Janice C. Honeyman-Buck

Hexagonal wavelet processing of digital mammography

This paper introduces a novel approach for accomplishing mammographic feature analysis through overcomplete multiresolution representations. We show that efficient representations may be identified from digital mammograms and used to enhance features of importance to mammography within a continuum of scale-space. We present a method of contrast enhancement based on an overcomplete, non-separable multiscale representation: The hexagonal wavelet transform.

Comparison of Two Commercial CAD Systems for Digital Mammography

I n late 2006, an outpatient imaging center began its conversion to digital mammography (DM). Part of this transition was a plan to purchase software for computer-aided detection (CAD) to assist in analysis of the digital mammography images. The preparation to purchase one of two systems included a comparison of several specifications, including DICOM compatibility and the ability of the systems to fit well into the digital mammography program of the imaging center. A small study was designed to determine whether one of the two different commercially available systems being considered [R2 ImageChecker (version 8.3.17) and iCAD Second Look (version 7.2-H)] was superior to the other for assisting in interpretation of digital mammography images from screening mammograms obtained with the newly installed General Electric Senographe DS unit. Patient images included in this study were given retrospective approval from the Institutional Review Board with waiver of informed consent for publication of the study results.

How does observer training affect imaging performance in digital mammography

Simulated mass lesions, superimposed onto an anthropomorphic breast phantom, were x-rayed using a small field of view digital mammography system. Eight radiologists and four scientists viewed the phantom images on a display monitor in a darkened room. Five readers had prior experience of reading these type of images. Readers assessed the probability of a simulated mass being present in each ROI, with the resultant data used to plot the corresponding Receiver Operating Characteristic (ROC) curves, and determine the corresponding area under the ROC curve (Az). Readers viewed the same set of images five successive times in a single session, and the time taken to read each image was recorded. The average time to complete the study for all twelve observers was 24 minutes (71 seconds/image). Experienced readers were quicker than novices, and radiologists were quicker than the scientists. The average Az value for the twelve readers for this detection task was 0.842 +/- 0.037 with coefficient of variations for individual readers ranging from 2.1% to 7.7%. Differences in imaging performance between the radiologists and scientists were very small. Analysis of the trends in measured imaging performance for each reader viewing successive (repeat) images showed that there was no improvement in imaging performance with increasing experience.

Novel dual screen-dual film combination for mammography

A new dual screen-dual film mammography combination was constructed which made of two phosphor screens and two films loaded into a single x-ray cassette. The screens and films were combined so that a single emulsion film (Film #1, Kodak Min-R E film) was placed in direct contact with the phosphor side of the first screen (Screen #1). Screen #1 was made of the Kodak Min-R phosphor (34.0 mg/cm2 Gd2O2S:Tb) coated on a 4 mil transparent Mylar backing. A double emulsion film (Film #2, Kodak T-Mat G film) was sandwiched between Screen #1 backing and the phosphor side of the second screen (Screen #2). Screen #2 was a Kodak Insight ME screen that has a Gd2O2S:Tb coating weight of 43.1 mg/cm2 and a reflective coating between its phosphor layer and support layer. The relative sensitometric responses and spatial resolution properties of the two films were measured as a function of x-ray tube potential (kVp). A series of a contrast-detail phantom images was obtained by varying the x- ray exposure level at 28 kVp. An observer performance study was subsequently carried out to investigate the low contrast performance using the dual screen-dual film combination. The effective speeds of the two films differed by a factor of 2.12 to 2.67 between 25 to 30 kVp. Film #2 contrast was a factor of two or greater than Film #1 in the range where Film #1 optical density values were under 0.7. Measured MTF curves from digitized edge images showed that Film #1 MTF performance was comparable to a standard Kodak Min-R screen-Min-R E film combination. The limiting spatial resolution was found to be 19.5 lp/mm for Film #1 and approximately 11 lp/mm for Film #2. Observer performance studies showed that the threshold contrast in detecting small (less than 10 mm) breast lesions could be up to a factor of two lower on Film #2 images when regions of interest are underexposed on Film #1 images.

PACS quality control and automatic problem notifier

One side effect of installing a clinical PACS Is that users become dependent upon the technology and in some cases it can be very difficult to revert back to a film based system if components fail. The nature of system failures range from slow deterioration of function as seen in the loss of monitor luminance through sudden catastrophic loss of the entire PACS networks. This paper describes the quality control procedures in place at the University of Florida and the automatic notification system that alerts PACS personnel when a failure has happened or is anticipated. The goal is to recover from a failure with a minimum of downtime and no data loss. Routine quality control is practiced on all aspects of PACS, from acquisition, through network routing, through display, and including archiving. Whenever possible, the system components perform self and between platform checks for active processes, file system status, errors in log files, and system uptime. When an error is detected or a exception occurs, an automatic page is sent to a pager with a diagnostic code. Documentation on each code, trouble shooting procedures, and repairs are kept on an intranet server accessible only to people involved in maintaining the PACS. In addition to the automatic paging system for error conditions, acquisition is assured by an automatic fax report sent on a daily basis to all technologists acquiring PACS images to be used as a cross check that all studies are archived prior to being removed from the acquisition systems. Daily quality control is preformed to assure that studies can be moved from each acquisition and contrast adjustment. The results of selected quality control reports will be presented. The intranet documentation server will be described with the automatic pager system. Monitor quality control reports will be described and the cost of quality control will be quantified. As PACS is accepted as a clinical tool, the same standards of quality control must be established as are expected on other equipment used in the diagnostic process.

On-line medical record/RIS/PACS interface

The University of Florida Medical Center has developed an On-Line Medical Record (OLMR) that serves as a repository of patient information from a number of individual department databases, the Radiology Information System included, and builds a comprehensive electronic patient based chart. The OLMR, widely used by clinicians to view information on test results, will be expanded to add image and graphics display capabilities, and will require pointers to PACS images.

