Improving performance of computer-aided detection scheme by combining results from two machine learning classifiers

[Comparing prediction performances of 18F-FDG PET and CGFL/Curie nomogram to predict pathologic complete response after neoadjuvant chemotherapy for HER2-positive breast cancers]

OBJECTIVES

The aim of this study was to compare the value of ¹⁸F-fluorodesoxyglucose positron emission tomography (¹⁸F-FDG PET/CT) with CGFL/Curie nomogram to predict a pathologic complete response (pCR) after neoadjuvant chemotherapy (NAC) in women with human epidermal growth factor 2 (HER2)-positive breast cancer treated by trastuzumab.

METHODS

Fifty-one women with HER2-positive breast cancer treated with trastuzumab plus taxane-based NAC were retrospectively included from January 2005 to December 2015. For ¹⁸F-FDG PET/CT, the analyzed predictor was the maximum standardized uptake value of the primary tumor and axillary nodes after the first course of NAC (PET₂.SUV_max). pCR was defined by no residual infiltrative tumor but in situ tumor was accepted. Accuracy of CGFL/Curie nomogram and PET₂.SUV_max was evaluated measuring sensitivity (Se), specificity (Sp), positive predictive value (PPV) and negative predictive value (NPV). Combined prediction was evaluated testing predictor's associations.

RESULTS

For CGFL/Curie nomogram's performances, Se, Sp, PPV and NPV were respectively: 76% (95%CI: 58-90%), 57% (95%CI: 43-66%), 55% (95%CI: 42-65), 77% (95%CI: 59-90%). For PET₂.SUV_max's performances, Se, Sp, PPV and NPV were respectively: 67% (95%CI: 48-81%), 77% (95%CI: 64-97%), 67% (95%CI: 48-82%), 77% (95%CI: 64-87%). ROC curves for these predictors were similar; the areas under the curve were 0.6 (95%CI: 0.56-0.64) for PET₂.SUV_max and 0.55 (95%CI: 0.50-0.59) for CGFL/Curie nomogram. Combined prediction was efficient with Se at 80%, VPN at 76%, Sp at 78% and VPP at 81%.

CONCLUSIONS

CGFL/Curie nomogram and PET₂.SUV_max were two efficient predictors of pCR in patients with HER2-positive breast cancer. Combined prediction has an improved accuracy.

A Comparative Study of Land Cover Classification by Using Multispectral and Texture Data

The main objective of this study is to find out the importance of machine vision approach for the classification of five types of land cover data such as bare land, desert rangeland, green pasture, fertile cultivated land, and Sutlej river land. A novel spectra-statistical framework is designed to classify the subjective land cover data types accurately. Multispectral data of these land covers were acquired by using a handheld device named multispectral radiometer in the form of five spectral bands (blue, green, red, near infrared, and shortwave infrared) while texture data were acquired with a digital camera by the transformation of acquired images into 229 texture features for each image. The most discriminant 30 features of each image were obtained by integrating the three statistical features selection techniques such as Fisher, Probability of Error plus Average Correlation, and Mutual Information (F + PA + MI). Selected texture data clustering was verified by nonlinear discriminant analysis while linear discriminant analysis approach was applied for multispectral data. For classification, the texture and multispectral data were deployed to artificial neural network (ANN: n-class). By implementing a cross validation method (80-20), we received an accuracy of 91.332% for texture data and 96.40% for multispectral data, respectively.

An interactive system for computer-aided diagnosis of breast masses

Although mammography is the only clinically accepted imaging modality for screening the general population to detect breast cancer, interpreting mammograms is difficult with lower sensitivity and specificity. To provide radiologists "a visual aid" in interpreting mammograms, we developed and tested an interactive system for computer-aided detection and diagnosis (CAD) of mass-like cancers. Using this system, an observer can view CAD-cued mass regions depicted on one image and then query any suspicious regions (either cued or not cued by CAD). CAD scheme automatically segments the suspicious region or accepts manually defined region and computes a set of image features. Using content-based image retrieval (CBIR) algorithm, CAD searches for a set of reference images depicting "abnormalities" similar to the queried region. Based on image retrieval results and a decision algorithm, a classification score is assigned to the queried region. In this study, a reference database with 1,800 malignant mass regions and 1,800 benign and CAD-generated false-positive regions was used. A modified CBIR algorithm with a new function of stretching the attributes in the multi-dimensional space and decision scheme was optimized using a genetic algorithm. Using a leave-one-out testing method to classify suspicious mass regions, we compared the classification performance using two CBIR algorithms with either equally weighted or optimally stretched attributes. Using the modified CBIR algorithm, the area under receiver operating characteristic curve was significantly increased from 0.865 ± 0.006 to 0.897 ± 0.005 (p < 0.001). This study demonstrated the feasibility of developing an interactive CAD system with a large reference database and achieving improved performance.

