Jeffrey W. Hoffmeister

Computer-aided breast cancer detection and diagnosis of masses using difference of Gaussians and derivative-based feature saliency

A new model-based vision (MBV) algorithm is developed to find regions of interest (ROIs) corresponding to masses in digitized mammograms and to classify the masses as malignant/benign. The MBV algorithm is comprised of 5 modules to structurally identify suspicious ROIs, eliminate false positives, and classify the remaining as malignant or benign. The focus of attention module uses a difference of Gaussians (DoG) filter to highlight suspicious regions in the mammogram. The index module uses tests to reduce the number of nonmalignant regions from 8.39 to 2.36 per full breast image. Size, shape, contrast, and Laws texture features are used to develop the prediction modules mass models. Derivative-based feature saliency techniques are used to determine the best features for classification. Nine features are chosen to define the malignant/benign models. The feature extraction module obtains these features from all suspicious ROIs. The matching module classifies the regions using a multilayer perceptron neural network architecture to obtain an overall classification accuracy of 100% for the segmented malignant masses with a false-positive rate of 1.8 per full breast image. This system has a sensitivity of 92% for locating malignant ROIs. The database contains 272 images (12 b, 100 /spl mu/m) with 36 malignant and 53 benign mass images. The results demonstrate that the MBV approach provides a structured order of integrating complex stages into a system for radiologists.

Detection of Breast Cancer with Full-Field Digital Mammography and Computer-Aided Detection

OBJECTIVE The purpose of this study was to evaluate computer-aided detection (CAD) performance with full-field digital mammography (FFDM). MATERIALS AND METHODS CAD (Second Look, version 7.2) was used to evaluate 123 cases of breast cancer detected with FFDM (Senographe DS). Retrospectively, CAD sensitivity was assessed using breast density, mammographic presentation, histopathology results, and lesion size. To determine the case-based false-positive rate, patients with four standard views per case were included in the study group. Eighteen unilateral mammography examinations with nonstandard views were excluded, resulting in a sample of 105 bilateral cases. RESULTS CAD detected 115 (94%) of 123 cancer cases: six of six (100%) in fatty breasts, 63 of 66 (95%) in breasts containing scattered fibroglandular densities, 43 of 46 (93%) in heterogeneously dense breasts, and three of five (60%) in extremely dense breasts. CAD detected 93% (41/44) of cancers manifesting as calcifications, 92% (57/62) as masses, and 100% (17/17) as mixed masses and calcifications. CAD detected 94% of the invasive ductal carcinomas (n = 63), 100% of the invasive lobular carcinomas (n = 7), 91% of the other invasive carcinomas (n = 11), and 93% of the ductal carcinomas in situ (n = 42). CAD sensitivity for cancers 1-10 mm (n = 55) was 89%; 11-20 mm (n = 37), 97%; 21-30 mm (n = 16), 100%; and larger than 30 mm (n = 15), 93%. The CAD false-positive rate was 2.3 marks per four-image case. CONCLUSION CAD with FFDM showed a high sensitivity in identifying cancers manifesting as calcifications and masses. Sensitivity was maintained in cancers with lower mammographic sensitivity, including invasive lobular carcinomas and small neoplasms (1-20 mm). CAD with FFDM should be effective in assisting radiologists with earlier detection of breast cancer. Future studies are needed to assess CAD accuracy in larger populations.

Computer-aided detection system applied to full-field digital mammograms.

Background: Although mammography remains the mainstay for breast cancer screening, it is an imperfect examination with a sensitivity of 75–92% for breast cancer. Computer-aided detection (CAD) has been developed to improve mammographic detection of breast cancer. Purpose: To retrospectively estimate CAD sensitivity and false-positive rate with full-field digital mammograms (FFDMs). Material and Methods: CAD was used to evaluate 151 cases of ductal carcinoma in situ (DCIS) (n=48) and invasive breast cancer (n=103) detected with FFDM. Retrospectively, CAD sensitivity was estimated based on breast density, mammographic presentation, histopathology type, and lesion size. CAD false-positive rate was estimated with screening FFDMs from 200 women. Results: CAD detected 93% (141/151) of cancer cases: 97% (28/29) in fatty breasts, 94% (81/86) in breasts containing scattered fibroglandular densities, 90% (28/31) in heterogeneously dense breasts, and 80% (4/5) in extremely dense breasts. CAD detected 98% (54/55) of cancers manifesting as calcifications, 89% (74/83) as masses, and 100% (13/13) as mixed masses and calcifications. CAD detected 92% (73/79) of invasive ductal carcinomas, 89% (8/9) of invasive lobular carcinomas, 93% (14/15) of other invasive carcinomas, and 96% (46/48) of DCIS. CAD sensitivity for cancers 1–10 mm was 87% (47/54); 11–20 mm, 99% (70/71); 21–30 mm, 86% (12/14); and larger than 30 mm, 100% (12/12). The CAD false-positive rate was 2.5 marks per case. Conclusion: CAD with FFDM showed a high sensitivity in identifying cancers manifesting as calcifications or masses. CAD sensitivity was maintained in small lesions (1–20 mm) and invasive lobular carcinomas, which have lower mammographic sensitivity.

Automated global matching of temporal thoracic helical CT studies: feasibility study

Abstract In this paper, we demonstrate the feasibility of a computer-aided diagnosis (CAD) method that automatically matches slices from temporal thoracic helical computed tomography (CT) studies. The proposed method uses three-dimensional (3D) anatomical information but does not require anatomical landmark identification or organ segmentation. Information from coronal maximum intensity projection (MIP) images is utilized in the registration process. The registration method was assessed using 10 pairs of sequential lung CT studies.

