OBJECTIVE: To investigate the utilization of text mining technique for evaluating the informational quality of electronic mammographic reports considering adherence to the BI-RADS® lexicon as a quality parameter. MATERIALS AND METHODS: A total of 22,247 mammography reports of the period between January, 2000 and June, 2006 were collected from the radiology information database of Hospital das Clínicas da Faculdade de Medicina de Ribeirão Preto, SP, Brazil. Two experiments were undertaken - experiment 1 to evaluate the accuracy in the adoption of the lexicon terms (text mining method specificity), and experiment 2 to identify all and any attempt to utilize or refer to the lexicon (text mining method sensitivity). RESULTS: Experiment 1: variation between 11% and 61% in reports including lexicon terms in their conclusion, randomly distributed over time since 2001. Experiment 2: variation between 44% and 100% in reports that somehow refer to the lexicon in their conclusion. CONCLUSION: Results indicate a good potential for text mining tool application for assessing the quality of information included in electronic mammography reports.

Keywords: Mammography, BI-RADS, Information theory, Medical information technology

OBJECTIVE: The present paper describes the implementation and evaluation of a medical images management system with content-based retrieval support (PACS-CBIR) integrating modules focused on images acquisition, storage and distribution, and text retrieval by keyword and images retrieval by similarity. MATERIALS AND METHODS: Internet-compatible technologies were utilized for the system implementation with freeware, and C++, PHP and Java languages on a Linux platform. There is a DICOM-compatible image management module and two query modules, one of them based on text and the other on similarity of image texture attributes. RESULTS: Results demonstrate an appropriate images management and storage, and that the images retrieval time, always < 15 sec, was found to be good by users. The evaluation of retrieval by similarity has demonstrated that the selected images extractor allowed the sorting of images according to anatomical areas. CONCLUSION: Based on these results, one can conclude that the PACS-CBIR implementation is feasible. The system has demonstrated to be DICOM-compatible, and that it can be integrated with the local information system. The similar images retrieval functionality can be enhanced by the introduction of further descriptors.

Keywords: CBIR, PACS, DICOM, Medical images, Health information system

OBJECTIVE: To develop an atlas of myocardial perfusion scintigraphy and evaluating its applicability in computer-aided detection of myocardial perfusion defects in patients with ischemic heart disease. MATERIALS AND METHODS: The atlas was created with rest-stress myocardial perfusion scintigraphic images of 20 patients of both genders with low probability of coronary artery disease and considered as normal by two experienced observers. Techniques of image registration and mathematical operations on images were utilized for obtaining template images depicting mean myocardial uptake and standard deviation for each gender and physiological condition. RESULTS: Myocardial perfusion scintigraphy images of one male and one female patient were aligned with the corresponding atlas template image, and voxels with myocardial uptake rates two standard deviations below the mean voxel value of the respective region in the atlas template image were highlighted on the tomographic sections and confirmed as perfusion defects by both observers. CONCLUSION: The present study demonstrated the creation of an atlas of myocardial perfusion scintigraphy with promising results of this tool as an aid in the detection of myocardial perfusion defects. However, further prospective validation with a more representative sample is recommended.

Keywords: Scintigraphy, SPECT, Myocardium, Nuclear medicine, Computer-assisted image processing, Computer-assisted diagnosis

OBJECTIVE: To utilize the grid computing technology to enable the utilization of a similarity measurement algorithm for content-based medical image retrieval. MATERIALS AND METHODS: The content-based images retrieval technique is comprised of two sequential steps: texture analysis and similarity measurement algorithm. These steps have been adopted for head and knee images for evaluation of accuracy in the retrieval of images of a single plane and acquisition sequence in a databank with 2,400 medical images. Initially, texture analysis was utilized as a preselection resource to obtain a set of the 1,000 most similar images as compared with a reference image selected by a clinician. Then, these 1,000 images were processed utilizing a similarity measurement algorithm on a computational grid. RESULTS: The texture analysis has demonstrated low accuracy for sagittal knee images (0.54) and axial head images (0.40). Nevertheless, this technique has shown effectiveness as a filter, pre-selecting images to be evaluated by the similarity measurement algorithm. Content-based images retrieval with similarity measurement algorithm applied on these pre-selected images has demonstrated satisfactory accuracy - 0.95 for sagittal knee images, and 0.92 for axial head images. The high computational cost of the similarity measurement algorithm was balanced by the utilization of grid computing. CONCLUSION: The approach combining texture analysis and similarity measurement algorithm for content-based images retrieval resulted in an accuracy of > 90%. Grid computing has shown to be essential for the utilization of similarity measurement algorithm in the content-based images retrieval that otherwise would be limited to supercomputers.

