Computer-aided diagnosis of breast tumors with different US systems

RATIONALE AND OBJECTIVES: The authors performed this study to determine whether a computer-aided diagnostic (CAD) system was suitable from one ultrasound (US) unit to another after parameters were adjusted by using intelligent selection algorithms. MATERIALS AND METHODS: The authors used texture analysis and data mining with a decision tree model to classify breast tumors with different US systems. The databases of training cases from one unit and testing cases from another were collected from different countries. Regions of interest on US scans and co-variance texture parameters were used in the diagnosis system. Proposed adjustment schemes for different US systems were used to transform the information needed for a differential diagnosis. RESULTS: Comparison of the diagnostic system with and without adjustment, respectively, yielded the following results: accuracy, 89.9% and 82.2%; sensitivity, 94.6% and 92.2%; specificity, 85.4% and 72.3%; positive predictive value, 86.5% and 76.8%; and negative predictive value, 94.1% and 90.4%. The improvement in accuracy, specificity, and positive predictive value was statistically significant. Diagnostic performance was improved after the adjustment. CONCLUSION: After parameters were adjusted by using intelligent selection algorithms, the performance of the proposed CAD system was better both with the same and with different systems. Different resolutions, different setting conditions, and different scanner ages are no longer obstacles to the application of such a CAD system.     

Through-transmission US applied to breast imaging

RATIONALE AND OBJECTIVES: The purpose of this study was to evaluate the application of a new method of through-transmission ultrasonography (US) to breast imaging. MATERIALS AND METHODS: Through-transmission US was used to image breast tissue in 18 women; 12 had no known lesions, and six had masses, five of which were palpable. Of those with no known lesions, two had implants, and two had mammary duct ectasia. Lesion sizes and locations on the through-transmission sonograms were correlated with findings from conventional imaging to determine successful imaging. Histopathologic findings and lesion size were recorded from pathology reports. Detailed comparative analyses of through-transmission US, conventional x-ray mammography, and conventional US were performed. The following lesion variables were compared: maximum diameter, shape, margins, and internal architecture. RESULTS: Through-transmission US produced images of all masses, some of which were mammographically occult. Comparisons among through-transmission US, mammography, and conventional US showed high correlation in lesion characteristics such as maximum diameter, shape, margins, and internal architecture. For two malignant masses, through-transmission US appeared to be more accurate in assessing tumor extent, compared with mammography and conventional US. CONCLUSION: This new method of through-transmission US produces images of breast tissue in women with a variety of breast types and can be used to detect and characterize a variety of lesions, some of which are mammographically occult.     

Computer-aided diagnosis of breast cancer

Computer-aided diagnosis (CAD) is generally defined as a diagnosis made by a radiologist who takes into account computer output as a "second opinion". The computer output usually indicates the location of suspicious lesions such as microcalcifications or masses in mammograms, and also provides quantitative analysis of potential lesions, yielding measures such as the likelihood of malignancy. The purpose of CAD is to improve diagnostic accuracy and the consistency of radiologists' image interpretation by using such computer output as a guide. This article provides a brief overview of some CAD schemes used for breast cancer.     

Computer-aided diagnosis: a shape classification of pulmonary nodules imaged by high-resolution CT.

We investigated the possibility of using computer analysis of high-resolution CT images to radiologically classify the shape of pulmonary nodules. Using a combination of circularity and second moment as quantitative measures we were able to classify pulmonary nodules in each shape group as effectively as could a radiologist. We found that pulmonary nodules with circularity < or =0.75 and second moment < or =0.18 were very likely to reveal lung cancer.     

Computer-aided detection and diagnosis of breast cancer

The limitations of radiologists when interpreting mammogram examinations provides a reasonable, if not compelling, basis for application of computer techniques that have the potential to improve diagnostic performance. Computer algorithms, at their present state of development, show great promise for clinical use. It can be expected that such use will only improve as computer technology and computer methods continue to become more formidable. The eventual role of computers in mammographic detection and diagnosis has not been fully defined, but their effect on practice may one day be very significant.     

Computer-aided periodontal disease diagnosis using computer vision.

