The following scientific studies have been published using the Senographe 2000D:

• Clinical Accuracy
• Productivity
• Physics
• Dose
• System Design

Clinical

• View the Abstract

Full-field digital mammography
Preliminary discussion of digital mammography systems compared to the conventional film-screen mammography systems. "The main advantage of digital mammography is the linear relationship between dose and detector signal with the possibility of a tailored optimization of image contrast. Other advantages include digital image storage, telemammography, and computer-assisted diagnosis."
U Bick, La Charité, Germany
German Fortschr Röntgenstr (RöFo), Dec 2000; vol. 172(12): 957-964.

• View the Abstract

Low-dose Tangential Visualization of Skin Calcifications by Digital Full-field Mammography
The study demonstrated skin calcifications on tangential views obtained by dose-reduced FFDM. The results proved that it was possible to prove the intracutaneous location of the calcifications with the radiation exposure of just one additional view. Thus, FFDM allowed to identify skin calcifications by tangential views obtained at a markedly reduced dose.
F Diekmann, SD Grebe, U Bick, BK Hamm III, La Charité, Germany
Abstract #9, presented at RSNA 2000.

• View the Abstract

Full-field digital mammography: soft-copy versus hard-copy
The study compared soft copy and hard copy readings of FFDM in terms of diagnostic accuracy. The soft copy reading was found to be equivalent to hard copy analysis. However, the advantage of the soft copy reading was that it allowed users to modify parameters in case of unclear findings.
M Funke, University of Göttingen, Germany
Göttingen Mammography Symposium, May 2000.

Full-field digital mammography: Is soft copy reviewing possible?
The study compared soft copy and hard copy readings of FFDM in terms of diagnostic accuracy. The soft copy reading was found to be equivalent to hard copy analysis. However, the advantage of the soft copy reading was that it allowed users to modify parameters in case of unclear findings.
M Funke, KP Hermann, S Obenauer, U Fischer, E Grabbe, University of Göttingen, Germany
Presented at the German Radiology Congress 2000, Wiesbaden, abstract in Fortschr Röntgenstr (RöFo), vol. 172, 2000, S42.

Interim clinical evaluation of FFDM in a screening cohort: comparison with screen-film mammography in 4,965 exams
The study compared FFDM to SFM for cancer detection in a screening population. In screening 4,965 women, the difference in sensitivity of FFDM and SFM was not statistically significant. However, FFDM had a significantly lower recall rate, and the positive biopsy rate for findings detected by FFDM was significantly higher than for findings detected by SFM.
RE Hendrick, JM Lewin, CJ D’Orsi, P Isaacs, L Moss, A Karellas, GA Sisney, CK Kuni, GR Cutter, University of Colorado, USA
Presentation #B4, presented at IWDM 2000, Toronto.

• View the Abstract

Non-Inferiority Study of FFDM in an Enriched Diagnostic Cohort: Comparison with Screen-film Mammography in 625 Women
Both FFDM and FSM exams were performed on 625 women aged 40 and over presenting for diagnostic mammography. FFDM resulted in a lower recall rate than FSM, meaning that fewer women were called back for unnecessary procedures. The sensitivity and ROC curve areas of FFDM and FSM were identical for breast cancer detection. Lesion visibility towards the skin line of the breast was superior with FFDM because of the tissue equalization post-processing.
RE Hendrick, A Sitzler*, K White, GR Cutter, Northwestern, USA
Presentation #B5, presented at IWDM 2000, Toronto.

• View the Abstract

Clinical Evaluation of Full-field Digital Mammography in a Screening Population
The study compared FFDM to FSM for cancer detection in a screening population. In the first 4945 subjects, no difference in cancer detection rate was been observed between FFDM and SFM. However, FFDM has so far led to fewer recalls than SFM. JM Lewin, RE Hendrick, CJ D'Orsi, LJ Moss, PK Isaacs, University of Colorado, USA
Abstract #1, presented at RSNA 2000.

