visible in at least one of the views. To evaluate performance, five-fold cross validation and FROC analysis was performed. Bootstrapping was used for statistical analysis. A significant increase of case-based detection performance was found when the proposed method was used. Mean sensitivity increased by 4.7% in the range of 0.01-0.5 false positives per image.

Computer-aided detection of breast masses depicted on full-field digital mammograms: a performance assessment

OBJECTIVES

To investigate the feasibility of converting a computer-aided detection (CAD) scheme for digitised screen-film mammograms to full-field digital mammograms (FFDMs) and assessing CAD performance on a large database.

METHODS

The database included 6478 FFDM images acquired on 1120 females, with 525 cancer cases and 595 negative cases. The database was divided into five case groups: (1) cancer detected during screening, (2) interval cancers, (3) "high-risk" recommended for surgical excision, (4) recalled but negative and (5) negative (not recalled). A previously developed CAD scheme for masses depicted on digitised images was converted and re-optimised for FFDM images while keeping the same image-processing structure. CAD performance was analysed on the entire database.

RESULTS

The case-based sensitivity was 75.6% (397/525) for the current mammograms and 40.8% (42/103) for the prior mammograms deemed negative during clinical interpretation but "visible" during retrospective review. The region-based sensitivity was 58.1% (618/1064) for the current mammograms and 28.4% (57/201) for the prior mammograms. The CAD scheme marked 55.7% (221/397) and 35.7% (15/42) of the masses on both views of the current and the prior examinations, respectively. The overall CAD-cued false-positive rate was 0.32 per image, ranging from 0.29 to 0.51 for the five case groups.

CONCLUSION

This study indicated that (1) digitised image-based CAD can be converted for FFDMs while performing at a comparable, or better, level; (2) CAD detects a substantial fraction of cancers depicted on prior examinations, albeit most having been marked only on one view; and (3) CAD tends to mark more false-positive results on "difficult" negative cases that are more visually difficult for radiologists to interpret.

Exploring the potential of context-sensitive CADe in screening mammography

PURPOSE

Conventional computer-assisted detection (CADe) systems in screening mammography provide the same decision support to all users. The aim of this study was to investigate the potential of a context-sensitive CADe system which provides decision support guided by each user's focus of attention during visual search and reporting patterns for a specific case.

METHODS

An observer study for the detection of malignant masses in screening mammograms was conducted in which six radiologists evaluated 20 mammograms while wearing an eye-tracking device. Eye-position data and diagnostic decisions were collected for each radiologist and case they reviewed. These cases were subsequently analyzed with an in-house knowledge-based CADe system using two different modes: Conventional mode with a globally fixed decision threshold and context-sensitive mode with a location-variable decision threshold based on the radiologists' eye dwelling data and reporting information.

RESULTS

The CADe system operating in conventional mode had 85.7% per-image malignant mass sensitivity at 3.15 false positives per image (FPsI). The same system operating in context-sensitive mode provided personalized decision support at 85.7%-100% sensitivity and 0.35-0.40 FPsI to all six radiologists. Furthermore, context-sensitive CADe system could improve the radiologists' sensitivity and reduce their performance gap more effectively than conventional CADe.

CONCLUSIONS

Context-sensitive CADe support shows promise in delineating and reducing the radiologists' perceptual and cognitive errors in the diagnostic interpretation of screening mammograms more effectively than conventional CADe.

CT colonography computer-aided polyp detection: Effect on radiologist observers of polyp identification by CAD on both the supine and prone scans

RATIONALE AND OBJECTIVES

To determine whether the display of computer-aided detection (CAD) marks on individual polyps on both the supine and prone scans leads to improved polyp detection by radiologists compared to the display of CAD marks on individual polyps on either the supine or the prone scan, but not both.