Periodontal diseases are the major cause of tooth loss. The study of the evolution of these diseases is crucial to achieve adequate planning and treatment. Depth probing is essential to know the periodontal disease stage. In this paper we present a new system for Computer-Aided Periodontal Disease Diagnosis using computer vision. The system automates the depth probing and incorporates a colour camera fitted together with a plastic probe that automatically and exactly obtains the depth probing measure. The system has been tested by several periodontists and with 125 teeth of different patients. The differences between the values taken by the system and two periodontists have not been significant.     

Classification of brain compartments and head injury lesions by neural networks applied to MRI

An automatic, neural network-based approach was applied to segment normal brain compartments and lesions on MR images. Two supervised networks, backpropagation (BPN) and counterpropagation, and two unsupervised networks, Kohonen learning vector quantizer and analog adaptive resonance theory, were trained on registered T2-weighted and proton density images. The classes of interest were background, gray matter, white matter, cerebrospinal fluid, macrocystic encephalomalacia, gliosis, and unknown. A comprehensive feature vector was chosen to discriminate these classes. The BPN combined with feature conditioning, multiple discriminant analysis followed by Hotelling transform, produced the most accurate and consistent classification results. Classifications of normal brain compartments were generally in agreement with expert interpretation of the images. Macrocystic encephalomalacia and gliosis were recognized and, except around the periphery, classified in agreement with the clinician's report used to train the neural network.     

Computer-aided diagnosis in breast DCE-MRI--quantification of the heterogeneity of breast lesions

Purpose

In our study we aim at the quantification of the heterogeneity for differential diagnosis of breast lesions in MRI.

Materials and methods

We tested a software tool for quantification of heterogeneity. The software tool provides a three-dimensional analysis of the whole breast lesion. The lesions were divided in regions with similar perfusion characteristics. Voxels were merged to the same region, if the perfusion parameters (wash-in, wash-out, integral, peak enhancement and time to peak) correlated to 99%. We evaluated 68 lesions from 50 patients. 31 lesions proved to be benign (45.6%) and 37 malignant (54.4%). We included small lesions which could only be detected with MRI.

Results

The analysis of heterogeneity showed significant differences (p < 0.005; AUC 0.7). Malignant lesions were more heterogeneous than benign ones. Significant differences were also found for morphologic parameters such as shape (p < 0.001) and margin (p < 0.007). The analysis of the enhancement dynamics did not prove successful in lesion discrimination.

Conclusion

Our study indicates that the region analysis for quantification of heterogeneity may be a helpful additional method to differentiate benign lesions from malignant ones.     

Computer-aided diagnosis of pulmonary nodules: results of a large-scale observer test

PURPOSE: To determine the frequency on magnetic resonance (MR) images of a widened pericholecystic space, which the authors call the expanded gallbladder fossa sign, and to assess the sensitivity and specificity of this sign for cirrhosis. MATERIALS AND METHODS: Three-hundred thirteen patients who underwent MR imaging were included in this study, including 190 with pathologically proved cirrhosis (cirrhosis patients) and 123 without history of chronic liver diseases (control subjects). MR images were qualitatively evaluated by three independent observers for the presence of the expanded gallbladder fossa sign. This sign was considered present if there was enlargement of the pericholecystic space (i.e., gallbladder fossa) and the space was bounded laterally by the edge of the right hepatic lobe and medially by the edge of the left lateral segment, in conjunction with nonvisualization of the left medial segment. RESULTS: The expanded gallbladder fossa sign was seen in 129 cirrhosis patients and in three control subjects (P < .001). The sensitivity, specificity, accuracy, and positive predictive value of this sign for the MR diagnosis of cirrhosis were 68%, 98%, 80%, and 98%, respectively. CONCLUSION: The expanded gallbladder fossa sign on MR images is a frequently present, specific indicator of cirrhosis. This sign can be used as a simple and highly specific sign of cirrhosis, if present, despite the overall sensitivity of 68%.     