Redundant Publication-How to Avoid Duplication.

We all know that duplicate publication of an article is considered self-plagiarism and is not allowed in peerreviewed journals. The Journal of Digital Imaging has a statement in the submission process that a manuscript has not been submitted to any other journals for publication. This is a common statement that peer-reviewed journals often use. Accusations of duplicate publications have wide ranging consequences. For example, such an accusation may bar all the authors on a manuscript from future submissions to a journal, the author’s Department Chair or Dean may be informed, an internal investigation may be launched, the local newspaper might report on scientific misconduct by faculty members, and the repercussions may effect a person’s promotion, tenure, and reputation. How can an author get caught in the position of unknowingly submitting a manuscript containing material that could be considered a duplicate publication? Duplicate publication includes the text in an article, but it also includes figures and data sets previously published. If an author uses a figure in an article published in a blog, an abstract, another journal article, a teaching file, or published lecture notes, that figure may have a copyright associated with it or it at the very least it has been published. This figure could be a graph or drawing produced by the author or a radiology image. Once it has been published, it cannot be included in a future article without acknowledgement and for most peer review journals, the ability to assign the copyright to that figure to the journal accepting the manuscript for publication. If the author uses a dataset for an article, that dataset has been published. Different parts of the dataset can be used for subsequent articles but not the prior published dataset. When an author wishes to present research at a scientific meeting, it is common to submit an abstract to the organization holding the meeting and if accepted, that abstract could be published by the organization either in a proceedings format or online. Often, the author assigns the copyright to the organization publishing the abstract. Signing copyright forms is part of the publication process and most of us sign them without much thought about the future consequences. But suppose the author who presented the paper at a scientific meeting went on to produce a manuscript and included figures and text from the original abstract. That is duplicate publication. Many manuscripts are placed in Open Access Institutional Repositories. There were 65 institutional repositories in the USA reported in August 2015. This list includes 36 major medical centers and 29 colleges and universities and more repositories are planned. Most, possibly all of the authors submitting the JDI are unaware if a repository exists and even if they are aware of the repository, they are not aware of the policies and restrictions. Many institutions allow the authors to maintain the copyright to their work. In this case, authors are allowed to post the article on other open access sites and on their personal website. Journals should seek out and publish work that has not been previously published in print, web, or other electronic publications. Redundant publication occurs when multiple papers are written without reference in the text, and share the same text, data or results. In order to help identify potential plagiarism of any kind journal editors, including myself, use a tool called iThenticate to search for similarities in any form of publication on the web. Because of the availability of these tools, there is a * Janice Honeyman-Buck honeymanbuck@aol.com

References and reading lists for imaging informatics professionals: preparing for certification.

As most picture archiving and communication system (PACS) administrators or imaging informatics professionals are aware, the Society for Imaging Informatics in Medicine (SIIM, formerly SCAR) is sponsoring an effort to provide certification in the growing field. SIIM members believe that this society is the most qualified to provide guidance in the development of an exam and certification because we are an unbiased nonprofit organization. We are working with the American Registry of Radiologic Technologists (ARRT) to help develop and administer the test, not because we believe a Certified Imaging Informatics Professional should have a background in radiologic technology but because the ARRT has vast experience with developing and administering examinations. Many of the people who intend to take the examination have asked how to prepare for it. Of course, attending professional meetings and PACS Administration courses, including those offered by SIIM, are a good way to meet others in the profession as well as to learn new ways to solve problems. We have attempted to put together a reading list for those who also wish to prepare for the examination by reading and studying at home.

ROC comparison between digital mammography and screen-film using an anthropomorphic breast phantom

Mass lesion detection performance of a LoRAD Digital Spot Mammography (DSM) system was compared with a Kodak Min R screen-film combination exposed either in front of the DSM, or in the Bucky of a GE 600T mammography unit. Low-contrast objects simulating small masses were superimposed on an RMI 165 anthropomorphic breast phantom and radiographs obtained at 28 kVp and an mAs value, which resulted in a mean film density of approximately 1.1. DSM images were obtained at the same radiation exposure as used with screen-film. Fully masked radiographs were viewed on a mammography light box, and the DSM images were viewed on the DSM monitor in a darkened room. Of the 64 regions of interest (ROI) in each type of image, 28 (44%) contained the test object. For each imaging modality, six radiologists and six scientists assessed the probability of a simulated mass being present in each ROI. The resultant data were used to plot receiver operating characteristic (ROC) curves of twelve readers for each of the three imaging modalities investigated. There was no significant difference in reader performance between the screen-film combination exposed in front of the DSM system and exposed in the GE 600T system. Both screen-film imaging systems resulted in the same average area under the ROC curve, Az, of 0.78. At the same level of radiation exposure, the DSM had an average ROC area, Az, of 0.71 which was significantly inferior to the average performance achieved using screen-film (p less than 0.005). For this detection task, there were no significant differences in performance between the radiologists and scientists. Reader performance was found to improve with the number of images read, demonstrating an observer learning curve for this specific detection task.

Comparison of a dyadic wavelet image enhancement algorithm with unsharp masking and median filtering for mammography

Image processing techniques using wavelet signal analysis techniques have shown promise in mammography. Wavelet algorithms are compared with traditional image enhancement techniques of unsharp masking and median filtering. Computer simulated phantom images were generated containing lesions mimicking masses and microcalcifications. The degree of image enhancement was evaluated by comparing processed and original signal-to-noise (SNR) ratios in such phantom images. Results obtained in this study suggest that image processing algorithms based on the wavelet transform are likely to enhance the visibility of low-contrast features in mammograms.