New image processing technique for evaluating breast microcalcifications: a comparative study. J Ultrasound Med

Department of Radiology, Thomas Jefferson University, 132 S 10th St, Philadelphia, PA 19107, USA.
Journal of ultrasound in medicine: official journal of the American Institute of Ultrasound in Medicine (Impact Factor: 1.54). 06/2012; 31(6):885-93.
Source: PubMed


The purpose of this study was to evaluate a new commercial image processing technique (MicroPure; Toshiba America Medical Systems, Tustin, CA) for identifying breast microcalcifications compared to gray scale ultrasound imaging (US) using mammography as the reference standard.
Twenty women, with breast calcifications identified mammographically, underwent gray scale US and MicroPure examinations of the breast. Still images and digital clips of the target area were acquired using gray scale US and MicroPure (at 3 different sensitivity levels: 0, 1, and 2). The images were analyzed by 4 independent and blinded readers (2 radiologists and 2 physicists) to determine the number of calcifications as well as to score image quality and artifacts.
For all 4 readers, there were significantly more calcifications seen with MicroPure (at the 2 highest sensitivity levels) compared to gray scale US (P < .009). Agreement between readers consistently increased from gray scale US to MicroPure imaging (gray scale intraclass correlation coefficient, 0.02-0.44; versus MicroPure intraclass correlation coefficient, 0.34-0.71). The agreement improved between mammography and MicroPure (13.2%-28.3%) when compared with mammography and gray scale US (1.7%-5.2%); the 2 radiologists saw a bigger improvement. Two readers preferred the MicroPure image quality over gray scale US (P < .001) and vice versa for the other 2 readers(P < .001). All 4 readers saw fewer artifacts with MicroPure (at level 2) than with gray scale US (P < .02).
MicroPure imaging identified significantly more breast microcalcifications than gray scale US.