MATERIALS AND METHODS

The acquisition of patient data for this study was approved by the Institutional Review Board and was Health Insurance Portability and Accountability Act-compliant. Subsequently, the use of the data was declared exempt from further institutional review board review. Four radiologists interpreted 33 computed tomography colonography cases, 21 of which had one adenoma 6-9 mm in size, with the assistance of a CAD system in the first reader mode (ie, the radiologists reviewed only the CAD marks). The radiologists were shown each case twice, with different sets of CAD marks for each of the two readings. In one reading, a true-positive CAD mark for the same polyp was displayed on both the supine and prone scans (a double-mark reading). In the other reading, a true-positive CAD mark was displayed either on the supine or prone scan, but not both (a single-mark reading). True-positive marks were randomized between readings and there was at least a 1-month delay between readings to minimize recall bias. Sensitivity and specificity were determined and receiver operating characteristic (ROC) and multiple-reader multiple-case analyses were performed.

RESULTS

The average per polyp sensitivities were 60% (38%-81%) versus 71% (52%-91%) (P = .03) for single-mark and double-mark readings, respectively. The areas (95% confidence intervals) under the ROC curves were 0.76 (0.62-0.88) and 0.79 (0.58-0.96), respectively (P = NS). Specificities were similar for the single-mark compared with the double-mark readings.

CONCLUSION

The display of CAD marks on a polyp on both the supine and prone scans led to more frequent detection of polyps by radiologists without adversely affecting specificity for detecting 6-9 mm adenomas.

Dynamic multiple thresholding breast boundary detection algorithm for mammograms

PURPOSE

Automated detection of breast boundary is one of the fundamental steps for computer-aided analysis of mammograms. In this study, the authors developed a new dynamic multiple thresholding based breast boundary (MTBB) detection method for digitized mammograms.

METHODS

A large data set of 716 screen-film mammograms (442 CC view and 274 MLO view) obtained from consecutive cases of an Institutional Review Board approved project were used. An experienced breast radiologist manually traced the breast boundary on each digitized image using a graphical interface to provide a reference standard. The initial breast boundary (MTBB-Initial) was obtained by dynamically adapting the threshold to the gray level range in local regions of the breast periphery. The initial breast boundary was then refined by using gradient information from horizontal and vertical Sobel filtering to obtain the final breast boundary (MTBB-Final). The accuracy of the breast boundary detection algorithm was evaluated by comparison with the reference standard using three performance metrics: The Hausdorff distance (HDist), the average minimum Euclidean distance (AMinDist), and the area overlap measure (AOM).

RESULTS

In comparison with the authors' previously developed gradient-based breast boundary (GBB) algorithm, it was found that 68%, 85%, and 94% of images had HDist errors less than 6 pixels (4.8 mm) for GBB, MTBB-Initial, and MTBB-Final, respectively. 89%, 90%, and 96% of images had AMinDist errors less than 1.5 pixels (1.2 mm) for GBB, MTBB-Initial, and MTBB-Final, respectively. 96%, 98%, and 99% of images had AOM values larger than 0.9 for GBB, MTBB-Initial, and MTBB-Final, respectively. The improvement by the MTBB-Final method was statistically significant for all the evaluation measures by the Wilcoxon signed rank test (p < 0.0001).

CONCLUSIONS

The MTBB approach that combined dynamic multiple thresholding and gradient information provided better performance than the breast boundary detection algorithm that mainly used gradient information.

Computer-aided detection of breast masses on mammograms: dual system approach with two-view analysis

PURPOSE

The purpose of this study is to develop a computer-aided detection (CAD) system that combined a dual system approach with a two-view fusion method to improve the accuracy of mass detection on mammograms.

