Improved artificial neural networks in prediction of malignancy of lesions in contrast-enhanced MR-mammography

A multicentric study of radiomics and artificial intelligence analysis on contrast-enhanced mammography to identify different histotypes of breast cancer

OBJECTIVE

To evaluate the performance of radiomic analysis on contrast-enhanced mammography images to identify different histotypes of breast cancer mainly in order to predict grading, to identify hormone receptors, to discriminate human epidermal growth factor receptor 2 (HER2) and to identify luminal histotype of the breast cancer.

METHODS

From four Italian centers were recruited 180 malignant lesions and 68 benign lesions. However, only the malignant lesions were considered for the analysis. All patients underwent contrast-enhanced mammography in cranium caudal (CC) and medium lateral oblique (MLO) view. Considering histological findings as the ground truth, four outcomes were considered: (1) G1 + G2 vs. G3; (2) HER2 + vs. HER2 - ; (3) HR + vs. HR - ; and (4) non-luminal vs. luminal A or HR + /HER2-  and luminal B or HR + /HER2 + . For multivariate analysis feature selection, balancing techniques and patter recognition approaches were considered.

RESULTS

The univariate findings showed that the diagnostic performance is low for each outcome, while the results of the multivariate analysis showed that better performances can be obtained. In the HER2 + detection, the best performance (73% of accuracy and AUC = 0.77) was obtained using a linear regression model (LRM) with 12 features extracted by MLO view. In the HR + detection, the best performance (77% of accuracy and AUC = 0.80) was obtained using a LRM with 14 features extracted by MLO view. In grading classification, the best performance was obtained by a decision tree trained with three predictors extracted by MLO view reaching an accuracy of 82% on validation set. In the luminal versus non-luminal histotype classification, the best performance was obtained by a bagged tree trained with 15 predictors extracted by CC view reaching an accuracy of 94% on validation set.

CONCLUSIONS

The results suggest that radiomics analysis can be effectively applied to design a tool to support physician decision making in breast cancer classification. In particular, the classification of luminal versus non-luminal histotypes can be performed with high accuracy.

Radiomics and artificial intelligence analysis by T2-weighted imaging and dynamic contrast-enhanced magnetic resonance imaging to predict Breast Cancer Histological Outcome

OBJECTIVE

The objective of the study was to evaluate the accuracy of radiomics features obtained by MR images to predict Breast Cancer Histological Outcome.

METHODS

A total of 217 patients with malignant lesions were analysed underwent MRI examinations. Considering histological findings as the ground truth, four different types of findings were used in both univariate and multivariate analyses: (1) G1 + G2 vs G3 classification; (2) presence of human epidermal growth factor receptor 2 (HER2 + vs HER2 -); (3) presence of the hormone receptor (HR + vs HR -); and (4) presence of luminal subtypes of breast cancer.

RESULTS

The best accuracy for discriminating HER2 + versus HER2 - breast cancers was obtained considering nine predictors by early phase T1-weighted subtraction images and a decision tree (accuracy of 88% on validation set). The best accuracy for discriminating HR + versus HR - breast cancers was obtained considering nine predictors by T2-weighted subtraction images and a decision tree (accuracy of 90% on validation set). The best accuracy for discriminating G1 + G2 versus G3 breast cancers was obtained considering 16 predictors by early phase T1-weighted subtraction images in a linear regression model with an accuracy of 75%. The best accuracy for discriminating luminal versus non-luminal breast cancers was obtained considering 27 predictors by early phase T1-weighted subtraction images and a decision tree (accuracy of 94% on validation set).

CONCLUSIONS

The combination of radiomics analysis and artificial intelligence techniques could be used to support physician decision-making in prediction of Breast Cancer Histological Outcome.

Prediction of Breast Cancer Histological Outcome by Radiomics and Artificial Intelligence Analysis in Contrast-Enhanced Mammography

Simple Summary

The assessment of breast lesions through mammographic images is currently challenging, especially in dense breasts. Contrast-enhanced mammography has been shown to overcome the limitations of standard mammography but it greatly depends on the interpretative skills of the physician. The aim of this study was to evaluate the potentialities of statistical and artificial intelligence algorithms as a tool for helping the radiologists in the interpretation of images. The most remarkable results were achieved in discriminating benign from malignant lesions and in the identification of the presence of the hormone receptor. A tool to support the physician’s decision-making process may be designed starting from simple logistic regression and tree-based algorithms. This type of tool may help the radiologist in assessing the investigated breast and in choosing the appropriate follow-up without resorting to histology.