• View the Abstract

Causes of Differences between Full-field Digital Mammography and Screen-film Mammography Interpretations in the Colorado/ Massachusetts Full-field Digital Mammography Screening Trial
The study determined the causes of differences in the interpretations of paired FFDM and SFM screening examinations. Breast cancer sensitivity was equal between FFDM and FSM, and in 19 cancers found on only one of FFDM or FSM, the cancers were equally conspicuous. The primary causes of differences in cancer detection were breast positioning and radiologist difference of opinion. FFDM had fewer false positive findings. The primary cause of differences in false positives was due to lesion conspicuity (53%), indicative of the high FFDM image quality.
JM Lewin, CJ D'Orsi, PK Isaacs, LJ Moss, RE Hendrick, University of Colorado, USA
Abstract #2, presented at RSNA 2000.

• View the Abstract

Clinical Evaluation of a Full Field Digital Mammography Prototype for Cancer Detection in a Screening Setting - Work in Progress
The study compared FFDM to FSM in the detection of breast cancer. The results did not demonstrate a significant difference between two modalities in sensitivity.
JM Lewin, RE Hendrick, CJ D’Orsi, LJ Moss, GA Sisney, A Karellas, et al, University of Colorado, USA
Abstract #418, presented at RSNA 1998.

• View the Abstract

Full-field digital mammography: A candid assessment
Interim results from the Army screening study indicated that FFDM had a lower recall rate and equal sensitivity to film-screen. To obtain valid comparative results, studies between FFDM and film-screen must use absolute truth, not using one modality as the gold standard. In addition, both modalities must have equal opportunity to select cases for recall or biopsy - otherwise the selection criteria will dominate the result, not the true difference between the accuracy of the modalities.
J Lewin, University of Colorado, USA
Diagnostic Imaging, Sept 99, pp. 40-45.

Initial Clinical Experience with Full-Field Digital Mammography
38 women were tested on FFDM and FSM to investigate grid vs. non-grid, Rhodium vs. Molybdenum, and standard dose vs. doubled dose. FFDM appeared comparable to the FSM in a cross section of clinical cases even when the film display parameters were not optimized.
RH Moore, DB Kopans, LT Niklason, PJ Slanetz, KA McCarthy, DA Hall, AA Giardino, S Kattapurian, MC Hanley, B Opsahl-Ong*, C Landberg*, D Castleberry*, J Eberhard*, Massachusetts General Hospital, USA
Abstract #570, presented at RSNA 1997.

• View the Abstract

Initial clinical experience with a high resolution full field digital mammography system
Initial clinical results of FFDM vs. FSM study were discussed. In this study, both systems were interpreted by different readers and the findings were compared. Further, discrepancies between readings were evaluated for etiology.
L Moss, C D’Orsi, A Karellas, E Hendrick, J Lewin, G Sisney, University of Massachusetts, USA
Presented at SCAR 1998, abstract in Journal of Digital Imaging, vol. 11, no. 3, Supplement 1, August 1998, p. 110.

• View the Abstract

Rate and Causes of Disagreement between Full-Field Digital Mammography and Film Screen Mammography in a Diagnostic Setting
The study determined the rate and causes of disagreement in interpretation of FFDM and FSM. The results showed that the disagreement rate was 2.9%, and the most frequent cause of disagreement was a difference in management approach between radiologists.
LA Venta, NI Kay, PM Mengoni, YT Adler, CA Beam, et al, Northwestern, USA
Abstract #962, presented at RSNA 1999.

• View the Abstract

Productivity

• View the Abstract

A Comparison of Reporting Times for Analog vs. Digital Mammography
The study compared the reporting time for hard-copy review of FSM with that of soft-copy review of FFDM. When loading time of hard copy films was taken into account, the study found no statistically significant difference in reporting time between hard versus soft-copy of reading of screening mammograms.
E O’Riordan, K Bukhanov, D Muradali, F Goldberg, University of Toronto, Canada
Abstract #7, presented at RSNA 2000.

• View the Abstract

Dose

• View the Abstract

Patient dose in full-field digital mammography in comparison with screen-film mammography
The study compared FFDM to FSM in terms of the radiation exposure. The results show that FFDM provided excellent image quality with a lower radiation exposure compared to FSM.
KP Hermann, S Obenauer, M Funke, E Grabbe, University of Göttingen, Germany
Presented at the German Radiology Congress 2000, Wiesbaden, abstract in Fortschr Röntgenstr (RöFo), vol. 172, 2000, S41.

