Computer-aided Detection Performance Research Articles

Computerized breast cancer analysis system using three stage semi-supervised learning method

Background and ObjectiveA large number of labeled medical image data is usually a requirement to train a well-performed computer-aided detection (CAD) system. But the process of data labeling is time consuming, and potential ethical and logistical problems may also present complications. As a result, incorporating unlabeled data into CAD system can be a feasible way to combat these obstacles. MethodsIn this study we developed a three stage semi-supervised learning (SSL) scheme that combines a small amount of labeled data and larger amount of unlabeled data. The scheme was modified on our existing CAD system using the following three stages: data weighing, feature selection, and newly proposed dividing co-training data labeling algorithm. Global density asymmetry features were incorporated to the feature pool to reduce the false positive rate. Area under the curve (AUC) and accuracy were computed using 10 fold cross validation method to evaluate the performance of our CAD system. The image dataset includes mammograms from 400 women who underwent routine screening examinations, and each pair contains either two cranio-caudal (CC) or two mediolateral-oblique (MLO) view mammograms from the right and the left breasts. From these mammograms 512 regions were extracted and used in this study, and among them 90 regions were treated as labeled while the rest were treated as unlabeled. ResultsUsing our proposed scheme, the highest AUC observed in our research was 0.841, which included the 90 labeled data and all the unlabeled data. It was 7.4% higher than using labeled data only. With the increasing amount of labeled data, AUC difference between using mixed data and using labeled data only reached its peak when the amount of labeled data was around 60. ConclusionsThis study demonstrated that our proposed three stage semi-supervised learning can improve the CAD performance by incorporating unlabeled data. Using unlabeled data is promising in computerized cancer research and may have a significant impact for future CAD system applications.

Technical Note: Image filtering to make computer-aided detection robust to image reconstruction kernel choice in lung cancer CT screening.

In lung cancer computed tomography (CT) screening, the performance of a computer-aided detection (CAD) system depends on the selection of the image reconstruction kernel. To reduce this dependence on reconstruction kernels, the authors propose a novel application of an image filtering method previously proposed by their group. The proposed filtering process uses the ratio of modulation transfer functions (MTFs) of two reconstruction kernels as a filtering function in the spatial-frequency domain. This method is referred to as MTFratio filtering. Test image data were obtained from CT screening scans of 67 subjects who each had one nodule. Images were reconstructed using two kernels: fSTD (for standard lung imaging) and fSHARP (for sharp edge-enhancement lung imaging). The MTFratio filtering was implemented using the MTFs measured for those kernels and was applied to the reconstructed fSHARP images to obtain images that were similar to the fSTD images. A mean filter and a median filter were applied (separately) for comparison. All reconstructed and filtered images were processed using their prototype CAD system. The MTFratio filtered images showed excellent agreement with the fSTD images. The standard deviation for the difference between these images was very small, ∼6.0 Hounsfield units (HU). However, the mean and median filtered images showed larger differences of ∼48.1 and ∼57.9 HU from the fSTD images, respectively. The free-response receiver operating characteristic (FROC) curve for the fSHARP images indicated poorer performance compared with the FROC curve for the fSTD images. The FROC curve for the MTFratio filtered images was equivalent to the curve for the fSTD images. However, this similarity was not achieved by using the mean filter or median filter. The accuracy of MTFratio image filtering was verified and the method was demonstrated to be effective for reducing the kernel dependence of CAD performance.