METHODS

The authors previously developed a dual CAD system that merged the decision from two mass detection systems in parallel, one trained with average masses and another trained with subtle masses, to improve sensitivity without excessively increasing false positives (FPs). In this study, they further designed a two-view fusion method to combine the information from different mammographic views. Mass candidates detected independently by the dual system on the two-view mammograms were first identified as potential pairs based on a regional registration technique. A similarity measure was designed to differentiate TP-TP pairs from other pairs (TP-FP and FP-FP pairs) using paired morphological features, Hessian feature, and texture features. A two-view fusion score for each object was generated by weighting the similarity measure with the cross correlation measure of the object pair. Finally, a linear discriminant analysis classifier was trained to combine the mass likelihood score of the object from the single-view dual system and the two-view fusion score for classification of masses and FPs. A total of 2332 mammograms from 735 subjects including 800 normal mammograms from 200 normal subjects was collected with Institutional Review Board (IRB) approval.

RESULTS

When the single-view CAD system that was trained with average masses only were applied to the test sets, the average case-based sensitivities were 50.6% and 63.6% for average masses on current mammograms and 22.6% and 36.2% for subtle masses on prior mammograms at 0.5 and 1 FPs/image, respectively. With the new two-view dual system approach, the average case-based sensitivities were improved to 67.4% and 83.7% for average masses and 44.8% and 57.0% for subtle masses at the same FP rates.

CONCLUSIONS

The improvement with the proposed method was found to be statistically significant (p<0.0001) by JAFROC analysis.

Use of normal tissue context in computer-aided detection of masses in mammograms

When reading mammograms, radiologists do not only look at local properties of suspicious regions but also take into account more general contextual information. This suggests that context may be used to improve the performance of computer-aided detection (CAD) of malignant masses in mammograms. In this study, we developed a set of context features that represent suspiciousness of normal tissue in the same case. For each candidate mass region, three normal reference areas were defined in the image at hand. Corresponding areas were also defined in the contralateral image and in different projections. Evaluation of the context features was done using 10-fold cross validation and case based bootstrapping. Free response receiver operating characteristic (FROC) curves were computed for feature sets including context features and a feature set without context. Results show that the mean sensitivity in the interval of 0.05-0.5 false positives/image increased more than 6% when context features were added. This increase was significant ( p < 0.0001). Context computed using multiple views yielded a better performance than using a single view (mean sensitivity increase of 2.9%, p < 0.0001). Besides the importance of using multiple views, results show that best CAD performance was obtained when multiple context features were combined that are based on different reference areas in the mammogram.

Matching breast masses depicted on different views a comparison of three methods

RATIONALE AND OBJECTIVES

Computerized determination of optimal search areas on mammograms for matching breast mass regions depicted on two ipsilateral views remains a challenge for developing multiview-based computer-aided detection (CAD) schemes. The purpose of this study was to compare three methods aimed at matching CAD-cued mass regions depicted on two views and the associated impact on CAD performance.

MATERIALS AND METHODS

The three search methods used (1) an annular (fan-shaped) band, (2) a straight strip perpendicular to the estimated centerline, and (3) a mixed search area bound on the chest wall side by a straight line and an annular arc on the nipple side, respectively. An image database of 200 examinations with positive results depicting the masses on two views and 200 examinations with negative results was used for testing. Two performance assessment experiments were conducted. The first investigated the maximum matching sensitivity as a function of the search area size, and the second assessed the change in CAD performance using these three search methods.

RESULTS

To include all 200 paired mass regions within the search areas, maximum widths were 28 and 68 mm for the use of the straight strip and the annular band search methods, respectively. When applying a single-image-based CAD scheme to this image database, 172 masses (86% sensitivity) and 523 false-positive (FP) regions (0.33 per image) were detected and cued. Among the positive findings, 92 were cued by the CAD system on both views, and 80 were cued on only one view. In an attempt to match as many of the 172 CAD-cued masses (true-positive [TP] regions) on two views by incrementally reducing the CAD threshold inside the different search areas, the CAD scheme generated 158 TP-TP paired matches with 14 TP-FP paired matches, 142 TP-TP paired matches with 30 TP-FP paired matches, and 146 TP-TP paired matches with 26 TP-FP paired matches, using the methods involving the straight strip, the annular band, and the mixed search areas, respectively. Using the straight strip search method, the CAD also eliminated 25% of FP regions initially cued by the single-image-based CAD scheme and generated the lowest case-based FP detection rate, namely, 15% less than that generated by the annular band method.