Detection of soft tissue densities from digital breast tomosynthesis: comparison of conventional and deep learning approaches

Computer-aided detection (CAD) has been used in screening mammography for many years and is likely to be utilized for digital breast tomosynthesis (DBT). Higher detection performance is desirable as it may have an impact on radiologists decisions and clinical outcomes. Recently the algorithms based on deep convolutional architectures have been shown to achieve state of the art performance in object classification and detection. Similarly, we trained a deep convolutional neural network directly on patches sampled from two-dimensional mammography and reconstructed DBT volumes and compared its performance to a conventional CAD algorithm that is based on computation and classification of hand-engineered features. The detection performance was evaluated on the independent test set of 344 DBT reconstructions (GE SenoClaire 3D, iterative reconstruction algorithm) containing 328 suspicious and 115 malignant soft tissue densities including masses and architectural distortions. Detection sensitivity was measured on a region of interest (ROI) basis at the rate of five detection marks per volume. Moving from conventional to deep learning approach resulted in increase of ROI sensitivity from 0:832 ± 0:040 to 0:893 ± 0:033 for suspicious ROIs; and from 0:852 ± 0:065 to 0:930 ± 0:046 for malignant ROIs. These results indicate the high utility of deep feature learning in the analysis of DBT data and high potential of the method for broader medical image analysis tasks.

Neural network feature selection for breast cancer diagnosis

More than 50 million women over the age of 40 are currently at risk for breast cancer in the United States. Computer-aided diagnosis, as a second opinion to radiologists, will aid in decreasing the number of false readings of mammograms. Neural network benefits are exploited at both the classification and feature selection stages in the development of a computer-aided breast cancer diagnostic system. The multilayer perceptron is used to classify and contrast three features (angular second moment, eigenmasses, and wavelets) developed to distinguish benign from malignant lesion in a database of 94 difficult-to-diagnose digitized microcalcification cases. System performance of 74 percent correct classifications is achieved. Feature selection techniques are presented which further improve performance. Neural and decision boundary-based methods are implemented, compared, and validated to isolate and remove useless features. The contribution from this analysis is an increase to 88 percent correct classification in system performance. These feature selection techniques can also process risk factor data.

Workflow improvements for digital breast tomosynthesis: computerized generation of enhanced synthetic images

In a typical 2D mammography workflow scenario, a computer-aided detection (CAD) algorithm is used as a second reader producing marks for a radiologist to review. In the case of 3D digital breast tomosynthesis (DBT), the display of CAD detections at multiple reconstruction heights would lead to an increased image browsing and interpretation time. We propose an alternative approach in which an algorithm automatically identifies suspicious regions of interest from 3D reconstructed DBT slices and then merges the findings with the corresponding 2D synthetic projection image which is then reviewed. The resultant enhanced synthetic 2D image combines the benefits of a familiar 2D breast view with superior appearance of suspicious locations from 3D slices. Moreover, clicking on 2D suspicious locations brings up the display of the corresponding 3D regions in a DBT volume allowing navigation between 2D and 3D images. We explored the use of these enhanced synthetic images in a concurrent read paradigm by conducting a study with 5 readers and 30 breast exams. We observed that the introduction of the enhanced synthetic view reduced radiologists average interpretation time by 5.4%, increased sensitivity by 6.7% and increased specificity by 15.6%.

Measurement of colonic polyp size from virtual colonoscopy studies: Comparison of manual and automated methods

Polyp size is an important feature descriptor for clinical classification and follow-up decision making in CT colonography. Currently, polyp size is measured from computed tomography (CT) studies manually as the single largest dimension of the polyp head, excluding the stalk if present, in either multi-planar reconstruction (MPR) or three-dimensional (3D) views. Manual measurements are subject to intra- and inter-reader variation, and can be time-consuming. Automated polyp segmentation and size measurement can reduce the variability and speed up the process. In this study, an automated polyp size measurement technique is developed. Using this technique, the polyp is segmented from the attached healthy tissue using a novel, model-based approach. The largest diameter of the segmented polyp is measured in axial, sagitttal and coronal MPR views. An expert radiologist identified 48 polyps from either supine or prone views of 52 cases of the Walter-Reed virtual colonoscopy database. Automated polyp size measurements were carried out and compared with the manual ones. For comparison, three different statistical methods were used: overall agreement using chance-corrected kappa indices; the mean absolute differences; and Bland-Altman limits of agreement. Manual and automated measurements show good agreement both in 2D and 3D views.

Derivative-based feature saliency for computer-aided breast cancer detection and diagnosis

Derivative-based feature saliency techniques were used to define the best of 25 Laws texture features for the classification of 101 malignant mass and benign mass regions. Statistical and derivative-based saliency techniques were used to select the best size, shape, contrast, and Laws texture features for the mass model. Nine features were chosen to define the model, of which four have been used by other researchers. Using this model, the regions were classified using a multilayer perceptron neural network architecture trained with an imbalanced training set weight update algorithm to obtain an overall classification accuracy of 100 percent for the segmented malignant masses with a false-positive rates of 1.8/image. The system has shown a sensitivity of 92 percent for locating malignant ROIs. The database contained 284 images (12 bit, 100 micrometers ).