Effect of radiation dose reduction and iterative reconstruction on computer-aided detection of pulmonary nodules: Intra-individual comparison

ObjectiveTo evaluate the effect of radiation dose reduction and iterative reconstruction (IR) on the performance of computer-aided detection (CAD) for pulmonary nodules. MethodsIn this prospective study twenty-five patients were included who were scanned for pulmonary nodule follow-up. Image acquisition was performed at routine dose and three reduced dose levels in a single session by decreasing mAs-values with 45%, 60% and 75%. Tube voltage was fixed at 120kVp for patients ≥80kg and 100kVp for patients <80kg. Data were reconstructed with filtered back projection (FBP), iDose4 (levels 1,4,6) and IMR (levels 1–3). All noncalcified solid pulmonary nodules ≥4mm identified by two radiologists in consensus served as the reference standard. Subsequently, nodule volume was measured with CAD software and compared to the reference consensus. The numbers of true-positives, false-positives and missed pulmonary nodules were evaluated as well as the sensitivity. ResultsMedian effective radiation dose was 2.2mSv at routine dose and 1.2, 0.9 and 0.6mSv at respectively 45%, 60% and 75% reduced dose. A total of 28 pulmonary nodules were included. With FBP at routine dose, 89% (25/28) of the nodules were correctly identified by CAD. This was similar at reduced dose levels with FBP, iDose4 and IMR. CAD resulted in a median number of false-positives findings of 11 per scan with FBP at routine dose (93% of the CAD marks) increasing to 15 per scan with iDose4 (95% of the CAD marks) and 26 per scan (96% of the CAD marks) with IMR at the lowest dose level. ConclusionCAD can identify pulmonary nodules at submillisievert dose levels with FBP, hybrid and model-based IR. However, the number of false-positive findings increased using hybrid and especially model-based IR at submillisievert dose while dose reduction did not affect the number of false-positives with FBP.

A new computer-aided detection scheme based on assessment of local bilateral mammographic feature asymmetry - a preliminary evaluation.

Accurate segmentation of breast lesions depicting on two-dimensional projection mammograms has been proven very difficult and unreliable. In this study we investigated a new approach of a computer-aided detection (CAD) scheme of mammograms without lesion segmentation. Our scheme was developed based on the detection and analysis of region-of-interest (ROI)-based bilateral mammographic tissue or feature asymmetry. A bilateral image registration, image feature selection process, and naïve Bayes linear classifier were implemented in CAD scheme. CAD performance predicting the likelihood of either an ROI or a subject (case) being abnormal was evaluated using 161 subjects from the mini-MIAS database and a leave-one-out testing method. The results showed that areas under receiver operating characteristic (ROC) curves were 0.87 and 0.72 on the ROI-based and case-based evaluation, respectively. The study demonstrated that using ROI-based bilateral mammographic tissue asymmetry can provide supplementary information with high discriminatory power in order to improve CAD performance.

Smart Colonography for Distributed Medical Databases with Group Kernel Feature Analysis

Computer-Aided Detection (CAD) of polyps in Computed Tomographic (CT) colonography is currently very limited since a single database at each hospital/institution doesn't provide sufficient data for training the CAD system's classification algorithm. To address this limitation, we propose to use multiple databases, (e.g., big data studies) to create multiple institution-wide databases using distributed computing technologies, which we call smart colonography. Smart colonography may be built by a larger colonography database networked through the participation of multiple institutions via distributed computing. The motivation herein is to create a distributed database that increases the detection accuracy of CAD diagnosis by covering many true-positive cases. Colonography data analysis is mutually accessible to increase the availability of resources so that the knowledge of radiologists is enhanced. In this article, we propose a scalable and efficient algorithm called Group Kernel Feature Analysis (GKFA), which can be applied to multiple cancer databases so that the overall performance of CAD is improved. The key idea behind the proposed GKFA method is to allow the feature space to be updated as the training proceeds with more data being fed from other institutions into the algorithm. Experimental results show that GKFA achieves very good classification accuracy.

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.

Potential clinical impact of advanced imaging and computer-aided diagnosis in chest radiology: importance of radiologist's role and successful observer study.