CONCLUSIONS

This study showed that among these three search methods, the straight strip method required a smaller search area and achieved the highest level of CAD performance.

Computer-Aided Diagnosis in Mammography Using Content-based Image Retrieval Approaches: Current Status and Future Perspectives

As the rapid advance of digital imaging technologies, the content-based image retrieval (CBIR) has became one of the most vivid research areas in computer vision. In the last several years, developing computer-aided detection and/or diagnosis (CAD) schemes that use CBIR to search for the clinically relevant and visually similar medical images (or regions) depicting suspicious lesions has also been attracting research interest. CBIR-based CAD schemes have potential to provide radiologists with "visual aid" and increase their confidence in accepting CAD-cued results in the decision making. The CAD performance and reliability depends on a number of factors including the optimization of lesion segmentation, feature selection, reference database size, computational efficiency, and relationship between the clinical relevance and visual similarity of the CAD results. By presenting and comparing a number of approaches commonly used in previous studies, this article identifies and discusses the optimal approaches in developing CBIR-based CAD schemes and assessing their performance. Although preliminary studies have suggested that using CBIR-based CAD schemes might improve radiologists' performance and/or increase their confidence in the decision making, this technology is still in the early development stage. Much research work is needed before the CBIR-based CAD schemes can be accepted in the clinical practice.

Automated regional registration and characterization of corresponding microcalcification clusters on temporal pairs of mammograms for interval change analysis

A computerized regional registration and characterization system for analysis of microcalcification clusters on serial mammograms is being developed in our laboratory. The system consists of two stages. In the first stage, based on the location of a detected cluster on the current mammogram, a regional registration procedure identifies the local area on the prior that may contain the corresponding cluster. A search program is used to detect cluster candidates within the local area. The detected cluster on the current image is then paired with the cluster candidates on the prior image to form true (TP-TP) or false (TP-FP) pairs. Automatically extracted features were used in a newly designed correspondence classifier to reduce the number of false pairs. In the second stage, a temporal classifier, based on both current and prior information, is used if a cluster has been detected on the prior image, and a current classifier, based on current information alone, is used if no prior cluster has been detected. The data set used in this study consisted of 261 serial pairs containing biopsy-proven calcification clusters. An MQSA radiologist identified the corresponding clusters on the mammograms. On the priors, the radiologist rated the subtlety of 30 clusters (out of the 261 clusters) as 9 or 10 on a scale of 1 (very obvious) to 10 (very subtle). Leave-one-case-out resampling was used for feature selection and classification in both the correspondence and malignant/benign classification schemes. The search program detected 91.2% (238/261) of the clusters on the priors with an average of 0.42 FPs/image. The correspondence classifier identified 86.6% (226/261) of the TP-TP pairs with 20 false matches (0.08 FPs/image) relative to the entire set of 261 image pairs. In the malignant/benign classification stage the temporal classifier achieved a test A(z) of 0.81 for the 246 pairs which contained a detection on the prior. In addition, a classifier was designed by using the clusters on the current mammograms only. It achieved a test A(z) of 0.72 in classifying the clusters as malignant and benign. The difference between the performance of the temporal classifier and the current classifier was statistically significant (p=0.0014). Our interval change analysis system can detect the corresponding cluster on the prior mammogram with high sensitivity, and classify them with a satisfactory accuracy.