Abstract

Purpose: To evaluate radiomics features in order to: differentiate malignant versus benign lesions; predict low versus moderate and high grading; identify positive or negative hormone receptors; and discriminate positive versus negative human epidermal growth factor receptor 2 related to breast cancer. Methods: A total of 182 patients with known breast lesions and that underwent Contrast-Enhanced Mammography were enrolled in this retrospective study. The reference standard was pathology (118 malignant lesions and 64 benign lesions). A total of 837 textural metrics were extracted by manually segmenting the region of interest from both craniocaudally (CC) and mediolateral oblique (MLO) views. Non-parametric Wilcoxon–Mann–Whitney test, receiver operating characteristic, logistic regression and tree-based machine learning algorithms were used. The Adaptive Synthetic Sampling balancing approach was used and a feature selection process was implemented. Results: In univariate analysis, the classification of malignant versus benign lesions achieved the best performance when considering the original_gldm_DependenceNonUniformity feature extracted on CC view (accuracy of 88.98%). An accuracy of 83.65% was reached in the classification of grading, whereas a slightly lower value of accuracy (81.65%) was found in the classification of the presence of the hormone receptor; the features extracted were the original_glrlm_RunEntropy and the original_gldm_DependenceNonUniformity, respectively. The results of multivariate analysis achieved the best performances when using two or more features as predictors for classifying malignant versus benign lesions from CC view images (max test accuracy of 95.83% with a non-regularized logistic regression). Considering the features extracted from MLO view images, the best test accuracy (91.67%) was obtained when predicting the grading using a classification-tree algorithm. Combinations of only two features, extracted from both CC and MLO views, always showed test accuracy values greater than or equal to 90.00%, with the only exception being the prediction of the human epidermal growth factor receptor 2, where the best performance (test accuracy of 89.29%) was obtained with the random forest algorithm. Conclusions: The results confirm that the identification of malignant breast lesions and the differentiation of histological outcomes and some molecular subtypes of tumors (mainly positive hormone receptor tumors) can be obtained with satisfactory accuracy through both univariate and multivariate analysis of textural features extracted from Contrast-Enhanced Mammography images.

Radiomic and Artificial Intelligence Analysis with Textural Metrics Extracted by Contrast-Enhanced Mammography and Dynamic Contrast Magnetic Resonance Imaging to Detect Breast Malignant Lesions

Purpose:The purpose of this study was to discriminate between benign and malignant breast lesions through several classifiers using, as predictors, radiomic metrics extracted from CEM and DCE-MRI images. In order to optimize the analysis, balancing and feature selection procedures were performed. Methods: Fifty-four patients with 79 histo-pathologically proven breast lesions (48 malignant lesions and 31 benign lesions) underwent both CEM and DCE-MRI. The lesions were retrospectively analyzed with radiomic and artificial intelligence approaches. Forty-eight textural metrics were extracted, and univariate and multivariate analyses were performed: non-parametric statistical test, receiver operating characteristic (ROC) and machine learning classifiers. Results: Considering the single metrics extracted from CEM, the best predictors were KURTOSIS (area under ROC curve (AUC) = 0.71) and SKEWNESS (AUC = 0.71) calculated on late MLO view. Considering the features calculated from DCE-MRI, the best predictors were RANGE (AUC = 0.72), ENERGY (AUC = 0.72), ENTROPY (AUC = 0.70) and GLN (gray-level nonuniformity) of the gray-level run-length matrix (AUC = 0.72). Considering the analysis with classifiers and an unbalanced dataset, no significant results were obtained. After the balancing and feature selection procedures, higher values of accuracy, specificity and AUC were reached. The best performance was obtained considering 18 robust features among all metrics derived from CEM and DCE-MRI, using a linear discriminant analysis (accuracy of 0.84 and AUC = 0.88). Conclusions: Classifiers, adjusted with adaptive synthetic sampling and feature selection, allowed for increased diagnostic performance of CEM and DCE-MRI in the differentiation between benign and malignant lesions.

Radiomics and Artificial Intelligence Analysis with Textural Metrics Extracted by Contrast-Enhanced Mammography in the Breast Lesions Classification

The aim of the study was to estimate the diagnostic accuracy of textural features extracted by dual-energy contrast-enhanced mammography (CEM) images, by carrying out univariate and multivariate statistical analyses including artificial intelligence approaches. In total, 80 patients with known breast lesion were enrolled in this prospective study according to regulations issued by the local Institutional Review Board. All patients underwent dual-energy CEM examination in both craniocaudally (CC) and double acquisition of mediolateral oblique (MLO) projections (early and late). The reference standard was pathology from a surgical specimen for malignant lesions and pathology from a surgical specimen or fine needle aspiration cytology, core or Tru-Cut needle biopsy, and vacuum assisted breast biopsy for benign lesions. In total, 104 samples of 80 patients were analyzed. Furthermore, 48 textural parameters were extracted by manually segmenting regions of interest. Univariate and multivariate approaches were performed: non-parametric Wilcoxon–Mann–Whitney test; receiver operating characteristic (ROC), linear classifier (LDA), decision tree (DT), k-nearest neighbors (KNN), artificial neural network (NNET), and support vector machine (SVM) were utilized. A balancing approach and feature selection methods were used. The univariate analysis showed low accuracy and area under the curve (AUC) for all considered features. Instead, in the multivariate textural analysis, the best performance considering the CC view (accuracy (ACC) = 0.75; AUC = 0.82) was reached with a DT trained with leave-one-out cross-variation (LOOCV) and balanced data (with adaptive synthetic (ADASYN) function) and a subset of three robust textural features (MAD, VARIANCE, and LRLGE). The best performance (ACC = 0.77; AUC = 0.83) considering the early-MLO view was reached with a NNET trained with LOOCV and balanced data (with ADASYN function) and a subset of ten robust features (MEAN, MAD, RANGE, IQR, VARIANCE, CORRELATION, RLV, COARSNESS, BUSYNESS, and STRENGTH). The best performance (ACC = 0.73; AUC = 0.82) considering the late-MLO view was reached with a NNET trained with LOOCV and balanced data (with ADASYN function) and a subset of eleven robust features (MODE, MEDIAN, RANGE, RLN, LRLGE, RLV, LZLGE, GLV_GLSZM, ZSV, COARSNESS, and BUSYNESS). Multivariate analyses using pattern recognition approaches, considering 144 textural features extracted from all three mammographic projections (CC, early MLO, and late MLO), optimized by adaptive synthetic sampling and feature selection operations obtained the best results (ACC = 0.87; AUC = 0.90) and showed the best performance in the discrimination of benign and malignant lesions.