This review paper is based on our research experience in the past 30years. The importance of radiologists' role is discussed in the development or evaluation of new medical images and of computer-aided detection (CAD) schemes in chest radiology. The four main topics include (1) introducing what diseases can be included in a research database for different imaging techniques or CAD systems and what imaging database can be built by radiologists, (2) understanding how radiologists' subjective judgment can be combined with technical objective features to improve CAD performance, (3) sharing our experience in the design of successful observer performance studies, and (4) finally, discussing whether the new images and CAD systems can improve radiologists' diagnostic ability in chest radiology. In conclusion, advanced imaging techniques and detection/classification of CAD systems have a potential clinical impact on improvement of radiologists' diagnostic ability, for both the detection and the differential diagnosis of various lung diseases, in chest radiology.

Computer-aided detection of artificial pulmonary nodules using an ex vivo lung phantom: Influence of exposure parameters and iterative reconstruction

ObjectivesTo evaluate the influence of exposure parameters and raw-data based iterative reconstruction (IR) on the performance of computer-aided detection (CAD) of pulmonary nodules on chest multidetector computed tomography (MDCT). Material and methodsSeven porcine lung explants were inflated in a dedicated ex vivo phantom shell and prepared with n=162 artificial nodules of a clinically relevant volume and maximum diameter (46–1063μl, and 6.2–21.5mm). n=118 nodules were solid and n=44 part-solid. MDCT was performed with different combinations of 120 and 80kV with 120, 60, 30 and 12mA*s, and reconstructed with both filtered back projection (FBP) and IR. Subsequently, 16 datasets per lung were subjected to dedicated CAD software. The rate of true positive, false negative and false positive CAD marks was measured for each reconstruction. ResultsThe rate of true positive findings ranged between 88.9–91.4% for FBP and 88.3–90.1% for IR (n.s.) with most exposure settings, but was significantly lower with the combination of 80kV and 12mA*s (80.9% and 81.5%, respectively, p<0.05). False positive findings ranged between 2.3 - 8.1 annotations per lung. For nodule volumes <200μl the rate of true positives was significantly lower than for >300μl (p<0.05). Similarly, it was significantly lower for diameters <12mm compared to ≥12mm (p<0.05). The rate of true positives for solid and part-solid nodules was similar. ConclusionsNodule CAD on chest MDCT is robust over a wide range of exposure settings. Noise reduction by IR is not detrimental for CAD, and may be used to improve image quality in the setting of low-dose MDCT for lung cancer screening.

Computer-aided detection of screening breast cancer: a novel approach based on genetic programming

Mammography is the keystone of breast cancer screening. Yet high sensitivity is achieved at the cost of low specificity - only one-third of patients recalled will have breast cancer. Computer-aided detection (CAD) is a potentially valuable tool for assisting the breast radiologist to improve positive prediction values. However, to date, CAD has not reliably altered screening outcomes and the large proportion of false positives remains a drawback. We describe a novel method to improve CAD performance called Cartesian Genetic Programming (CGP); a machine-based learning algorithm, akin to genetic evolution.

Mammographic Image Denoising and Enhancement Using the Anscombe Transformation, Adaptive Wiener Filtering, and the Modulation Transfer Function

A new restoration methodology is proposed to enhance mammographic images through the improvement of contrast features and the simultaneous suppression of noise. Denoising is performed in the first step using the Anscombe transformation to convert the signal-dependent quantum noise into an approximately signal-independent Gaussian additive noise. In the Anscombe domain, noise is filtered through an adaptive Wiener filter, whose parameters are obtained by considering local image statistics. In the second step, a filter based on the modulation transfer function of the imaging system in the whole radiation field is applied for image enhancement. This methodology can be used as a preprocessing module for computer-aided detection (CAD) systems to improve the performance of breast cancer screening. A preliminary assessment of the restoration algorithm was performed using synthetic images with different levels of quantum noise. Afterward, we evaluated the effect of the preprocessing on the performance of a previously developed CAD system for clustered microcalcification detection in mammographic images. The results from the synthetic images showed an increase of up to 11.5 dB (p = 0.002) in the peak signal-to-noise ratio. Moreover, the mean structural similarity index increased up to 8.3 % (p < 0.001). Regarding CAD performance, the results suggested that the preprocessing increased the detectability of microcalcifications in mammographic images without increasing the false-positive rates. Receiver operating characteristic analysis revealed an average increase of 14.1 % (p = 0.01) in overall CAD performance when restored image sets were used.