Optimizing Radiation Quality in Full-Field Digital Mammography for Reducing the Patient Doses in Comparison to Screen-Film Mammography
The study evaluated the potential of radiation dose reduction with FFDM compared to FSM by using automatic mode controlled radiation. The results showed that dose values of FFDM were not significantly different from those of FSM. Some of the results obtained are: (1) raising peak voltage by 2 kVp decreases the average glandular dose (AGD) by 8 - 13 % for the same target/filter combination, and (2) switching the target/filter material from Mo/Mo to Rh/Rh for 40 mm PMMA lowered AGD by approximately 40 % at the same peak voltage values. Further, exposures with lower doses resulted in low SNR, which was the limiting factor for further decrease in dose.
KP Hermann, U Fischer, S Obenauer, E Grabbe, University of Göttingen, Germany
Abstract #B-0296, presented at ECR 2001.

• View the Abstract

Physics

• View the Abstract

Evaluation of a Full-field Clinical Prototype Flat Panel Imager for Digital Mammography
Detailed physics measurements of FFDM were investigated. Particularly noted was the improved DQE of the digital imager at low and mid-frequencies that could be advantageous in the imaging of low contrast soft tissue lesions.
Karellas, S Vedantham, I Levis, D Albagli*, S Han*, CE Landberg*, BH Opsahl-Ong*, RE Hendrick, CJ D'Orsi, University of Massachusetts, USA
Abstract #52, presented at RSNA 1998.

• View the Abstract

QC variations between four manufacturers whole breast digital mammography units
A QC phantom with automated image analysis was tested on four digital mammography systems. GE system had one of the highest SNR, which only varied 10%.
CM Kimme-Smith, L Yang, C Blatscher, LW Bassett, UCLA, USA
Abstract #967, presented at RSNA 1999.

• View the Abstract

Theoretical and Experimental Assessment of the Influence of Pixel Size on Microcalcification Detection and Diagnosis in Digital Mammography
The effect of pixel size of FFDM was analyzed in terms of diagnostic accuracy. Based on detectability and clinical accuracy studies, no significant difference was observed between 50 and 100-micron pixel size.
S Muller*, A Rick*, K Priday*, GE Medical Systems, France
Presentation #H4, presented at IWDM 2000, Toronto.

• View the Abstract

A quantitative model of microcalcification detection in digital mammography
This article discusses a simulation framework, used for analysis of microcalcification detection, for the image acquisition of FFDM. An improved selection scheme for the detection threshold is developed and evaluated clinically.
A Rick*, S Muller*, S Bothorel*, M Grimaud*, GE Medical Systems, France
Presented at MICCAI'99, Cambridge, England, September 1999.

• View the Abstract

Performance Evaluation of a Flat-Panel Filmless Full-Field Digital Mammography System
HM Rougeot*, B Opsahl-Ong*, DE Castleberry*, C Landberg*, JQ Liu, C Kimme-Smith, D Albagli*, M DeJule*, P Fitzgerald*, D Fobare*, B Giambattista, R Kwasnick*, S Lubowski*, G Possin*, J Richotte*, C Wei*, R Wirth*, GE Medical Systems, USA
Abstract #341, presented at RSNA 1996.

Physics of different breast imaging modalities
Different breast imaging modalities (i.e. infrared and ultrasound) were examined as well as other x-ray improvements such as microfiches and CAD.
M Thijssen, University of Nijmegen, Netherlands
Abstract #719, presented at ECR 2000.

• View the Abstract

Modeling the X-ray energy characteristics of DQE for full field digital mammography
The measurements of MTF, NPS and DQE for a FFDM detector were discussed. These metrics were measured using such factors as different exposure levels, x-ray tube voltages, and target composition and beam filtrations. Further, pitfalls to be avoided in the measurement methodology and modeling due to the effects of x-ray scatter were discussed, and available x-ray tube emission spectrum models were evaluated.
JE Tkaczyk*, J LeBlanc*, R Nevin*, G Kautz*, D Albagli*, J Sandrik*, P Granfors*, GE CRD, USA
Abstract #4320-21, presented at SPIE 2001.