Development of machine learning models to predict RT-PCR results for severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) in patients with influenza-like symptoms using only basic clinical data

BACKGROUND

Reverse Transcription-Polymerase Chain Reaction (RT-PCR) for Severe Acute Respiratory Syndrome Coronavirus 2 (SARS-COV-2) diagnosis currently requires quite a long time span. A quicker and more efficient diagnostic tool in emergency departments could improve management during this global crisis. Our main goal was assessing the accuracy of artificial intelligence in predicting the results of RT-PCR for SARS-COV-2, using basic information at hand in all emergency departments.

METHODS

This is a retrospective study carried out between February 22, 2020 and March 16, 2020 in one of the main hospitals in Milan, Italy. We screened for eligibility all patients admitted with influenza-like symptoms tested for SARS-COV-2. Patients under 12 years old and patients in whom the leukocyte formula was not performed in the ED were excluded. Input data through artificial intelligence were made up of a combination of clinical, radiological and routine laboratory data upon hospital admission. Different Machine Learning algorithms available on WEKA data mining software and on Semeion Research Centre depository were trained using both the Training and Testing and the K-fold cross-validation protocol.

RESULTS

Among 199 patients subject to study (median [interquartile range] age 65 [46-78] years; 127 [63.8%] men), 124 [62.3%] resulted positive to SARS-COV-2. The best Machine Learning System reached an accuracy of 91.4% with 94.1% sensitivity and 88.7% specificity.

CONCLUSION

Our study suggests that properly trained artificial intelligence algorithms may be able to predict correct results in RT-PCR for SARS-COV-2, using basic clinical data. If confirmed, on a larger-scale study, this approach could have important clinical and organizational implications.

Decision support system for breast lesions via dynamic contrast enhanced magnetic resonance imaging

The presented study aims to design a computer-aided detection and diagnosis system for breast dynamic contrast enhanced magnetic resonance imaging. In the proposed system, the segmentation task is performed in two stages. The first stage is called breast region segmentation in which adaptive noise filtering, local adaptive thresholding, connected component analysis, integral of horizontal projection, and breast region of interest detection algorithms are applied to the breast images consecutively. The second stage of segmentation is breast lesion detection that consists of 32-class Otsu thresholding and Markov random field techniques. Histogram, gray level co-occurrence matrix and neighboring gray tone difference matrix based feature extraction, Fisher score based feature selection and, tenfold and leave-one-out cross-validation steps are carried out after segmentation to increase the reliability of the designed system while decreasing the computational time. Finally, support vector machines, k- nearest neighbor, and artificial neural network classifiers are performed to separate the breast lesions as benign and malignant. The average accuracy, sensitivity, specificity, and positive predictive values of each classifier are calculated and the best results are compared with the existing similar studies. According to the achieved results, the proposed decision support system for breast lesion segmentation distinguishes the breast lesions with 86%, 100%, 67%, and 85% accuracy, sensitivity, specificity, and positive predictive values, respectively. These results show that the proposed system can be used to support the radiologists during a breast cancer diagnosis.

The "MS-ROM/IFAST" Model, a Novel Parallel Nonlinear EEG Analysis Technique, Distinguishes ASD Subjects From Children Affected With Other Neuropsychiatric Disorders With High Degree of Accuracy

Background and Objective. In a previous study, we showed a new EEG processing methodology called Multi-Scale Ranked Organizing Map/Implicit Function As Squashing Time (MS-ROM/IFAST) performing an almost perfect distinction between computerized EEG of Italian children with autism spectrum disorder (ASD) and typically developing children. In this study, we assessed this system in distinguishing ASD subjects from children affected with other neuropsychiatric disorders (NPD). Methods. At a psychiatric practice in Texas, 20 children diagnosed with ASD and 20 children diagnosed with NPD were entered into the study. Continuous segments of artifact-free EEG data lasting 10 minutes were entered in MS-ROM/IFAST. From the new variables created by MS-ROM/IFAST, only 12 has been selected according to a correlation criterion. The selected features represent the input on which supervised machine learning systems (MLS) acted as blind classifiers. Results. The overall predictive capability in distinguishing ASD from other NPD cases ranged from 93% to 97.5%. The results were confirmed in further experiments in which Italian and US data have been combined. In this analysis, the best MLS reached 95.0% global accuracy in 1 out of 3 classes distinction (ASD, NPD, controls). This study demonstrates the value of EEG processing with advanced MLS in the differential diagnosis between ASD and NPD cases. The results were not affected by age, ethnicity and technicalities of EEG acquisition, confirming the existence of a specific EEG signature in ASD cases. To further support these findings, it was decided to test the behavior of already trained neural networks on 10 Italian very young ASD children (25-37 months). In this test, 9 out of 10 cases have been correctly recognized as ASD subjects in the best case. Conclusions. These results confirm the possibility of an early automatic autism detection based on standard EEG.

Hybrid ANN optimized artificial fish swarm algorithm based classifier for classification of suspicious lesions in breast DCE-MRI

Automatic mass or lesion classification systems are developed to aid in distinguishing between malignant and benign lesions present in the breast DCE-MR images, the systems need to improve both the sensitivity and specificity of DCE-MR image interpretation in order to be successful for clinical use. A new classifier (a set of features together with a classification method) based on artificial neural networks trained using artificial fish swarm optimization (AFSO) algorithm is proposed in this paper. The basic idea behind the proposed classifier is to use AFSO algorithm for searching the best combination of synaptic weights for the neural network. An optimal set of features based on the statistical textural features is presented. The investigational outcomes of the proposed suspicious lesion classifier algorithm therefore confirm that the resulting classifier performs better than other such classifiers reported in the literature. Therefore this classifier demonstrates that the improvement in both the sensitivity and specificity are possible through automated image analysis.