Standalone computer-aided detection compared to radiologists’ performance for the detection of mammographic masses

We developed a computer-aided detection (CAD) system aimed at decision support for detection of malignant masses and architectural distortions in mammograms. The effect of this system on radiologists' performance depends strongly on its standalone performance. The purpose of this study was to compare the standalone performance of this CAD system to that of radiologists. In a retrospective study, nine certified screening radiologists and three residents read 200 digital screening mammograms without the use of CAD. Performances of the individual readers and of CAD were computed as the true-positive fraction (TPF) at a false-positive fraction of 0.05 and 0.2. Differences were analysed using an independent one-sample t-test. At a false-positive fraction of 0.05, the performance of CAD (TPF = 0.487) was similar to that of the certified screening radiologists (TPF = 0.518, P = 0.17). At a false-positive fraction of 0.2, CAD performance (TPF = 0.620) was significantly lower than the radiologist performance (TPF = 0.736, P <0.001). Compared to the residents, CAD performance was similar for all false-positive fractions. The sensitivity of CAD at a high specificity was comparable to that of human readers. These results show potential for CAD to be used as an independent reader in breast cancer screening.

Impact of breast density on computer-aided detection (CAD) of breast cancer

To evaluate the impact of breast density on computer-aided detection (CAD) for breast cancer and the CAD false-positive rate of normal controls. Two hundred and seventy-one histologically proven breast malignant lesions (from Feb. 2008 to Dec. 2009) and 238 randomly selected normal cases were classified by mammographic density according to the American College of Radiology breast imaging reporting and data system (BI-RADS). Mammograms of BI-RADS 1 or BI-RADS 2 density were categorized as non-dense breasts, and those of BI-RADS 3 or BI-RADS 4 density were categorized as dense breasts. Full-field digital mammography (GEMS Senographe) were performed in all patients and controls with craniocaudal (CC) and mediolateral oblique (MLO) views. Then the image data were transferred to review workstation (SenoAdvantage), and the lesions were marked by Second Look Digital CAD system (version 7.2, iCAD). The differences of sensitivity and false-positive rate between dense and non-dense breasts were compared. Overall, the sensitivity of CAD in detection of cancers was 84.1% (228/271), there was a statistically significant difference in CAD of cancers in dense versus non-dense breasts (P = 0.015). The sensitivity of CAD in detection of mass cancers was 76.5% (186/243), in detection of calcification cancers was 79.1% (125/158), there was no statistically significant difference in CAD performance for the detection of mass cancers versus calcification cancers (P = 0.547). There was a significant difference in the CAD performance for the detection of mass cancer cases in non-dense versus dense breasts (P = 0.001), but no significant difference in the CAD for the detection of calcification cancers in non-dense versus dense breasts (P = 0.216). In the controls, the distribution of mass false-positive marks did not differ significantly between non-dense and dense breast tissue cases (P = 0.207), but the distribution of calcification false-positive marks differed significantly between non-dense and dense breast tissue cases (P = 0.001). There was a statistically significant difference of false-positive marks in non-dense versus dense breasts (P = 0.043). The sensitivity of CAD in the detection of breast cancers is impacted by breast density. There is a statistically significant difference in the CAD performance for the detection of cancer cases in non-dense versus dense breasts. The false-positive rate of CAD is lower in dense versus non-dense breasts. It appears difficult for CAD in the early detection of breast cancer in the absence of microcalcifications, particularly in dense breasts.

Mammography screening using independent double reading with consensus: Is there a potential benefit for computer-aided detection?