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.

Assessment of performance and reliability of computer-aided detection scheme using content-based image retrieval approach and limited reference database

Content-based image retrieval approach was used in our computer-aided detection (CAD) schemes for breast cancer detection with mammography. In this study, we assessed CAD performance and reliability using a reference database including 1500 positive (breast mass) regions of interest (ROIs) and 1500 normal ROIs. To test the relationship between CAD performance and the similarity level between the queried ROI and the retrieved ROIs, we applied a set of similarity thresholds to the retrieved similar ROIs selected by the CAD schemes for all queried suspicious regions, and used only the ROIs that were above the threshold for assessing CAD performance at each threshold level. Using the leave-one-out testing method, we computed areas under receiver operating characteristic (ROC) curves (A(Z)) to assess CAD performance. The experimental results showed that as threshold increase, (1) less true positive ROIs can be referenced in the database than normal ROIs and (2) the A(Z) value was monotonically increased from 0.854 ± 0.004 to 0.932 ± 0.016. This study suggests that (1) in order to more accurately detect and diagnose subtle masses, a large and diverse database is required, and (2) assessing the reliability of the decision scores based on the similarity measurement is important in application of the CBIR-based CAD schemes when the limited database is used.

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.

Computer-aided detection of early interstitial lung diseases using low-dose CT images

This study aims to develop a new computer-aided detection (CAD) scheme to detect early interstitial lung disease (ILD) using low-dose computed tomography (CT) examinations. The CAD scheme classifies each pixel depicted on the segmented lung areas into positive or negative groups for ILD using a mesh-grid-based region growth method and a multi-feature-based artificial neural network (ANN). A genetic algorithm was applied to select optimal image features and the ANN structure. In testing each CT examination, only pixels selected by the mesh-grid region growth method were analyzed and classified by the ANN to improve computational efficiency. All unselected pixels were classified as negative for ILD. After classifying all pixels into the positive and negative groups, CAD computed a detection score based on the ratio of the number of positive pixels to all pixels in the segmented lung areas, which indicates the likelihood of the test case being positive for ILD. When applying to an independent testing dataset of 15 positive and 15 negative cases, the CAD scheme yielded the area under receiver operating characteristic curve (AUC = 0.884 ± 0.064) and 80.0% sensitivity at 85.7% specificity. The results demonstrated the feasibility of applying the CAD scheme to automatically detect early ILD using low-dose CT examinations.

Computer-aided detection; the effect of training databases on detection of subtle breast masses

RATIONALE AND OBJECTIVES

Lesion conspicuity is typically highly correlated with visual difficulty for lesion detection, and computer-aided detection (CAD) has been widely used as a "second reader" in mammography. Hence, increasing CAD sensitivity in detecting subtle cancers without increasing false-positive rates is important. The aim of this study was to investigate the effect of training database case selection on CAD performance in detecting low-conspicuity breast masses.

MATERIALS AND METHODS

A full-field digital mammographic image database that included 525 cases depicting malignant masses was randomly partitioned into three subsets. A CAD scheme was applied to detect all initially suspected mass regions and compute region conspicuity. Training samples were iteratively selected from two of the subsets. Four types of training data sets-(1) one including all available true-positive mass regions in the two subsets ("all"), (2) one including 350 randomly selected mass regions ("diverse"), (3) one including 350 high-conspicuity mass regions ("easy"), and (4) one including 350 low-conspicuity mass regions ("difficult")-were assembled. In each training data set, the same number of randomly selected false-positive regions as the true-positives were also included. Two classifiers, an artificial neural network (ANN) and a k-nearest neighbor (KNN) algorithm, were trained using each of the four training data sets and tested on all suspected regions in the remaining data set. Using a threefold cross-validation method, the performance changes of the CAD schemes trained using one of the four training data sets were computed and compared.