Detection of masses and microcalcifications in digital mammogram images using fuzzy logic

Background

Detection of small breast lesions is a challenging task for radiologists. Computer aided detection (CAD) systems are implemented to aid radiologists in detecting masses and microcalcifications. This has the potential to raise the level of sensitivity in breast cancer detection.

Objectives

To evaluate a new system to detect suggestions of suspicious small lesions.

Methods

Small samples were extracted from different tissue types. Texture features were calculated, and the best features were selected using Waikato Environment for Knowledge Analysis (WEKA) software. Subsequently, 7 selected features were used to form a decision tree. To reduce false negative cases, fuzzy logic was used. In the implementation phase, input images were divided into 8 pixel ´ 8 pixel tiles. For each tile, all selected features were computed as fuzzy inputs.

Results

To evaluate the technique, the suggested system was applied to 326 images obtained from the National Cancer Society of Malaysia. Based on this application, results showed that the suggested system has an acceptable sensitivity of 85.6% and specificity of 90.7%.

Conclusions

The fuzzy system is a promising technique for early detection of breast cancer.

Pattern Recognition Approaches for Breast Cancer DCE-MRI Classification: A Systematic Review

We performed a systematic review of several pattern analysis approaches for classifying breast lesions using dynamic, morphological, and textural features in dynamic contrast-enhanced magnetic resonance imaging (DCE-MRI). Several machine learning approaches, namely artificial neural networks (ANN), support vector machines (SVM), linear discriminant analysis (LDA), tree-based classifiers (TC), and Bayesian classifiers (BC), and features used for classification are described. The findings of a systematic review of 26 studies are presented. The sensitivity and specificity are respectively 91 and 83 % for ANN, 85 and 82 % for SVM, 96 and 85 % for LDA, 92 and 87 % for TC, and 82 and 85 % for BC. The sensitivity and specificity are respectively 82 and 74 % for dynamic features, 93 and 60 % for morphological features, 88 and 81 % for textural features, 95 and 86 % for a combination of dynamic and morphological features, and 88 and 84 % for a combination of dynamic, morphological, and other features. LDA and TC have the best performance. A combination of dynamic and morphological features gives the best performance.

Outcome predictors in autism spectrum disorders preschoolers undergoing treatment as usual: insights from an observational study using artificial neural networks

BACKGROUND

Treatment as usual (TAU) for autism spectrum disorders (ASDs) includes eclectic treatments usually available in the community and school inclusion with an individual support teacher. Artificial neural networks (ANNs) have never been used to study the effects of treatment in ASDs. The Auto Contractive Map (Auto-CM) is a kind of ANN able to discover trends and associations among variables creating a semantic connectivity map. The matrix of connections, visualized through a minimum spanning tree filter, takes into account nonlinear associations among variables and captures connection schemes among clusters. Our aim is to use Auto-CM to recognize variables to discriminate between responders versus no responders at TAU.

METHODS

A total of 56 preschoolers with ASDs were recruited at different sites in Italy. They were evaluated at T0 and after 6 months of treatment (T1). The children were referred to community providers for usual treatments.

RESULTS

At T1, the severity of autism measured through the Autism Diagnostic Observation Schedule decreased in 62% of involved children (Response), whereas it was the same or worse in 37% of the children (No Response). The application of the Semeion ANNs overcomes the 85% of global accuracy (Sine Net almost reaching 90%). Consequently, some of the tested algorithms were able to find a good correlation between some variables and TAU outcome. The semantic connectivity map obtained with the application of the Auto-CM system showed results that clearly indicated that "Response" cases can be visually separated from the "No Response" cases. It was possible to visualize a response area characterized by "Parents Involvement high". The resultant No Response area strongly connected with "Parents Involvement low".

CONCLUSION

The ANN model used in this study seems to be a promising tool for the identification of the variables involved in the positive response to TAU in autism.

A vector machine formulation with application to the computer-aided diagnosis of breast cancer from DCE-MRI screening examinations

This study investigates the use of a proposed vector machine formulation with application to dynamic contrast-enhanced magnetic resonance imaging examinations in the context of the computer-aided diagnosis of breast cancer. This paper describes a method for generating feature measurements that characterize a lesion's vascular heterogeneity as well as a supervised learning formulation that represents an improvement over the conventional support vector machine in this application. Spatially varying signal-intensity measures were extracted from the examinations using principal components analysis and the machine learning technique known as the support vector machine (SVM) was used to classify the results. An alternative vector machine formulation was found to improve on the results produced by the established SVM in randomized bootstrap validation trials, yielding a receiver-operating characteristic curve area of 0.82 which represents a statistically significant improvement over the SVM technique in this application.

A new endoscopic ultrasonography image processing method to evaluate the prognosis for pancreatic cancer treated with interstitial brachytherapy

AIM: To develop a fuzzy classification method to score the texture features of pancreatic cancer in endoscopic ultrasonography (EUS) images and evaluate its utility in making prognosis judgments for patients with unresectable pancreatic cancer treated by EUS-guided interstitial brachytherapy.

METHODS: EUS images from our retrospective database were analyzed. The regions of interest were drawn, and texture features were extracted, selected, and scored with a fuzzy classification method using a C++ program. Then, patients with unresectable pancreatic cancer were enrolled to receive EUS-guided iodine 125 radioactive seed implantation. Their fuzzy classification scores, tumor volumes, and carbohydrate antigen 199 (CA199) levels before and after the brachytherapy were recorded. The association between the changes in these parameters and overall survival was analyzed statistically.