Double reading improves the cancer detection rate in mammography screening. Single reading with computer-aided detection (CAD) has been considered to be an alternative to double reading. Little is known about the potential benefit of CAD in breast cancer screening with double reading. To compare prospective independent double reading of screen-film (SFM) and full-field digital (FFDM) mammography in population-based screening with retrospective standalone CAD performance on the baseline mammograms of the screen-detected cancers and subsequent cancers diagnosed during the follow-up period. The study had ethics committee approval. A 5-point rating scale for probability of cancer was used for 23,923 (SFM = 16,983; FFDM = 6940) screening mammograms. Of 208 evaluable cancers, 104 were screen-detected and 104 were subsequent (44 interval and 60 next screening round) cancers. Baseline mammograms of subsequent cancers were retrospectively classified in consensus without information about cancer location, histology, or CAD prompting as normal, non-specific minimal signs, significant minimal signs, and false-negatives. The baseline mammograms of the screen-detected cancers and subsequent cancers were evaluated by CAD. Significant minimal signs and false-negatives were considered 'actionable' and potentially diagnosable if correctly prompted by CAD. CAD correctly marked 94% (98/104) of the baseline mammograms of the screen-detected cancers (SFM = 95% [61/64]; FFDM = 93% [37/40]), including 96% (23/24) of those with discordant interpretations. Considering only those baseline examinations of subsequent cancers prospectively interpreted as normal and retrospectively categorized as 'actionable', CAD input at baseline screening had the potential to increase the cancer detection rate from 0.43% to 0.51% (P = 0.13); and to increase cancer detection by 16% ([104 + 17]/104) and decrease interval cancers by 20% (from 44 to 35). CAD may have the potential to increase cancer detection by up to 16%, and to reduce the number of interval cancers by up to 20% in SFM and FFDM screening programs using independent double reading with consensus review. The influence of true- and false-positive CAD marks on decision-making can, however, only be evaluated in a prospective clinical study.

FluorescenceIn SituHybridization (FISH) Signal Analysis Using Automated Generated Projection Images

Fluorescencein situhybridization (FISH) tests provide promising molecular imaging biomarkers to more accurately and reliably detect and diagnose cancers and genetic disorders. Since current manual FISH signal analysis is low-efficient and inconsistent, which limits its clinical utility, developing automated FISH image scanning systems and computer-aided detection (CAD) schemes has been attracting research interests. To acquire high-resolution FISH images in a multi-spectral scanning mode, a huge amount of image data with the stack of the multiple three-dimensional (3-D) image slices is generated from a single specimen. Automated preprocessing these scanned images to eliminate the non-useful and redundant data is important to make the automated FISH tests acceptable in clinical applications. In this study, a dual-detector fluorescence image scanning system was applied to scan four specimen slides with FISH-probed chromosome X. A CAD scheme was developed to detect analyzable interphase cells and map the multiple imaging slices recorded FISH-probed signals into the 2-D projection images. CAD scheme was then applied to each projection image to detect analyzable interphase cells using an adaptive multiple-threshold algorithm, identify FISH-probed signals using a top-hat transform, and compute the ratios between the normal and abnormal cells. To assess CAD performance, the FISH-probed signals were also independently visually detected by an observer. The Kappa coefficients for agreement between CAD and observer ranged from 0.69 to 1.0 in detecting/counting FISH signal spots in four testing samples. The study demonstrated the feasibility of automated FISH signal analysis that applying a CAD scheme to the automated generated 2-D projection images.

Fluorescence in situ hybridization (FISH) signal analysis using automated generated projection images.