Correlative feature analysis on FFDM

Identifying the corresponding images of a lesion in different views is an essential step in improving the diagnostic ability of both radiologists and computer-aided diagnosis (CAD) systems. Because of the nonrigidity of the breasts and the 2D projective property of mammograms, this task is not trivial. In this pilot study, we present a computerized framework that differentiates between corresponding images of the same lesion in different views and noncorresponding images, i.e., images of different lesions. A dual-stage segmentation method, which employs an initial radial gradient index (RGI) based segmentation and an active contour model, is applied to extract mass lesions from the surrounding parenchyma. Then various lesion features are automatically extracted from each of the two views of each lesion to quantify the characteristics of density, size, texture and the neighborhood of the lesion, as well as its distance to the nipple. A two-step scheme is employed to estimate the probability that the two lesion images from different mammographic views are of the same physical lesion. In the first step, a correspondence metric for each pairwise feature is estimated by a Bayesian artificial neural network (BANN). Then, these pairwise correspondence metrics are combined using another BANN to yield an overall probability of correspondence. Receiver operating characteristic (ROC) analysis was used to evaluate the performance of the individual features and the selected feature subset in the task of distinguishing corresponding pairs from noncorresponding pairs. Using a FFDM database with 123 corresponding image pairs and 82 noncorresponding pairs, the distance feature yielded an area under the ROC curve (AUC) of 0.81 +/- 0.02 with leave-one-out (by physical lesion) evaluation, and the feature metric subset, which included distance, gradient texture, and ROI-based correlation, yielded an AUC of 0.87 +/- 0.02. The improvement by using multiple feature metrics was statistically significant compared to single feature performance.

Anniversary paper: History and status of CAD and quantitative image analysis: the role of Medical Physics and AAPM

The roles of physicists in medical imaging have expanded over the years, from the study of imaging systems (sources and detectors) and dose to the assessment of image quality and perception, the development of image processing techniques, and the development of image analysis methods to assist in detection and diagnosis. The latter is a natural extension of medical physicists' goals in developing imaging techniques to help physicians acquire diagnostic information and improve clinical decisions. Studies indicate that radiologists do not detect all abnormalities on images that are visible on retrospective review, and they do not always correctly characterize abnormalities that are found. Since the 1950s, the potential use of computers had been considered for analysis of radiographic abnormalities. In the mid-1980s, however, medical physicists and radiologists began major research efforts for computer-aided detection or computer-aided diagnosis (CAD), that is, using the computer output as an aid to radiologists-as opposed to a completely automatic computer interpretation-focusing initially on methods for the detection of lesions on chest radiographs and mammograms. Since then, extensive investigations of computerized image analysis for detection or diagnosis of abnormalities in a variety of 2D and 3D medical images have been conducted. The growth of CAD over the past 20 years has been tremendous-from the early days of time-consuming film digitization and CPU-intensive computations on a limited number of cases to its current status in which developed CAD approaches are evaluated rigorously on large clinically relevant databases. CAD research by medical physicists includes many aspects-collecting relevant normal and pathological cases; developing computer algorithms appropriate for the medical interpretation task including those for segmentation, feature extraction, and classifier design; developing methodology for assessing CAD performance; validating the algorithms using appropriate cases to measure performance and robustness; conducting observer studies with which to evaluate radiologists in the diagnostic task without and with the use of the computer aid; and ultimately assessing performance with a clinical trial. Medical physicists also have an important role in quantitative imaging, by validating the quantitative integrity of scanners and developing imaging techniques, and image analysis tools that extract quantitative data in a more accurate and automated fashion. As imaging systems become more complex and the need for better quantitative information from images grows, the future includes the combined research efforts from physicists working in CAD with those working on quantitative imaging systems to readily yield information on morphology, function, molecular structure, and more-from animal imaging research to clinical patient care. A historical review of CAD and a discussion of challenges for the future are presented here, along with the extension to quantitative image analysis.

KDD cup 2008 and the workshop on mining medical data

In this report we summarize the KDD Cup 2008 task, which addressed a problem of early breast cancer detection. We describe the data and the challenges, the results and summarize the algorithms used by the winning teams. We also summarize the workshop on Mining Medical Data held in conjunction with SIGKDD on August 24, 2008 in Las Vegas, NV that brought together researchers working on various aspects of applying machine learning and data mining to challenging tasks in medical and health care domains.