RESULTS: EUS images of 153 patients with pancreatic cancer and 63 non-cancer patients were analyzed. A total of 25 consecutive patients were enrolled, and they tolerated the brachytherapy well without any complications. There was a correlation between the change in the fuzzy classification score and overall survival (Spearman test, r = 0.616, P = 0.001), whereas no correlation was found to be significant between the change in tumor volume (P = 0.663), CA199 level (P = 0.659), and overall survival. There were 15 patients with a decrease in their fuzzy classification score after brachytherapy, whereas the fuzzy classification score increased in another 10 patients. There was a significant difference in overall survival between the two groups (67 d vs 151 d, P = 0.001), but not in the change of tumor volume and CA199 level.

CONCLUSION: Using the fuzzy classification method to analyze EUS images of pancreatic cancer is feasible, and the method can be used to make prognosis judgments for patients with unresectable pancreatic cancer treated by interstitial brachytherapy.

Feature selection in computer-aided breast cancer diagnosis via dynamic contrast-enhanced magnetic resonance images

The accuracy of computer-aided diagnosis (CAD) for early detection and classification of breast cancer in dynamic contrast-enhanced magnetic resonance imaging (DCE-MRI) is dependent upon the features used by the CAD classifier. Here, we show that fast orthogonal search (FOS), which provides a more efficient iterative manner of computing stepwise regression feature selection, can select features with predictive value from a set of kinetic and texture candidate features computed from dynamic contrast-enhanced magnetic resonance images. FOS can in minutes search candidate feature sets of millions of terms, which may include cross-products of features up to second-, third- or fourth-order. This method is tested on a set of 83 DCE-MRI images, of which 20 are for cancerous and 63 for benign cases, using a leave-one-out trial. The features selected by FOS were used in a FOS predictor and nearest-neighbour predictor and had an area under the receiver operating curve (AUC) of 0.889 and 0.791 respectively. The FOS predictor AUC is significantly improved over the signal enhancement ratio predictor with an AUC of 0.706 (p = 0.0035 for the difference in the AUCs). Moreover, using FOS-selected features in a support vector machine increased the AUC over that resulting when the features were manually selected.

Combined use of T2-weighted MRI and T1-weighted dynamic contrast-enhanced MRI in the automated analysis of breast lesions

A multiparametric computer-aided diagnosis scheme that combines information from T1-weighted dynamic contrast-enhanced (DCE)-MRI and T2-weighted MRI was investigated using a database of 110 malignant and 86 benign breast lesions. Automatic lesion segmentation was performed, and three categories of lesion features (geometric, T1-weighted DCE, and T2-weighted) were automatically extracted. Stepwise feature selection was performed considering only geometric features, only T1-weighted DCE features, only T2-weighted features, and all features. Features were merged with Bayesian artificial neural networks, and diagnostic performance was evaluated by ROC analysis. With leave-one-lesion-out cross-validation, an area under the ROC curve value of 0.77±0.03 was achieved with T2-weighted-only features, indicating high diagnostic value of information in T2-weighted images. Area under the ROC curve values of 0.79±0.03 and 0.80 ± 0.03 were obtained for geometric-only features and T1-weighted DCE-only features, respectively. When all features were considered, an area under the ROC curve value of 0.85±0.03 was achieved. We observed P values of 0.006, 0.023, and 0.0014 between the geometric-only, T1-weighted DCE-only, and T2-weighted-only features and all features conditions, respectively. When ranked, the P values satisfied the Holm-Bonferroni multiple-comparison test; thus, the improvement of multiparametric computer-aided diagnosis was statistically significant. A computer-aided diagnosis scheme that combines information from T1-weighted DCE and T2-weighted MRI may be advantageous over conventional T1-weighted DCE-MRI computer-aided diagnosis.

Differential diagnosis of pancreatic cancer from normal tissue with digital imaging processing and pattern recognition based on a support vector machine of EUS images

BACKGROUND

EUS can detect morphologic abnormalities of pancreatic cancer with high sensitivity but with limited specificity.

OBJECTIVE

To develop a classification model for differential diagnosis of pancreatic cancer by using a digital imaging processing (DIP) technique to analyze EUS images of the pancreas.

DESIGN

A retrospective, controlled, single-center design was used.

SETTING

The study took place at the Second Military Medical University, Shanghai, China.

PATIENTS

There were 153 pancreatic cancer and 63 noncancer patients in this study.

INTERVENTION

All patients underwent EUS-guided FNA and pathologic analysis.

MAIN OUTCOME MEASUREMENTS

EUS images were obtained and correlated with cytologic findings after FNA. Texture features were extracted from the region of interest, and multifractal dimension vectors were introduced in the feature selection to the frame of the M-band wavelet transform. The sequential forward selection process was used for a better combination of features. By using the area under the receiver operating characteristic curve and other texture features based on separability criteria, a predictive model was built, trained, and validated according to the support vector machine theory.

RESULTS

From 67 frequently used texture features, 20 better features were selected, resulting in a classification accuracy of 99.07% after being added to 9 other features. A predictive model was then built and trained. After 50 random tests, the average accuracy, sensitivity, specificity, positive predictive value, and negative predictive value for the diagnosis of pancreatic cancer were 97.98 ± 1.23%, 94.32 ± 0.03%, 99.45 ± 0.01%, 98.65 ± 0.02%, and 97.77 ± 0.01%, respectively.