Fluorescence in situ hybridization (FISH) tests provide promising molecular imaging biomarkers to more accurately and reliably detect and diagnose cancers and genetic disorders. Since current manual FISH signal analysis is low-efficient and inconsistent, which limits its clinical utility, developing automated FISH image scanning systems and computer-aided detection (CAD) schemes has been attracting research interests. To acquire high-resolution FISH images in a multi-spectral scanning mode, a huge amount of image data with the stack of the multiple three-dimensional (3-D) image slices is generated from a single specimen. Automated preprocessing these scanned images to eliminate the non-useful and redundant data is important to make the automated FISH tests acceptable in clinical applications. In this study, a dual-detector fluorescence image scanning system was applied to scan four specimen slides with FISH-probed chromosome X. A CAD scheme was developed to detect analyzable interphase cells and map the multiple imaging slices recorded FISH-probed signals into the 2-D projection images. CAD scheme was then applied to each projection image to detect analyzable interphase cells using an adaptive multiple-threshold algorithm, identify FISH-probed signals using a top-hat transform, and compute the ratios between the normal and abnormal cells. To assess CAD performance, the FISH-probed signals were also independently visually detected by an observer. The Kappa coefficients for agreement between CAD and observer ranged from 0.69 to 1.0 in detecting/counting FISH signal spots in four testing samples. The study demonstrated the feasibility of automated FISH signal analysis that applying a CAD scheme to the automated generated 2-D projection images.

Improving Performance of Computer-aided Detection of Masses by Incorporating Bilateral Mammographic Density Asymmetry: An Assessment

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. 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. 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). 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.

A robust and efficient approach to detect 3D rectal tubes from CT colonography

The rectal tube (RT) is a common source of false positives (FPs) in computer-aided detection (CAD) systems for CT colonography. A robust and efficient detection of RT can improve CAD performance by eliminating such "obvious" FPs and increase radiologists' confidence in CAD. In this paper, we present a novel and robust bottom-up approach to detect the RT. Probabilistic models, trained using kernel density estimation on simple low-level features, are employed to rank and select the most likely RT tube candidate on each axial slice. Then, a shape model, robustly estimated using random sample consensus (RANSAC), infers the global RT path from the selected local detections. Subimages around the RT path are projected into a subspace formed from training subimages of the RT. A quadratic discriminant analysis (QDA) provides a classification of a subimage as RT or non-RT based on the projection. Finally, a bottom-top clustering method is proposed to merge the classification predictions together to locate the tip position of the RT. Our method is validated using a diverse database, including data from five hospitals. On a testing data with 21 patients (42 volumes), 99.5% of annotated RT paths have been successfully detected. Evaluated with CAD, 98.4% of FPs caused by the RT have been detected and removed without any loss of sensitivity. The proposed method demonstrates a high detection rate of the RT path, and when tested in a CAD system, reduces FPs caused by the RT without the loss of sensitivity.

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 Detection: The Effect of Training Databases on Detection of Subtle Breast Masses

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. 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. 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. 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.

Electronic cleansing for CT colonography: does it help CAD software performance in a high-risk population for colorectal cancer?

To compare the performance of computer-aided detection (CAD) for CT colonography (CTC) with and without electronic cleansing (EC) in a high-risk population tagged with a faecal tagging (FT) protocol. Thirty-two patients underwent CTC followed by same-day colonoscopy. All patients underwent bowel preparation and FT with barium and gastrografin. Each CTC dataset was processed with colon CAD with and without EC. Per-polyp sensitivity was calculated. The average number of false-positive (FP) results and their causes were also analysed and compared. Eighty-six polyps were detected in 29 patients. Per-polyp sensitivities of CAD with EC (93.8% and 100%) were higher than those without EC (84.4% and 87.5%) for polyps >or=6 mm and >or=10 mm, respectively. However, the differences were not significant. The average number (6.3) of FPs of CAD with EC was significantly larger than that (3.1) without EC. The distribution of FPs in both CAD settings was also significantly different. The most common cause of FPs was the ileocaecal valve in both datasets. However, untagged faeces was a significantly less common cause of FPs with EC, EC-related artefacts being more common. Electronic cleansing has the potential to improve per-polyp sensitivity of CTC CAD, although the significantly larger number of FPs with EC remains to be improved.