Keywords: Content-based image retrieval, Texture analysis, Image registration, Grid computing

OBJECTIVE: Integration of the Radiology Information System (RIS) and the Picture Archiving and Communication System (PACS) of the Radiodiagnosis Service of "Hospital das Clínicas de Ribeirão Preto da Universidade de São Paulo", Brazil, to allow remote access to reports and associated images. MATERIALS AND METHODS: RIS/PACS integration occurs in real time at the moment of consultation, using web technologies and programming techniques for intranet/internet. RESULTS: This web application allows reports and associated images viewing through the medical center intranet. Search is based on the patients' first name, last name, hospital number identification and exam modality, within a time period. The viewer allows image navigation and presents some basic functions such as zoom, brightness/contrast control and side-by-side image visualization. CONCLUSION: RIS/PACS integration minimizes the risk of inconsistencies, through the reduction of the number of interfaces among data bases with great information redundancy, providing a fast and safe work environment for radiology reports, consultation and visualization of associated images.

Keywords: PACS, RIS, Filmless radiology, Web technology

OBJECTIVE: To characterize interstitial lesions in anterior-posterior chest X-rays based on the analysis of textural statistical features that allow the detection of abnormalities with diffuse pattern. MATERIALS AND METHODS: Image analysis begins with the semiautomatic segmentation of the lungs, marking the external contour of the lung manually followed by an automatic division of each lung in six regions. The data base of images used in this study consisted of 482 regions obtained from examinations in which lesions were detected and 324 regions from normal examinations. Textural features were automatically extracted from each area and the selection of the best set of features was made based on the Jeffries-Matusita distance. The regions were classified as normal or suspected using the k nearest-neighbor method and half-half, and cross-correlation methodologies were used for training the classifier. RESULTS: Results were assessed based on the value of the area under the ROC (receiver operating characteristic) curve that indicates an ideal response for an area equal to 1. The results showed an area under the ROC curve (A Z) of 0.887, sensitivity of 0.804, and specificity of 0.793. CONCLUSION: These results indicate that the implemented system has a good potential for computer-aided diagnosis of interstitial lung lesions.

Keywords: Computer-aided diagnosis, Interstitial lung lesions, Texture characteristics, Pattern recognition

OBJECTIVE: To present the process of implementation of toolkits for visualization of medical images using DICOM and to review the basis and characteristics of this protocol. In this paper we present a VDTApplication for visualization of local and remote images, and a VDTApplet that allows the visualization of images on a web browser. MATERIALS AND METHODS: The toolkits were implemented using Java programming language. For the development of this toolkit we considered the different version of the DICOM protocol, creating a generic system capable of reading and opening images generated by different modalities of different manufacturers. The tools were evaluated qualitatively by doctors taking into account the system's interface, image quality and image adjustment. RESULTS: The evaluation score ranged from 1 (very poor ) to 5 (very good) for each item, with the following results: 5 for system interface, 4 for image quality and 3 for the image adjustment. Based on the scores obtained we concluded that the tools can be considered good. CONCLUSION: These results allowed us to conclude that these tools are practical, user friendly and should be added to the PACS (picture archiving and communication system) of "Hospital das Clínicas da Faculdade de Medicina de Ribeirão Preto - Universidade de São Paulo", Ribeirão Preto, SP, Brazil.

Keywords: Medical images, DICOM, Filmless radiology, Medical images visualization

This work describes a content-based image retrieval system developed to aid in the diagnosis of breast lesions by visual inspection through comparison of images. When an unknown image is presented the system extracts a vector of texture features and searches a database for images with similar characteristics, within a previously established approximation. The retrieved images are presented to the user for evaluation of the associated diagnoses.

Keywords: Content-based image retrieval, Pattern recognition, Texture features

Several technological developments are contributing to increase the use of imaging techniques in biomedical research and clinical medicine. Many investigators have attempted to develop computerized routines for automated detection and quantitative analysis of abnormalities in radiological images. These routines have been referred to as computer-aided diagnosis. In this paper we discuss the basic concepts of computer-aided diagnosis and review the literature.

Keywords: Computer-aided diagnosis, Digital radiography, Image processing, Computer vision, Artificial intelligence

This paper describes the implementation of a mini-PACS (picture archiving and communication system) at a university hospital ("Hospital das Clínicas da Faculdade de Medicina de Ribeirão Preto da Universidade de São Paulo"), as a component of a project for a filmless radiology facility.

Keywords: Digital radiology. Picture archiving and communication system. PACS. Filmless radiology.