LIMITATIONS

The limitations of this study include the small sample size and that the support vector machine was not performed in real time.

CONCLUSION

The classification of EUS images for differentiating pancreatic cancer from normal tissue by DIP is quite useful. Further refinements of such a model could increase the accuracy of EUS diagnosis of tumors.

Multispectral co-occurrence with three random variables in dynamic contrast enhanced magnetic resonance imaging of breast cancer

Presented is a new computer-aided multispectral image processing method which is used in three spatial dimensions and one spectral dimension where the dynamic, contrast enhanced magnetic resonance parameter maps derived from voxel-wise model-fitting represent the spectral dimension. The method is based on co-occurrence analysis using a 3-D window of observation which introduces an automated identification of suspicious lesions. The co-occurrence analysis defines 21 different statistical features, a subset of which were input to a neural network classifier where the assessments of the voxel-wise majority of a group of radiologist readings were used as the gold standard. The voxel-wise true positive fraction (TPF) and false positive fraction (FPF) results of the computer classifier were statistically indistinguishable from the TPF and FPF results of the readers using a one sample paired t-test. In order to observe the generality of the method, two different groups of studies were used with widely different image acquisition specifications.

Classification of dynamic contrast-enhanced magnetic resonance breast lesions by support vector machines

Early detection of breast cancer is one of the most important factors in determining prognosis for women with malignant tumors. Dynamic contrast-enhanced magnetic resonance imaging (DCE-MRI) has been shown to be the most sensitive modality for screening high-risk women. Computer-aided diagnosis (CAD) systems have the potential to assist radiologists in the early detection of cancer. A key component of the development of such a CAD system will be the selection of an appropriate classification function responsible for separating malignant and benign lesions. The purpose of this study is to evaluate the effects of variations in temporal feature vectors and kernel functions on the separation of malignant and benign DCE-MRI breast lesions by support vector machines (SVMs). We also propose and demonstrate a classifier visualization and evaluation technique. We show that SVMs provide an effective and flexible framework from which to base CAD techniques for breast MRI, and that the proposed classifier visualization technique has potential as a mechanism for the evaluation of classification solutions.

MR spectroscopy of the breast

This literature review assesses the clinical potential of proton ((1)H) magnetic resonance spectroscopy (MRS) of breast lesions. We here illustrate the basic principles of spectrum acquisition for volumes of interest, determined on the basis of dynamic magnetic resonance imaging (MRI) and of MRS postprocessing. We discuss the criteria for interpreting the spectrum with particular reference to the metabolic significance of the peak of total choline containing compounds at 3.2 ppm, a marker that is correlated with malignancy. We then summarise the findings obtained in lesion characterisation (with a possible gain in specificity with respect to dynamic MRI), the assessment of the effects of neoadjuvant chemotherapy and the correlation reported at high-field between the tumour tissue concentration of choline-containing compounds and the presence of lymph node metastases. Lastly, we outline the clinical use of this technique as the final phase of a complete breast MR examination after intravenous administration of paramagnetic contrast material for the dynamic study, with reference to its use by radiologists dedicated to breast imaging.

Classification of small contrast enhancing breast lesions in dynamic magnetic resonance imaging using a combination of morphological criteria and dynamic analysis based on unsupervised vector-quantization

PURPOSE

To evaluate the diagnostic value of breast magnetic resonance imaging (MRI) in small focal lesions using dynamic analysis based on unsupervised vector quantization in combination with a score for morphologic criteria.

MATERIALS AND METHODS

We examined 85 mammographically indetermintate lesions (BIRADS 3-4; 47 malignant, mean lesion size 1.2 cm; 38 benign, mean lesion size 1.1 cm). MRI was performed with a dynamic T1-weighted gradient echo sequence (1 precontrast and 5 postcontrast series). Lesions with an initial contrast enhancement >/=50% were selected with semiautomatic segmentation. For conventional dynamic analysis, we calculated the mean initial signal increase and postinitial course of all voxels included in a lesion. Secondly, all voxels within the lesions were assigned to 4 clusters using minimal-free-energy vector quantization. Dynamic and morphologic criteria were summarized in a diagnostic score and evaluated by receiver operating characteristic analysis.

RESULTS

In the present collection of small lesions, morphologic criteria [area under the curve (AUC) = 0.610] were inferior to dynamic criteria in the detection of breast cancer. Dynamic analysis with vector quantization (AUC = 0.760) presented slightly better results compared with standard dynamic analysis (AUC = 0.693). There was no benefit for combined morphologic and dynamic analysis.

CONCLUSION

In small MR-mammographic lesions, dynamic analysis with vector quantization alone tends to result in a higher diagnostic accuracy compared with combined morphologic and dynamic analysis.

High frame-rate simultaneous bilateral breast DCE-MRI

A simultaneous bilateral back-projection method for 3D dynamic contrast-enhanced (DCE)-MRI of the breasts was developed and evaluated. Using a double-side band modulation of the RF slab excitation pulse, discontinuous volumes that included both breasts were simultaneously selected. The number of slice phase-encoding steps was undersampled by a factor of 2, and the resulting signal aliasing from one volume to the other was removed using SENSE processing. In-plane encoding was performed with an interleaved radial acquisition reconstructed using dynamic k-space-weighted image contrast (KWIC) temporal filtering. Image resolution was 0.5 x 0.5 x 3.0 mm(3) with an effective temporal resolution of 15 s for both breast volumes. Combined with the 2x acceleration from SENSE encoding, this is a 16x acceleration factor over a conventional MR bilateral breast scan. An initial evaluation of these methods was performed on a cohort of women who presented with palpable or mammographically visible breast abnormalities. A total of 73 abnormalities were found in 45 of the 54 bilateral examinations that were performed. In 11 of these cases there was a significant finding in the contralateral breast. DCE images of both breasts can be acquired simultaneously, resulting in high-resolution images as well as rapid sampling of the contrast kinetics.