Computer-aided detection of solid lung nodules in lossy compressed multidetector computed tomography chest exams

RATIONALE AND OBJECTIVES: To assess the effect of three-dimensional (3D) lossy image compression of multidetector computed tomography chest scans on computer-aided detection (CAD) of solid lung nodules greater than 4 mm in size. MATERIALS AND METHODS: A total of 120 cases, acquired with 1.25-mm collimation, were collected from 5 different sites, of which 66/120 were low-dose cases. Two chest radiologists established that 37 cases had no actionable lung nodules; the remaining 83 cases contained 169 nodules (range 3.8-35.0 mm, mean 5.8 mm +/- 3.0 [SD]). All cases were compressed using the 3D Set Partitioning in Hierarchical Trees algorithm to 24:1, 48:1, and 96:1 levels. A study of the effect of compression on computer-aided detection (CAD) sensitivity was performed at operating points of 2.5 false marks (FM), 5 FM, and 10 FM per case using McNemar's test. Logistic regression models were used to evaluate the impact on CAD sensitivity by compression level on nodule and image characteristics. RESULTS: Compared with no compression, there was no significant degradation in CAD sensitivity found at any of the studied compression levels and operating points. However, between compression levels, there was marginal association with sensitivity. Specifically, 24:1 level was significantly better than 96:1 at all operating points, and occasionally better than no compression at 10 FM/case. Based on multivariate analysis, nodule location was found to be a significant predictor (P = .01) with a lower sensitivity associated with juxtapleural nodules. Nodule size, dose, reconstruction filter, and contrast medium were not significant predictors. CONCLUSION: CAD detection performance of solid lung nodules did not suffer until 48:1 compression.     

Role of combined grey scale US and US tissue elastography in differentiating solid thyroid nodules

Background

Ultrasonographic (US) examination is an accurate method for detecting thyroid nodules, but its use in differentiating between benign and malignant thyroid nodules is relatively low. US elastography has been applied to study the hardness/elasticity of nodules to differentiate malignant from benign lesions thus deviating a significant group of patients from unnecessary FNAB.

Objectives

The aim of the study is to evaluate the validity of combined grey scale US and tissue elastography in differentiating benign form malignant solid thyroid nodules.

Methods

The study included 46 selected patients with solid thyroid nodules according to our inclusion and exclusion criteria. The patients underwent surgery for compressive symptoms or suspicion of malignancy on FNA cytology. US features and tissue elastography were scored according to the Rago criteria (1).

Results

On US elastography: all the 31 cases with a final diagnosis of benign nodule had a score of 1–3, while 14 of 15 (94.1%) with a final diagnosis of carcinoma had a score of 4–5, with a sensitivity of 93.3%, a specificity of 100% and an accuracy of 97.8%. Combined US and elastography reveals that hypoechogenicity/score 4–5 was most predictive of malignancy with sensitivity 80% and specificity 100%; and accuracy 93.4%.

Conclusions

US elastography seems to have great potential as a new tool for differentiating solid thyroid nodules and for recommending FNAC. Combined grey scale US features and US elastography added no significant value when compared with US elastography alone. Further prospective studies are needed.     

Computer-aided diagnosis for the differentiation of malignant from benign thyroid nodules on ultrasonography

RATIONALE AND OBJECTIVES: We sought to evaluate the diagnostic performance of an artificial neural network (ANN) and binary logistic regression (BLR) in differentiating malignant from benign thyroid nodules on ultrasonography. MATERIALS AND METHODS: Two experienced radiologists, who were unaware of the histopathological diagnosis, analyzed ultrasonographic (US) features of 109 pathologically proven thyroid lesions (49 malignant and 60 benign) in 96 patients. Each radiologist was asked to evaluate US findings and categorize nodules into one of the two groups (malignant vs. benign) in each case. The following 8 US parameters were assessed for each nodule: size, shape, margin, echogenicity, cystic change, microcalcification, macrocalcification, and halo sign. Statistically significant US findings were obtained with backward stepwise logistic regression and were used for training and testing of the ANN and the BLR. The performance of the ANN and BLR was compared to that of the radiologists using receiver-operating characteristic (ROC) analysis. RESULTS: Statistically significant US findings were size, margin, echogenicity, cystic change, and macrocalcification of the nodules. The area under the ROC curve (Az) values of ANN and BLR were 0.9492 +/- 0.0195 and 0.9046 +/- 0.0289, respectively. The Az value was 0.8300 +/- 0.0359 for reader 1 and 0.7600 +/- 0.0409 for reader 2. The Az values for ANN and BLR were significantly higher than those for both radiologists (all p < .05). CONCLUSION: The performance of the ANN and the BLR was better than that of the radiologists in the distinction of benign and malignant thyroid nodules.