OBJECTIVE: To study the feasibility of reducing radiation dose in protocols for acquisition of helical computed tomography images in a University Hospital. MATERIALS AND METHODS: A survey of radiation doses in computed tomography protocols was performed with phantoms and ionization chamber. Changes in kVp and mAs were proposed, determining the average noise. Protocols with noise values < 1% were submitted to qualitative assessment of contrast and spatial resolution by three observers. RESULTS: Tests of variations were performed with 22 protocols for pediatric skulls, 26 for adult skulls, 28 for abdomen, and 18 for chest. The reduction in dose achieved ranged between 7.4% and13% for pediatric skull, 3.8% and 25% for adult skull, 9.6% and 34.3% for abdomen, 6.4% and 12% for chest. It was also noted that the use of windowing and zoom tools supported the acceptance of images by the observers. CONCLUSION: Radiation dose levels can be reduced by up to 34.4% in comparison with routine protocols, keeping the noise at acceptable levels. The use of digital manipulation tools allowed the acceptance of images with higher noise levels, thus resulting in radiation dose reduction.

Keywords: Computed tomography; Radiation dose reduction; ALARA; Signal-to-noise ratio; CTDI; Optimization.

Objective: To analyze the Brazilian technological innovation in the field of radiology and imaging diagnosis, in terms of patent indicators. Materials and Methods: Exploratory analytical study of information retrieved through cross queries in databases of intellectual property offices in Brazil (State Industrial Property Office [Instituto Nacional da Propriedade Industrial – INPI]), United States (United States Patent and Trademark Office – USPTO) and Europe (European Patent Office – EPO). Results: There were 277,057 patents in the field of radiology and imaging diagnosis. Of this total, 7,800 were registered at INPI (3%), 65,428 (24%) registered at the EPO and 203,829 (73%) issued by USPTO. Brazil is a signatory to 1,732 patents published by the INPI, 80 by the EPO and 26 by the USPTO. Overall, 219,993 (79%) patents were related to electronic devices strongly linked to information technology in health care, ultrasonography, tomography, magnetic resonance imaging procedures as well as to images generation, communication and archiving, and 57,064 patents (21%) dealt with technologies related to radiation protection and dosimetry, nuclear physics, electrotherapy, magnetic therapy and radiotherapy. Conclusion: The results indicate the weakness of the domestic production of patented technological innovation in the field of radiology and imaging diagnosis.

Keywords: Medical education; Science and technology; Technological innovation; Intellectual property; Research.

Objective: This paper presents a model for integration of computer-aided diagnosis algorithms into the picture archiving and communication systems workflow that has been developed on the basis of the dcm4che2 open source toolkit. Materials and Methods: The proposed integration model consists of an image processing server and communication services. The data management follows the workflow defined by the post-processing workflow profile (PWF) developed by Integrating the Healthcare Enterprise (IHE) and utilizes the DICOM secondary capture functionality. An application for diffuse lung disease has been utilized as proof of concept. Results: Based on a cross validation method, the standard classification algorithm presented 78% accuracy. The integration enables the visualization of processed images as a new series in the original study. Conclusion: The proposed integration model is based on IHE profiles and allows the establishment of standardized procedures. The principles used to integrate the workflow are applicable to any non-interactive post-processing task.

Keywords: Computer-aided image processing; Picture archiving and communication systems; DICOM secondary capture; IHE.

OBJECTIVE: To construct a Portuguese language index of information on the practice of diagnostic radiology in order to improve the standardization of the medical language and terminology. MATERIALS AND METHODS: A total of 61,461 definitive reports were collected from the database of the Radiology Information System at Hospital das Clínicas - Faculdade de Medicina de Ribeirão Preto (RIS/HCFMRP) as follows: 30,000 chest x-ray reports; 27,000 mammography reports; and 4,461 thyroid ultrasonography reports. The text mining technique was applied for the selection of terms, and the ANSI/NISO Z39.19-2005 standard was utilized to construct the index based on a thesaurus structure. The system was created in ^*html. RESULTS: The text mining resulted in a set of 358,236 (n = 100%) words. Out of this total, 76,347 (n = 21%) terms were selected to form the index. Such terms refer to anatomical pathology description, imaging techniques, equipment, type of study and some other composite terms. The index system was developed with 78,538 *html web pages. CONCLUSION: The utilization of text mining on a radiological reports database has allowed the construction of a lexical system in Portuguese language consistent with the clinical practice in Radiology.

Keywords: Radiology reports; RadLex; Medical education; Information technology; Health informatics.