Technology transfer from the science of medicine to the real world: the potential role played by artificial adaptive systems

The author describes a refiguration of medical thought that originates from nonlinear dynamics and chaos theory. The coupling of computer science and these new theoretical bases coming from complex systems mathematics allows the creation of "intelligent" agents capable of adapting themselves dynamically to problems of high complexity: the artificial neural networks (ANNs). ANNs are able to reproduce the dynamic interaction of multiple factors simultaneously, allowing the study of complexity; they can also draw conclusions on an individual basis and not as average trends. These tools can allow a more efficient technology transfer from the science of medicine to the real world, overcoming many obstacles responsible for the present translational failure. They also contribute to a new holistic vision of the human subject person, contrasting the statistical reductionism that tends to squeeze or even delete the single subject, sacrificing him to his group of belongingness. A remarkable contribution to this individual approach comes from fuzzy logic, according to which there are no sharp limits between opposite things, such as wealth and disease. This approach allows one to partially escape from the probability theory trap in situations where it is fundamental to express a judgement based on a single case and favor a novel humanism directed to the management of the patient as an individual subject person.

Breast MR imaging in women at high-risk of breast cancer. Is something changing in early breast cancer detection?

In the last few years, several papers have addressed the introduction of contrast-enhanced MR imaging for screening women at high risk for breast cancer. Taking in consideration five prospective studies, on 3,571 screened women with hereditary predisposition to the disease and 9,652 rounds, we found that 168 patients were diagnosed with breast cancer (155 screen-detected, eight interval, and five cancers excluded from analysis) with a detection rate per year of 1.7%. These cancers were small (49% equal to or less than 10 mm in diameter) but aggressive, 82% being invasive and 49% with histologic grade 3; however, only 19% of these invasive cancers were associated with nodal involvement. The pooled sensitivity was 16% for clinical breast examination, 40% for mammography, 43% for ultrasound, and 81% for MR. The positive predictive value (calculated on the basis of the number of invasive diagnostic procedures due to false positives) was 33%, 47%, 18%, and 53%, respectively. Aim of the present article is to present the historical development of MR imaging of breast tumors that made this application theoretically and technically possible, to explain what strategic problems we face in the presence of a hereditary predisposition to the disease, to review the main results of the published studies, and to outline open problems and future perspectives.

Model-based region-of-interest selection in dynamic breast MRI

Magnetic resonance imaging (MRI) is emerging as a powerful tool for the diagnosis of breast abnormalities. Dynamic analysis of the temporal pattern of contrast uptake has been applied in differential diagnosis of benign and malignant lesions to improve specificity. Selecting a region of interest (ROI) is an almost universal step in the process of examining the contrast uptake characteristics of a breast lesion. We propose an ROI selection method that combines model-based clustering of the pixels with Bayesian morphology, a new statistical image segmentation method. We then investigate tools for subsequent analysis of signal intensity time course data in the selected region. Results on a database of 19 patients indicate that the method provides informative segmentations and good detection rates.

Breast lesion analysis of shape technique: semiautomated vs. manual morphological description

PURPOSE

To investigate the efficacy of an automated method of shape measurement for improving the discrimination of benign and malignant breast lesions.

MATERIALS AND METHODS

A total of 47 breast lesions (32 malignant and 15 benign) were examined using a 1.5 Tesla system. Regions of interest (ROIs) were manually drawn and extracted from high-resolution, fat-suppressed, postcontrast images, or were extracted with the use of a semiautomated computer algorithm. Shape parameters (i.e., complexity, convexity, circularity, and degree of elongation) were determined to assess whether they could be used to discriminate breast lesions.

RESULTS

Convexity differed significantly between the benign and malignant groups for both ROI methods. In addition, the semiautomated method demonstrated significantly different values of complexity.

CONCLUSION

This work demonstrates the usefulness of several shape descriptors for characterizing breast lesions, and shows that the automated method of analysis improves the discrimination and standardization of data.

The use of artificial neural networks in decision support in cancer: a systematic review

Artificial neural networks have featured in a wide range of medical journals, often with promising results. This paper reports on a systematic review that was conducted to assess the benefit of artificial neural networks (ANNs) as decision making tools in the field of cancer. The number of clinical trials (CTs) and randomised controlled trials (RCTs) involving the use of ANNs in diagnosis and prognosis increased from 1 to 38 in the last decade. However, out of 396 studies involving the use of ANNs in cancer, only 27 were either CTs or RCTs. Out of these trials, 21 showed an increase in benefit to healthcare provision and 6 did not. None of these studies however showed a decrease in benefit. This paper reviews the clinical fields where neural network methods figure most prominently, the main algorithms featured, methodologies for model selection and the need for rigorous evaluation of results.

Cluster analysis of signal-intensity time course in dynamic breast MRI: does unsupervised vector quantization help to evaluate small mammographic lesions?