An ellipse-fitting based method for efficient registration of breast masses on two mammographic views

When reading mammograms, radiologists routinely search for and compare suspicious breast lesions identified on two corresponding craniocaudal (CC) and mediolateral oblique (MLO) views. Automatically identifying and matching the same true-positive breast lesions depicted on two views is an important step for developing successful multiview based computer-aided detection (CAD) schemes. The authors developed a method to automatically register breast areas and detect matching strips of interest used to identify the matched mass regions depicted on CC and MLO views. The method uses an ellipse based model to fit the breast boundary contour (skin line) and set a local Cartesian coordinate system for each view. One intersection point between the major/minor axis and the fitted ellipse perimeter passed through breast boundary is selected as the origin and the majoraxis and the minoraxis of the ellipse are used as the two axis of the Cartesian coordinate system. When a mass is identified on one view, the scheme computes its position in the local coordinate system. Then, the distance is mapped onto the local coordinate of the other view. At the end of the mapped distance a registered centerline of the matching strip is created. The authors established an image database that includes 200 test examinations each depicting one verified mass visible on the two views. They tested whether the registered centerline identified on another view can be used to locate the matched mass region. The experiments show that the average distance between the mass region centers and the registered centerlines was +/- 8.3 mm and in 91% of testing cases the registered centerline actually passes through the matched mass regions. A matching strip width of 47 mm was required to achieve 100% sensitivity for the test database. The results demonstrate the feasibility of the proposed method to automatically identify masses depicted on CC and MLO views, which may improve future development of multiview based CAD schemes.

Evaluating computer-aided detection algorithms

Computer-aided detection (CAD) has been attracting extensive research interest during the last two decades. It is recognized that the full potential of CAD can only be realized by improving the performance and robustness of CAD algorithms and this requires good evaluation methodology that would permit CAD designers to optimize their algorithms. Free-response receiver operating characteristic (FROC) curves are widely used to assess CAD performance, however, evaluation rarely proceeds beyond determination of lesion localization fraction (sensitivity) at an arbitrarily selected value of nonlesion localizations (false marks) per image. This work describes a FROC curve fitting procedure that uses a recent model of visual search that serves as a framework for the free-response task. A maximum likelihood procedure for estimating the parameters of the model from free-response data and fitting CAD generated FROC curves was implemented. Procedures were implemented to estimate two figures of merit and associated statistics such as 95% confidence intervals and goodness of fit. One of the figures of merit does not require the arbitrary specification of an operating point at which to evaluate CAD performance. For comparison a related method termed initial detection and candidate analysis was also implemented that is applicable when all suspicious regions are reported. The two methods were tested on seven mammography CAD data sets and both yielded good to excellent fits. The search model approach has the advantage that it can potentially be applied to radiologist generated free-response data where not all suspicious regions are reported, only the ones that are deemed sufficiently suspicious to warrant clinical follow-up. This work represents the first practical application of the search model to an important evaluation problem in diagnostic radiology. Software based on this work is expected to benefit CAD developers working in diverse areas of medical imaging.

Information-theoretic CAD system in mammography: Entropy-based indexing for computational efficiency and robust performance

We have previously presented a knowledge-based computer-assisted detection (KB-CADe) system for the detection of mammographic masses. The system is designed to compare a query mammographic region with mammographic templates of known ground truth. The templates are stored in an adaptive knowledge database. Image similarity is assessed with information theoretic measures (e.g., mutual information) derived directly from the image histograms. A previous study suggested that the diagnostic performance of the system steadily improves as the knowledge database is initially enriched with more templates. However, as the database increases in size, an exhaustive comparison of the query case with each stored template becomes computationally burdensome. Furthermore, blind storing of new templates may result in redundancies that do not necessarily improve diagnostic performance. To address these concerns we investigated an entropy-based indexing scheme for improving the speed of analysis and for satisfying database storage restrictions without compromising the overall diagnostic performance of our KB-CADe system. The indexing scheme was evaluated on two different datasets as (i) a search mechanism to sort through the knowledge database, and (ii) a selection mechanism to build a smaller, concise knowledge database that is easier to maintain but still effective. There were two important findings in the study. First, entropy-based indexing is an effective strategy to identify fast a subset of templates that are most relevant to a given query. Only this subset could be analyzed in more detail using mutual information for optimized decision making regarding the query. Second, a selective entropy-based deposit strategy may be preferable where only high entropy cases are maintained in the knowledge database. Overall, the proposed entropy-based indexing scheme was shown to reduce the computational cost of our KB-CADe system by 55% to 80% while maintaining the system's diagnostic performance.

