It’s October 1^st and the beginning of Breast Cancer Awareness Month.  I am reminded that both my mother and grandmother had breast cancer, two close friends have died from breast cancer, and my daughter’s best friend was diagnosed at the age of 34.   When my sister and I investigated BRCA genetic testing, we were discouraged because if we had a genetic risk of getting cancer ourselves, and our health insurance companies found out, they may drop our coverage. 

We have five  JDI online-first articles with Breast Imaging topics I want to draw to your attention.  The first one is entitled “Effect of Dose Reduction on the Ability of Digital Mammography to Detect Simulated Microcalcifications”  by Yakabe, Sakai, Yabuuchi, Matsuo, Kamitani, Setoguchi, Cho, Masuda and Sasaki.  In their article, their research suggests that a certain level of dose reduction in digital mammography may be an option.  The second article is entitled “A Statistical Approach for Breast Density Segmentation” by Oliver,  Llado, Perez, Pont, Denton, Freixenet, and Marti.   Their research centers around evaluating the density of a breast by segmenting its internal parenchyma in either fatty or dense  surrounding tissue.  The third article is entitled “A New Fast Fractal Modeling Approach for the Detection of Microcalcifications in Mammograms” by Sankar and Thomas.  As the title suggests, they describe their fast method for modeling mammograms by using deterministic fractal coding to enhance microcalcifications.    The fourth article is entitled “Effect of Pixel Resolution on Texture Features of Breast Masses in Mammograms” by Rangayyan, Nguyen, ayres, and Nandi.  This group analyzed breast masses at various pixel sizes to discriminate mammographic breast lesions as benign masses or malignant tumors.   And last but not least, the fifth article is entitled “Validation of Results from Knowledge Discovery: Mass Density as a Predictor of Breast Cancer” by Woods, Oliphant, Shinki, Page, Shavlik, and Burnside.   The purpose of the study was to identify and quantify the association between high breast mass density and breast malignancy using inductive logic programming and conditional probabilities.  Their results show that both measures indicate that mass density is an important adjunct predictor of malignancy.  This article is also open access, provided by SIIM for articles deemed to be of high interest to the SIIM community.

All women readers should consider joining the Army of Women, www.armyofwomen.org.  This organization has one research goal – to prevent breast cancer.  They need women without breast cancer as well as those who are newly diagnosed or who are survivers.

We have several new online first articles for everyone interested in medical imaging informatics and the technologies used or created for it.  Be sure to visit Springerlink for new content, it is updated almost daily.

From time to time, I’ll write about a new article that catches my eye.  Today’s is from Steve Langer, a pioneer and leader in our field, who writes about “DCMTB: A Virtual Appliance DICOM Toolbox”.   His group uses many of the free and/or open source tools developed recently for toolkits to help in imaging informatics such as DCM4CHEE, MIRTH, XNAT, PostGres SQL, programming environments and a group of Diagnostic Medical Physics tools.  The problem was that multiple team members could not use the same instance of a toolkit for fear of overwriting another’s work and the installation of these toolkits is not a trivial task.  Their solution is using virtual machines to build a configurable setting that gives each team member the “ownership” of a set of tools as needed and supports their multi-project environment.  This is interesting reading, don’t miss it.

Three new articles have been added to the JDI SpringerLink Online-First site in addition to one of the articles described in my last post. They will be of interest to researchers who are working on methods of improving interpretation for radiologists.

PET/CT Fusion Viewing Software for Use with Picture Archiving and Communication Systems by Ki Chun Im, Ik Soo Choi, Jin-Sook Ryu, Gi Seoung Eo, Jae Seung Kim and Dae Hyuk Moon.

This is a description of PET-viewer, software developed by the authors to display co-registered PET and CT images.  Reconstructed datasets can be archived in the PACS. 

Assessing the Accuracy Factors in the Determination of Postoperative Acetabular Cup Orientation Using Hybrid 2D–3D Registration by Guoyan Zheng

The program developed by this author, called HipMatch, has been described previously.  This article describes the process of assessing accuracy of the program using a cadaver pelvis which was imaged in different positions.  Since it is sometimes difficult to precisely position a patient, this tool promises to be valuable in routine clinical use.

A New Fast Fractal Modeling Approach for the Detection of Microcalcifications in Mammograms by Deepa Sankar and Tessamma Thomas.

This article investigates a new way to use fractal modeling to help detect microcalcifications in mammograms.  Because modeling using fractal encoding takes a tremendous amount of time, the authors used mean and variance, the dynamic range of the image blocks, and mass center features to reduce the amount of time needed for encoding.  The authors demonstrate significant reduction in time and may provide insight to others performing this type of modeling.

If you are interested in compression, and who isn’t in these days of huge data acquisition, we have many JDI compression articles ready to be read on Springerlink online-first. These have been officially published, but not yet in the print copy – we have a huge backlog of articles to get to print. I’ll report on two of the compression studies today. To get to Springerlink, I always start at the SIIM website: www.siimweb.org and then follow the links to publications and JDI. It’s a good idea to check out Springerlink every week or two for new publications, but I’ll suggest a couple each week to get everyone interested.

My first article is titled: “A Multicenter Observer Performance Study of 3D JPEG2000 Compression of Thin-Slice CT” by Bradley J. Erickson, Elizabeth Krupinski and Katherine P. Andriole
The goal of this multi-center study was to determine the compression level at which 3D JPEG 2000 compression if thin slice CTs of the chest and abdomen-pelvis becomes visually perceptible. The results show that even mild compression is perceptible with current technology but does not evaluate the changes in diagnostic accuracy.This article is online first and open access on Springer.

The second article has similar goals and is titled: “Pan-Canadian Evaluation of Irreversible Compression Ratios (“Lossy” Compression) for Development of National Guidelines” by David Koff, Peter Bak, Paul Brownrigg, Danoush Hosseinzadeh, April Khademi, Alex Kiss, Luigi Lepanto, Tracy Michalak, Harry Shulman and Andrew Volkening
This study from Canada evaluates lossy compression and recommends the level at which lossy compression can be confidently used in diagnostic imaging applications. The authors tested JPEG and JPEG-2000 DICOM supported compression algorithms. Their results showed that at low levels of compression, there was no significant difference between the performance of lossy JPEG and lossy JPEG 2000 and that both could be used without effecting diagnostic accuracy. At higher levels of compression, lossy JPEG proved to be more effective than JPEG 2000 in some cases. They provide a table of recommended compression ratios for each modality and anatomical area investigated.
This article is online first on Springer.