RESULTS

CAD initially detected 1025 true-positive mass regions depicted on 507 cases (97% case-based sensitivity) and 9569 false-positive regions (3.5 per image) in the entire database. Using the all training data set, CAD achieved the highest overall performance on the entire testing database. However, CAD detected the highest number of low-conspicuity masses when the difficult training data set was used for training. Results did agree for both ANN-based and KNN-based classifiers in all tests. Compared to the use of the all training data set, the sensitivity of the schemes trained using the difficult data set decreased by 8.6% and 8.4% for the ANN and KNN algorithm on the entire database, respectively, but the detection of low-conspicuity masses increased by 7.1% and 15.1% for the ANN and KNN algorithm at a false-positive rate of 0.3 per image.

CONCLUSIONS

CAD performance depends on the size, diversity, and difficulty level of the training database. To increase CAD sensitivity in detecting subtle cancer, one should increase the fraction of difficult cases in the training database rather than simply increasing the training data set size.

Hybrid segmentation of mass in mammograms using template matching and dynamic programming

RATIONALE AND OBJECTIVES

Accurate image segmentation for breast lesions is a critical step in computer-aided diagnosis systems. The objective of this study was to develop a robust method for the automatic segmentation of breast masses on mammograms to extract feasible features for computer-aided diagnosis systems.

MATERIALS AND METHODS

The data set used in this study consisted of 483 regions of interest extracted from 328 patients. A hybrid method for segmenting breast masses was proposed on the basis of the template-matching and dynamic programming techniques. First, a template-matching technique was used to locate and obtain the rough region of masses. Then, on the basis of this rough region, a local cost function for dynamic programming was defined. Finally, the optimal contour was derived by applying dynamic programming as an optimization technique. The performance of this proposed segmentation method was evaluated using area-based and boundary distance-based similarity measures based on radiologists' manually marked annotations. A comparison with three different segmentation algorithms on the data set was provided.

RESULTS

The mean overlap percentage for our proposed hybrid method was 0.727 ± 0.127, whereas those for Timp and Karssemeijer's dynamic programming method, Song et al's plane-fitting and dynamic programming method, and the normalized cut segmentation method were 0.657 ± 0.216, 0.636 ± 0.190, and 0.562 ± 0.199, respectively. All P values for the measure distribution of our proposed method and the other three algorithms were <.001.

CONCLUSIONS

A hybrid method based on the template-matching and dynamic programming techniques was proposed to segment breast masses on mammograms. Evaluation results indicate that the proposed segmentation method can improve the accuracy of mass segmentation compared to three other algorithms. The proposed segmentation method shows better performance and has great potential in improving the accuracy of computer-aided diagnosis systems in interpreting mammograms.

A multistage approach to improve performance of computer-aided detection of pulmonary embolisms depicted on CT images: preliminary investigation

This study developed a computer-aided detection (CAD) scheme for pulmonary embolism (PE) detection and investigated several approaches to improve CAD performance. In the study, 20 computed tomography examinations with various lung diseases were selected, which include 44 verified PE lesions. The proposed CAD scheme consists of five basic steps: 1) lung segmentation; 2) PE candidate extraction using an intensity mask and tobogganing region growing; 3) PE candidate feature extraction; 4) false-positive (FP) reduction using an artificial neural network (ANN); and 5) a multifeature-based k-nearest neighbor for positive/negative classification. In this study, we also investigated the following additional methods to improve CAD performance: 1) grouping 2-D detected features into a single 3-D object; 2) selecting features with a genetic algorithm (GA); and 3) limiting the number of allowed suspicious lesions to be cued in one examination. The results showed that 1) CAD scheme using tobogganing, an ANN, and grouping method achieved the maximum detection sensitivity of 79.2%; 2) the maximum scoring method achieved the superior performance over other scoring fusion methods; 3) GA was able to delete "redundant" features and further improve CAD performance; and 4) limiting the maximum number of cued lesions in an examination reduced FP rate by 5.3 times. Combining these approaches, CAD scheme achieved 63.2% detection sensitivity with 18.4 FP lesions per examination. The study suggested that performance of CAD schemes for PE detection depends on many factors that include 1) optimizing the 2-D region grouping and scoring methods; 2) selecting the optimal feature set; and 3) limiting the number of allowed cueing lesions per examination.

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.