• View the Abstract

Full breast digital mammography with an amorphous silicon-based flat panel detector: Physical characteristics of a clinical prototype
Detailed physics measurements of FFDM were investigated. DQE(0) of the FFDM system was 55% vs. 35% for measurements previously reported for FSM. The improved DQE of the digital imager at low and mid-frequencies could be advantageous in the imaging of low contrast soft tissue lesions.
S Vedantham, A Karellas, S Suryanarayanan, D Albagli*, S Han*, E Tkaczyk*, C Landberg*, B Opahl-Ong*, P Granfors*, I Levis, C D’Orsi, RE Hendrick, University of Massachusetts, USA
Medical Physics, vol. 27, no. 3, March 2000, pp. 558-567.

• View the Abstract

Breast imaging using an amorphous silicon-based full-field digital mammographic system: Stability of a clinical prototype
Physics measurement of FFDM was investigated. "Observations made over this 1-year period indicated that the maximum variation from the average of the measurements were less than 0.5% for conversion factor, 3% for presampling MTF over all spatial frequencies, 5% for signal response, linearity and sensitivity, 12% for SNR over seven locations for all 3 target-filter combinations, and 0% for ACR accreditation phantom scores." Furthermore, "flat field uniformity for all 3 target-filter combinations displayed no artifacts such as gridlines, bad detector rows or columns, horizontal or vertical streaks, or bad pixels. Wire mesh screen images indicated uniform resolution and no geometric distortion."
S Vedantham, A Karellas, S Suryanarayanan, CJ D'Orsi, RE Hendrick, University of Massachusetts, USA
Journal of Digital Imaging, Vol. 13, No. 4 (November), 2000: pp. 191-199.

• View the Abstract

Spectral Optimization for Digital Mammography
The study investigated the optimum spectra for FFDM. Higher energy spectra were required for denser breast tissue, while lower energy spectra were required for breast tissue with a higher percentage of fat. In general, FFDM required higher energy spectra and a wider range of peak kilovoltage than SFM.
V Venkatakrishnan, LT Niklason, C Landberg*, B Opsahl-Ong*, DE Castleberry*, DB Kopans, et al, Massachusetts General Hospital, USA
Abstract #680, presented at RSNA 1997.

• View the Abstract

Temporal Stability of Image Quality Between Screen-Film and Digital Mammography
The temporal stability of the FFDM system was equal or superior to FSM over a 6 month period in measurements of contrast, optical density, ACR phantom score stability.
L Yarusso, I Castro, R Nishikawa, University of Chicago, USA
Abstract #TU-D4-5, presented at AAPM 1999.

• View the Abstract

System Design

• View the Abstract

Phantoms used for Evaluation of Full Field Digital Mammography Systems
Existing image quality phantoms used by mammography facilities and FDA inspectors for evaluation of FSM may not be appropriate for FFDM systems. The effects of different phantoms and phantom material on the performance of an FFDM detector are discussed.
RV Kaczmarek, JA Thomas, K Chakrabarti, National Naval Medical Center, USA
Poster #TU-FXH-12, presented at the 2000 World Congress on Medical Physics, Chicago, July.

• View the Abstract

An automated quality control program for whole breast digital image receptors
Digital QC tests were evaluated with three digital mammo units.
C Kimme-Smith, C Lewis, L Yang, LW Bassett, UCLA, USA
4th International Workshop on Digital Mammography, June 7-10, 1998, University of Nijmegen, The Netherlands.

Influence of the pixel size on the diagnostic precision in digital mammography
A ROC study was performed to evaluate the impact of the pixel size on the diagnostic accuracy of microcalcifications. 38 digital images with biopsy proven microcalcification clusters were randomly selected from patient files. Images with pixel sizes of 50, 100, 150 and 200 µm were derived from images of 30 µm pixel size acquired on a 6 × 6 cm field of view digital system. Ten radiologists rated the images on a 7 level scale, and ROC curves corresponding to each pixel size have been derived. The result showed that ROC curve areas were similar for 50 and 100 µm pixel sizes. The null hypothesis (D Az > 0.1) can be rejected with a 95 % confidence (p-value = 0.006) and consequently, any potential degradation of the characterization performance when increasing the pixel size from 50 to 100 µm is smaller than 0.1.
L Levy, S Muller*, K Priday, A Rick*, Institut De Radiologie De Paris, France
Presented at JFR 2000, October 24, Room #05.