A concentric morphology model for the detection of masses in mammography

We propose a technique for the automated detection of malignant masses in screening mammography. The technique is based on the presence of concentric layers surrounding a focal area with suspicious morphological characteristics and low relative incidence in the breast region. Mammographic locations with high concentration of concentric layers with progressively lower average intensity are considered suspicious deviations from normal parenchyma. The multiple concentric layers (MCLs) technique was trained and tested using the craniocaudal views of 270 mammographic cases with biopsy proven malignant masses from the digital database of screening mammography. One-half of the available cases were used for optimizing the parameters of the detection algorithm. The remaining cases were used for testing. During testing, malignant masses were detected with 92%, 88%, and 81% sensitivity at 5.4, 2.4, and 0.6 false positive marks per image. Testing on 82 normal screening mammograms showed a false positive rate of 5.0, 1.7, and 0.2 marks per image at the previously reported operating points. Furthermore, additional evaluation on 135 benign cases produced a significantly lower detection rate for benign masses (61.6%, 58.3%, and 43.7% at 5.1, 2.8, and 1.2 false positives per image, respectively). Overall, MCL is a promising computer-assisted detection strategy for screening mammograms to identify malignant masses while maintaining the detection rate of benign masses considerably lower.

Spatial localization accuracy of radiologists in free-response studies: Inferring perceptual FROC curves from mark-rating data

RATIONALE AND OBJECTIVES

Free-response data consist of a set of mark-ratings pairs. Before analysis, the data are classified or "scored" into lesion and non-lesion localizations. The scoring is done by choosing an acceptance-radius and classifying marks within the acceptance-radius of lesion centers as lesion localizations, and all other marks are classified as non-lesion localizations. The scored data are plotted as a free-response receiver operating characteristic (FROC) curve, essentially a plot of appropriately normalized numbers of lesion localizations vs. non-lesion localizations. Scored FROC curves are frequently used to compare imaging systems and computer-aided detection (CAD) algorithms. However, the choice of acceptance-radius is arbitrary. This makes it difficult to compare curves from different studies and to estimate true performance.

MATERIALS AND METHODS

To resolve this issue the concept of two types of marks is introduced: perceptual hits and perceptual misses. A perceptual hit is a mark made in response to the observer seeing the lesion. A perceptual miss is a mark made in response to the observer seeing a (lesion-like) non-lesion. A method of estimating the most probable numbers of perceptual hits and misses is described. This allows one to plot a perceptual FROC operating point and by extension a perceptual FROC curve. Unlike a scored FROC operating point, a perceptual point is independent of the choice of acceptance-radius. The method does not allow one to identify individual marks as perceptual hits or misses-only the most probable numbers. It is based on a three-parameter statistical model of the spatial distributions of perceptual hits and misses relative to lesion centers.

RESULTS

The method has been applied to an observer dataset in which mammographers and residents with different levels of experience were asked to locate lesions in mammograms. The perceptual operating points suggest superior performance for the mammographers and equivalent performance for residents in the first and second mammography rotations. These results and the model validation are preliminary as they are based on a small dataset.

CONCLUSION

The significance of this study is showing that it is possible to probabilistically determine if a mark resulted from seeing a lesion or a non-lesion. Using the method developed in this study one could perform acceptance-radius independent estimation of observer performance.