Charles E. Kahn

Construction of a Bayesian network for mammographic diagnosis of breast cancer

Bayesian networks use the techniques of probability theory to reason under uncertainty, and have become an important formalism for medical decision support systems. We describe the development and validation of a Bayesian network (MammoNet) to assist in mammographic diagnosis of breast cancer. MammoNet integrates five patient-history features, two physical findings, and 15 mammographic features extracted by experienced radiologists to determine the probability of malignancy. We outline the methods and issues in the systems design, implementation, and evaluation. Bayesian networks provide a potentially useful tool for mammographic decision support.

Toward Best Practices in Radiology Reporting

The goals and current efforts of the Radiological Society of North America Radiology Reporting Committee are described. The committees charter provides an opportunity to improve the organization, content, readability, and usefulness of the radiology report and to advance the efficiency and effectiveness of the reporting process.

GoldMiner: A Radiology Image Search Engine

OBJECTIVEnWe sought to create an Internet-based search engine to retrieve images from a large collection of figures published in peer-reviewed journals.nnnCONCLUSIONnThe GoldMiner search engine provides easy, rapid access to a large library of images and their associated text, and it is freely available for use on the Internet.

Probabilistic Computer Model Developed from Clinical Data in National Mammography Database Format to Classify Mammographic Findings

PURPOSEnTo determine whether a Bayesian network trained on a large database of patient demographic risk factors and radiologist-observed findings from consecutive clinical mammography examinations can exceed radiologist performance in the classification of mammographic findings as benign or malignant.nnnMATERIALS AND METHODSnThe institutional review board exempted this HIPAA-compliant retrospective study from requiring informed consent. Structured reports from 48 744 consecutive pooled screening and diagnostic mammography examinations in 18 269 patients from April 5, 1999 to February 9, 2004 were collected. Mammographic findings were matched with a state cancer registry, which served as the reference standard. By using 10-fold cross validation, the Bayesian network was tested and trained to estimate breast cancer risk by using demographic risk factors (age, family and personal history of breast cancer, and use of hormone replacement therapy) and mammographic findings recorded in the Breast Imaging Reporting and Data System lexicon. The performance of radiologists compared with the Bayesian network was evaluated by using area under the receiver operating characteristic curve (AUC), sensitivity, and specificity.nnnRESULTSnThe Bayesian network significantly exceeded the performance of interpreting radiologists in terms of AUC (0.960 vs 0.939, P = .002), sensitivity (90.0% vs 85.3%, P < .001), and specificity (93.0% vs 88.1%, P < .001).nnnCONCLUSIONnOn the basis of prospectively collected variables, the evaluated Bayesian network can predict the probability of breast cancer and exceed interpreting radiologist performance. Bayesian networks may help radiologists improve mammographic interpretation.

Informatics in Radiology: Comparison of Logistic Regression and Artificial Neural Network Models in Breast Cancer Risk Estimation

Computer models in medical diagnosis are being developed to help physicians differentiate between healthy patients and patients with disease. These models can aid in successful decision making by allowing calculation of disease likelihood on the basis of known patient characteristics and clinical test results. Two of the most frequently used computer models in clinical risk estimation are logistic regression and an artificial neural network. A study was conducted to review and compare these two models, elucidate the advantages and disadvantages of each, and provide criteria for model selection. The two models were used for estimation of breast cancer risk on the basis of mammographic descriptors and demographic risk factors. Although they demonstrated similar performance, the two models have unique characteristics-strengths as well as limitations-that must be considered and may prove complementary in contributing to improved clinical decision making.