OBJECTIVE: Several studies have been published regarding the use of bismuth shielding to protect the breast in computed tomography (CT) scans and, up to the writing of this article, only one publication about barium shielding was found. The present study was aimed at characterizing, for the first time, a lead breast shielding. MATERIALS AND METHODS: The percentage dose reduction and the influence of the shielding on quantitative imaging parameters were evaluated. Dose measurements were made on a CT equipment with the aid of specific phantoms and radiation detectors. A processing software assisted in the qualitative analysis evaluating variations in average CT number and noise on images. RESULTS: The authors observed a reduction in entrance dose by 30% and in CTDI_vol by 17%. In all measurements, in agreement with studies in the literature, the utilization of cotton fiber as spacer object reduced significantly the presence of artifacts on the images. All the measurements demonstrated increase in the average CT number and noise on the images with the presence of the shielding. CONCLUSION: As expected, the data observed with the use of lead shielding were of the same order as those found in the literature about bismuth shielding.

Keywords: Computed tomography; Shielding; Bismuth; Dose reduction; Lead.

The discipline of radiology and diagnostic imaging has evolved greatly in recent years. We have observed an exponential increase in the number of exams performed, subspecialization of medical fields, and increases in accuracy of the various imaging methods, making it a challenge for the radiologist to “know everything about all exams and regions”. In addition, imaging exams are no longer only qualitative and diagnostic, providing now quantitative information on disease severity, as well as identifying biomarkers of prognosis and treatment response. In view of this, computer-aided diagnosis systems have been developed with the objective of complementing diagnostic imaging and helping the therapeutic decision-making process. With the advent of artificial intelligence, “big data”, and machine learning, we are moving toward the rapid expansion of the use of these tools in daily life of physicians, making each patient unique, as well as leading radiology toward the concept of multidisciplinary approach and precision medicine. In this article, we will present the main aspects of the computational tools currently available for analysis of images and the principles of such analysis, together with the main terms and concepts involved, as well as examining the impact that the development of artificial intelligence has had on radiology and diagnostic imaging.

Keywords: Artificial intelligence; Machine learning; Computer aided diagnosis; Radiomics.

OBJECTIVE: To determine whether the radiomic features of lung lesions on computed tomography correlate with overall survival in lung cancer patients. MATERIALS AND METHODS: This was a retrospective study involving 101 consecutive patients with malignant neoplasms confirmed by biopsy or surgery. On computed tomography images, the lesions were submitted to semi-automated segmentation and were characterized on the basis of 2,465 radiomic variables. The prognostic assessment was based on Kaplan–Meier analysis and log-rank tests, according to the median value of the radiomic variables. RESULTS: Of the 101 patients evaluated, 28 died (16 dying from lung cancer), and 73 were censored, with a mean overall survival time of 1,819.4 days (95% confidence interval [95% CI]: 1,481.2–2,157.5). One radiomic feature (the mean of the Fourier transform) presented a difference on Kaplan–Meier curves (p < 0.05). A high-risk group of patients was identified on the basis of high values for the mean of the Fourier transform. In that group, the mean survival time was 1,465.4 days (95% CI: 985.2–1,945.6), with a hazard ratio of 2.12 (95% CI: 1.01–4.48). We also identified a low-risk group, in which the mean of the Fourier transform was low (mean survival time of 2,164.8 days; 95% CI: 1,745.4–2,584.1). CONCLUSION: A radiomic signature based on the Fourier transform correlates with overall survival, representing a prognostic biomarker for risk stratification in patients with lung cancer.

Keywords: Tomography, X-ray computed; Radiographic image interpretation, computer-assisted; Lung neoplasms; Prognosis.

OBJECTIVE: To evaluate the degree of similarity between manual and semiautomatic segmentation of soft-tissue sarcomas on magnetic resonance imaging (MRI). MATERIALS AND METHODS: This was a retrospective study of 15 MRI examinations of patients with histopathologically confirmed soft-tissue sarcomas acquired before therapeutic intervention. Manual and semiautomatic segmentations were performed by three radiologists, working independently, using the software 3D Slicer. The Dice similarity coefficient (DSC) and the Hausdorf distance were calculated in order to evaluate the similarity between manual and semiautomatic segmentation. To compare the two modalities in terms of the tumor volumes obtained, we also calculated descriptive statistics and intraclass correlation coef ficients (ICCs). RESULTS: In the comparison between manual and semiautomatic segmentation, the DSC values ranged from 0.871 to 0.973. The comparison of the volumes segmented by the two modalities resulted in ICCs between 0.9927 and 0.9990. The DSC values ranged from 0.849 to 0.979 for intraobserver variability and from 0.741 to 0.972 for interobserver variability. There was no significant difference between the semiautomatic and manual modalities in terms of the segmentation times (p > 0.05). CONCLUSION: There appears to be a high degree of similarity between manual and semiautomatic segmentation, with no significant difference between the two modalities in terms of the time required for segmentation.

Keywords: Image processing, computer-assisted; Sarcoma/diagnostic imaging; Soft tissue neoplasms/diagnostic imaging; Magnetic resonance imaging; Reproducibility of results.