We examined whether neural network clustering could support the characterization of diagnostically challenging breast lesions in dynamic magnetic resonance imaging (MRI). We examined 88 patients with 92 breast lesions (51 malignant, 41 benign). Lesions were detected by mammography and classified Breast Imaging and Reporting Data System (BIRADS) III (median diameter 14 mm). MRI was performed with a dynamic T1-weighted gradient echo sequence (one precontrast and five postcontrast series). Lesions with an initial contrast enhancement >or=50% were selected with semiautomatic segmentation. For conventional analysis, we calculated the mean initial signal increase and postinitial course of all voxels included in a lesion. Secondly, all voxels within the lesions were divided into four clusters using minimal-free-energy vector quantization (VQ). With conventional analysis, maximum accuracy in detecting breast cancer was 71%. With VQ, a maximum accuracy of 75% was observed. The slight improvement using VQ was mainly achieved by an increase of sensitivity, especially in invasive lobular carcinoma and ductal carcinoma in situ (DCIS). For lesion size, a high correlation between different observers was found (R(2) = 0.98). VQ slightly improved the discrimination between malignant and benign indeterminate lesions (BIRADS III) in comparison with a standard evaluation method.

Classification of Signal-Time Curves Obtained by Dynamic Magnetic Resonance Mammography

OBJECTIVE

This study compares the performance of quantitative methods for the characterization of signal-time curves acquired by dynamic contrast-enhanced magnetic resonance mammography from 253 females.

MATERIALS AND METHODS

Signal-time curves obtained from 105 parenchyma, 162 malignant, and 91 benign tissue regions were examined (243 lesions were histopathologically validated). A neural network, a nearest-neighbor, and a threshold classifier were applied to either the entire signal-time curve or pharmacokinetic and descriptive parameters calculated from the curves to differentiate between 2 (malignant or benign) or 3 tissue classes (malignant, benign, or parenchyma). The classifiers were tuned and evaluated according to their performance on 2 distinct subsets of the curves.

RESULTS

The accuracy determined for the neural network and the nearest-neighbor classifiers was nearly identical (approximately 80% in case of 3 tissue classes, and approximately 76% in case of the 2 classes). In contrast, the accuracy of the threshold classifier applied to the discrimination of 3 classes was low (65%).

CONCLUSION

Quantitative classifiers can support the radiologist in the diagnosis of breast lesions.

Artificial neural network modelling of megavoltage photon dose distributions

An artificial neural network (NN) has been used to model the two-dimensional dose distributions from a Varian 2100C linac. The network was trained using depth dose data for 6 and 10 MV x-rays, collected during the linac commissioning phase. During training, the number of iterations and hidden nodes was adjusted manually until acceptable agreement between measured and predicted data was obtained. In order to validate the network a subset of the data was set aside and not used for training. This enabled the performance of the network to be investigated in terms of generalization and accuracy, together with its ability to interpolate between different field sizes and positions in the beam. Finally, the network was used to generate data points over a 2D grid so that isodose distributions could be visualized. Good agreement was found between measured data and that produced by the trained neural network.

Combination of low and high resolution sequences in two orientations for dynamic contrast-enhanced MRI of the breast: more than a compromise

The purpose was to combine T1-weighted 3D gradient echo sequences at low and high spatial resolution (and short and longer acquisition time, respectively) in two orientations without compromising signal/time curve analysis and to evaluate the incremental value of assessing architectural features in high resolution images in dynamic contrast-enhanced MR mammography. T1-weighted 3D-FLASH sequences in a 1.5-T scanner (512 x 256 pixel matrix at high resolution; 256 x 128 pixels at low resolution sequences, 72 slices, 1.7-mm slice thickness, TR 8.8 ms, TE 4.5 ms, flip angle 25 degrees) were acquired in a special order during a single investigation. Three observers evaluated architectural features of 36 histopathologically proven lesions using high or low resolution images independently. Architectural features of each lesion were assessed by rating on two three-point scales. Kappa statistics verified the decrease of inter-observer variability. All observers improved assessment of architectural features regarding high resolution images in transversal and coronal orientation (observer A: eight positive, three negative corrections; B: 12/5; C: 16/4). Most positive corrections resulted from improved detection of morphologic criteria of malignancy. Mean inter-observer agreement significantly (P<0.05) increased from "slight" to "moderate" (mean weighted kappa increased from 0.185 to 0.422). This protocol at the charge of slightly enlarged time for measurement offers an elegant way to improve analysis of architectural features in MRM.

Artificial neural networks and artificial organisms can predict Alzheimer pathology in individual patients only on the basis of cognitive and functional status

Data from several studies have pointed out the existence of a strong correlation between Alzheimer's disease (AD) neuropathology and cognitive state. However, because of their highly complex and nonlinear relationship, it has been difficult to develop a predictive model for individual patient classification through traditional statistical approaches. When exposed to complex data sets, artificial neural networks (ANNs) can recognize patterns, learn the relationship of different variables, and address classification tasks. To predict the results of postmortem brain examinations, we applied ANNs to the Nun Study data set, a longitudinal epidemiological study, which includes annual cognitive and functional evaluation. One hundred seventeen subjects from the study participated in this analysis. We determined how demographic data and the cognitive and functional variables of each subject during the last year of her life could predict the presence of brain pathology expressed as Braak stages, neurofibrillary tangles (NFTs) and neuritic plaques (NPs) count in the neocortex and hippocampus, and brain atrophy. The result of this analysis was then compared with traditional statistical models. ANNs proved to be better predictors than Linear Discriminant Analysis in all experimentations (+ approximately 10% in overall accuracy), especially when assembled in Artificial Organisms (+ approximately 20% in overall accuracy). Demographic, cognitive, and clinical variables were better predictors of tangles count in the neocortex and in the hippocampus when compared to NPs count. These findings strengthen the hypothesis that neurofibrillary pathology may represent the major anatomic substrate of the cognitive impairment found in AD.