Impact of Pixel Size for the Differentiation of Benign and Malignant Microcalcification Clusters in Digital Mammography
A ROC study was performed to evaluate the impact of the pixel size on the diagnostic accuracy of radiologists in the process to differentiate malignant from benign clusters of microcalcifications. The ROC area for 100 micron pixels was higher than 50 micron pixels. With statistical significance, diagnosis with 100 micron pixels was not inferior to that of 50 micron pixels.
LD Levy, SL Muller*, K Priday*, A Rick*, Institut De Radiologie De Paris, France
Abstract #019BR, presented at RSNA 2000.

• View the Abstract

Influence of pixel size on diagnostic accuracy with digital mammography
A ROC study was performed to evaluate the impact of the pixel size on the diagnostic accuracy of microcalcifications. 38 digital images with biopsy proven microcalcification clusters were randomly selected from patient files. Images with pixel sizes of 50, 100, 150 and 200 µm were derived from images of 30 µm pixel size acquired on a 6 × 6 cm field of view digital system. Ten radiologists rated the images on a 7 level scale, and ROC curves corresponding to each pixel size have been derived. The result showed that ROC curve areas were similar for 50 and 100 µm pixel sizes. The null hypothesis (D Az > 0.1) can be rejected with a 95 % confidence (p-value = 0.006) and consequently, any potential degradation of the characterization performance when increasing the pixel size from 50 to 100 µm is smaller than 0.1.
L Levy, S Muller*, K Priday*, A Rick*, Institut De Radiologie De Paris, France
Abstract #A-0324, presented at ECR 2001.

• View the Abstract

Full-Field Digital Mammography designed as a complete system
A complete FFDM system was described in detail. Several advantages of the Senographe 2000D system include an excellent DQE curve and a very high exposure dynamic range.
S Muller*, GE Medical Systems, France
European Journal of Radiology, vol.31 (1), pp. 25-34, 1999.

• View the Abstract

Antiscatter Grids: Are Grids Required for Digital Mammography?
The study examined the usefulness of antiscatter grids on FFDM. Based on SNR measurements, a conventional antiscatter grid was beneficial and resulted in a loss of SNR for thicker breast.
LT Niklason, V Venkatakrishnan, BH Opsahl-Ong*, CE Landberg*, Massachusetts General Hospital, USA
Abstract #53, presented at RSNA 1998.

• View the Abstract

Quality assurance and dose in (digital) mammography
In imaging modalities, the higher dose might be preferred since it gives higher SNR in the image. A regular QC inspection must be performed since the easy manipulation of the image allows the observer to look at suboptimal images without being able to notice any difference.
M Thijssen, University of Nijmegen, Netherlands
Abstract #236, presented at ECR 2000.

• View the Abstract

Quality control phantom for full field digital mammographic systems
The development of a full field quality control phantom, that is used for acceptance testing and daily quality control measurements on FFDM, is described. Further presented are phantom design characteristics and measured parameters used for the feasibility system and FFDM.
JA Thomas, K Chakrabarti, RV Kaczmarek, MF Loscocco, J Gaskill, National Naval Medical Center, USA
Poster, presented at IWDM 2000, Toronto.

• View the Abstract

Grid Usefulness in Direct Digital Mammography
The study investigated if and when an anti-scatter grid is needed for a GE FFDM flat panel detector. The experiments involved the uses of CDMAM phantom with the smallest object diameter of 60 µm, and a homogeneous layer of perspex, simulating breast thickness. The results suggested that below a breast thickness of 5 cm the grid can be omitted which may allow lower dose to the patient.
WJH Veldkamp, MAO Thijssen, N Karssemeijer, University of Nijmegen, Netherlands
Abstract #B-0863, presented ECR 2001.

• View the Abstract

Experimental and Theoretical Spectral Optimization for Digital Mammography
Optimum technical parameters required for FFDM were determined by experimental and computer simulation methods. Using breast phantoms, peak kilovoltage, breast composition and thickness were varied, and usefulness of antiscatter grid were further determined.
V Venkatakrishnan, M Yavuz*, L Niklason, B Opsahl-Ong*, S Han*, C. Landberg*, R. Nevin*, L Hamberg, D Kopans, Massachusetts General Hospital, USA
Abstract #3659-13, Proceedings of SPIE 1999, Physics of Medical Imaging vol. 3659, pp. 142-149.

• View the Abstract

* GE Healthcare