A Logistic Regression Model Based on the National Mammography Database Format to Aid Breast Cancer Diagnosis

OBJECTIVEnThe purpose of our study was to create a breast cancer risk estimation model based on the descriptors of the National Mammography Database using logistic regression that can aid in decision making for the early detection of breast cancer.nnnMATERIALS AND METHODSnWe created two logistic regression models based on the mammography features and demographic data for 62,219 consecutive mammography records from 48,744 studies in 18,269 [corrected] patients reported using the Breast Imaging Reporting and Data System (BI-RADS) lexicon and the National Mammography Database format between April 5, 1999 and February 9, 2004. State cancer registry outcomes matched with our data served as the reference standard. The probability of cancer was the outcome in both models. Model 2 was built using all variables in Model 1 plus radiologists BI-RADS assessment categories. We used 10-fold cross-validation to train and test the model and to calculate the area under the receiver operating characteristic curves (A(z)) to measure the performance. Both models were compared with the radiologists BI-RADS assessments.nnnRESULTSnRadiologists achieved an A(z) value of 0.939 +/- 0.011. The A(z) was 0.927 +/- 0.015 for Model 1 and 0.963 +/- 0.009 for Model 2. At 90% specificity, the sensitivity of Model 2 (90%) was significantly better (p < 0.001) than that of radiologists (82%) and Model 1 (83%). At 85% sensitivity, the specificity of Model 2 (96%) was significantly better (p < 0.001) than that of radiologists (88%) and Model 1 (87%).nnnCONCLUSIONnOur logistic regression model can effectively discriminate between benign and malignant breast disease and can identify the most important features associated with breast cancer.

Actionable Findings and the Role of IT Support: Report of the ACR Actionable Reporting Work Group

The ACR formed the Actionable Reporting Work Group to address the potential role of IT in the communication of imaging findings, especially in cases that require nonroutine communication because of the urgency of the findings or their unexpected nature. These findings that require special communication with referring clinicians are classified as actionable findings. The work group defines 3 categories of actionable findings that require, respectively, communication and clinical decision within minutes (category 1), hours (category 2), or days (category 3). Although the work group does not believe that there can be definitive lists of such findings, it developed lists in each category that would apply in most general hospital settings. For each category, the work group discusses ways in which IT can assist interpreting radiologists in successfully communicating to the relevant clinicians to ensure optimal patient care. IT systems can also help document the communication and facilitate auditing of the documentation. The work group recommends that vendors develop platforms that can be customized on the basis of local preferences and needs. Whatever system is used, it should be highly reliable and fit seamlessly into radiologists workflow.

Breast cancer risk estimation with artificial neural networks revisited: discrimination and calibration.

Discriminating malignant breast lesions from benign ones and accurately predicting the risk of breast cancer for individual patients are crucial to successful clinical decisions. In the past, several artificial neural network (ANN) models have been developed for breast cancer‐risk prediction. All studies have reported discrimination performance, but not one has assessed calibration, which is an equivalently important measure for accurate risk prediction. In this study, the authors have evaluated whether an artificial neural network (ANN) trained on a large prospectively collected dataset of consecutive mammography findings can discriminate between benign and malignant disease and accurately predict the probability of breast cancer for individual patients.

From Guidelines to Practice: How Reporting Templates Promote the Use of Radiology Practice Guidelines

Radiology practice guidelines have been developed to help radiologists achieve quality and safety in their clinical practice. One means to promote the use of practice guidelines in radiology is through the wider use of reporting templates, also known as structured reporting. This article presents specific examples in which radiology reporting templates can promote adherence to guidelines, gather data for quality improvement efforts, and facilitate compliance with performance incentive programs.

BANTER: a Bayesian network tutoring shell

We present an educational tool for bringing the information contained in a Bayesian network to the end user in an easily intelligible form. The BANTER shell is designed to tutor users in evaluation of hypotheses and selection of optimal diagnostic procedures. BANTER can be used with any Bayesian network containing nodes that can be classified into hypotheses, observations, and diagnostic procedures. The system enables one to present various types of queries to the network, to test ones ability to select optimal diagnostic procedures, and the request explanations. We describe the systems capabilities by illustrating how it functions with two structurally different network models of real-world medical problems.