8 Reads
  • [Show abstract] [Hide abstract]
    ABSTRACT: This edition of Ultraschall in der Medizin - European Journal of Ultrasound - (UiM-EJU) includes 4 articles discussing breast ultrasound (US) 1234. In addition to prenatal sonography, a focal point of Volume 03/2012 of UiM-EJU 5678910, as well as contrast-enhanced US (CEUS) , contained in a separate supplementary volume 2012 1112131415161718, breast US is among the "Top 3" sonographic sub-specialties described in manuscripts accepted by UiM. This is reason enough to spotlight the broad spectrum of US technical applications and clinical breast US aspects of mammasonography in an editorial 19202122232425262728293031323334353637383940414243444546474849505152535455565758.Two aspects and articles should be particularly noted, as they treat US interventional evaluation of lymph node status of patients with breast cancer 1 and sonographic detection of microcalcifications 2.In a detailed review 1, I. Gruber et al. summarize the current state of the art US-guided tissue diagnosis, of sonographically suspicious axillary lymph nodes. Relatively recently this topic was also the subject of an editorial 41 discussing an article on using dynamic CEUS in distinguishing between benign and malignant lymph nodes in patients with breast cancer 42. One of the conclusions at that time stated: "If, using sonographic criteria, we find suspect axillary lymph nodes, then we should substantiate our suspicion histologically (preferably by performing US-guided core biopsy) 41. The accuracy of the biopsy is high, the complication rate low, and the positive verification of lymphatic metastasis avoids unnecessary sentinel lymph node excision (SLNE) 41. The present overview now illuminates the relevant literature on tissue diagnosis of suspect axillary lymph nodes 1 in cases of breast cancer and proposes an algorythm plan for pretherapeutic lymph node evaluation using sonographically guided, minimally-invasive interventions. This algorithm initially includes the use of fine needle aspiration (FNA, needle diameter 20G); if this is inconclusive, then a US-guided core biopsy, which, using the same high specificity as the FNA (both methods 100 % or nearly 100 %), possesses a somewhat higher sensitivity. The authors properly continue to explain that both an FNA as well as core biopsy with a negative zytology or histology for malignancy cannot completely exclude axillary lymphatic involvement. Thus the high positive predictive value of both techniques is relevant in avoiding SLNE in cases of positive findings of malignancy. SLNE should be performed if the biopsy is negativ for malignancy 1. The determination of FNA or core biopsy is certainly worthy of discussion; in general clinical practice, the choice of FNA or core biopsy relies on individual factors, such as size of the suspicious lymph node, its relation to blood vessels, examiner's experience, availability of zytologically competent pathologists, etc.The original article by T. Fischer et al. 2, describing US detection of breast microcalcifications is fascinating. In an in vitro experimental design, the authors compared sonographic detection of microcalcifications in 105 breast core specimens (using frequency compounding, tissue harmonic imaging and device-specific software [EMD - easy microcalcification detection, MicroPureTM, Toshiba, Otaware, Japan]) with specimen radiography and histology 2. Whereas in conventional B-mode US, microcalcifications could not be shown, microcalcification could be detected with EMB; the mean number of calcifications in US was 3.5 ± 3.1, in radiography, 4.3 ± 4.8; the difference was not significant. However, mammography, compared to the "gold standard" of histology in the detection of microcalcification, was significantly superior to ultrasound 2. And, subjectively, the authors write that particularly granular calcium clusters with more than 14 individual calcifications were better visualized using mammography 2. Further, size determination of calcifications was problematic in US 2. Setting 1 (blue color) of the EMD software was found to be the most sensitive compared to other settings and showed fewer artifacts 2.Isn't it about time to use high-resolution US with modern ultrasound technologies providing improved spatial resolution, using dedicated software instead of mammography? Would US be able to detect about 40 % of the non-palpable carcinomas in which microcalcification is the sole or primary malignancy criterion? The authors say "no", and in the introduction to their study 2 they properly state that "mammography remains the gold standard in the detection, characterization and localization of microcalcification." Why is this the case?Technical advancements in US have improved the detection of microcalcifications 2323452. According to BI-RADS 59, microcalcifications are characterized as "probably benign", "suspicious abnormality" (amorphous or indistinct calcifications) and "highly suggestive of malignancy" (pleomorphic or heterogeneous calcifications [granular], fine, linear or fine, linear dendritic calcifications), whereby the distribution pattern (clustered, linear, segmented, regional, diffuse/scattered) is an important characterization parameter. DEGUM US criteria from 2006 58, analogous to BI-RADS, distinguishes macrocalcification (> 0.5 mm) from microcalcification, within and outside a lesion. The ACR-BI-RADS US lexicon 60 differentiates between microcalcification and macrocalcification, the last within and outside a mass. In this instance hyperechoic foci in a hypoechoic mass are considered "conspicuous" 60. This means that the morphology of microcalcifications (and thus benign/malignant characterization) is not possible in US, as is likewise indicated in the study 2 of breast core specimens. It should also be taken into account that not all "hyperechoic foci" in a breast US are microcalcifications; collagen fibers and other changes might sumulate microcalcifications 37.Even if in vitro microcalcification can be detected with US, the result cannot be applied to a routine ultrasound breast examination. Some authors are of the opinion that calcium can be easily detected in hypoechoic lesions and dilated milk ducts 1920212324252735, whereas isolated microcalcification without an encompassing mass are harder to detect; there is little literature on this. H. Majdar et al. 44, found that of 21 carcinomas primarily discovered by sonography (of a total of 86 carcinomas), only one ductal carcinoma in situ (DCIS) was identified, the other carcinomas were invasive ductal (n = 16), invasive lobular (n = 3) and invasive tubular (n = 1), that is, masses with an average diameter of 12.6 mm. Therefore ultrasound is inappropriate for screening for suspicious microcalcifications, since about 25 % of all carcinomas discovered in mammographic screening are DCIS 28. In the Austrian screening program (introduction is planned for 2013), ultrasound will also not be used for detection of microcalcification, but will be employed to improve the detection of carcinomas in mammographically dense breasts (ACR stage 3 and 4) 2661.Most studies of US and microcalcifications therefore investigated microcalcifications previously detected and identified mammographically 1921222324252730323334353738; they do, however agree with the advantage of using US-guided core or vacuum biopsy of the microcalcification if can also be sonographically classified. This topic continues to be discussed, and very poor results for US detection of mammographically-visible microcalcification have been reported 38. At the other end of the spectrum is a recent publication describing the useful employment of US (together with mammography) in the case of DCIS in the context of a breast-conserving operation 36.The final sentence in the article by T. Fischer et al. reads: "Mammography remains the gold standard for the detection of microcalcifications" 2. This statement could be soon disproved by non-overlapping digital tomosynthesis of the breast. However, we should recognize the appearance of microcalcifications inside of and outside of masses in order to use ultrasound in US-guided tissue diagnosis of these changes for reasons of economy and for the benefit of the patient.
    Ultraschall in der Medizin 08/2012; 33(4):316-9. DOI:10.1055/s-0032-1313061 · 4.92 Impact Factor
  • [Show abstract] [Hide abstract]
    ABSTRACT: The process of echo-signal demodulation within the display stage of ultrasonic image formation discards signal phase. It has long been hypothesized that demodulation could be eliminating important clinical task information but the tools to study this effect were not developed. We have now developed a task-energy analysis to show how signal energy flows through different stages of image formation. In this paper, we show how traditional displaystage processing eliminates high spatial-frequency task information, and how simple methods can recover the loss for improved diagnostic performance. We also study the improvement in detecting breast microcalcifications using the proposed method.
    Proceedings of SPIE - The International Society for Optical Engineering 02/2014; 9037. DOI:10.1117/12.2043955 · 0.20 Impact Factor
  • Source
    [Show abstract] [Hide abstract]
    ABSTRACT: Background We presented the photoacoustic imaging (PAI) tool and to evaluate whether microcalcifications in breast tissue can be detected on photoacoustic (PA) images. Methods We collected 21 cores containing microcalcifications (n = 11, microcalcification group) and none (n = 10, control group) in stereotactic or ultrasound (US) guided 8-gauge vacuum-assisted biopsies. Photoacoustic (PA) images were acquired through ex vivo experiments by transmitting laser pulses with two different wavelengths (700 nm and 800 nm). The presence of microcalcifications in PA images were blindly assessed by two radiologists and compared with specimen mammography. A ratio of the signal amplitude occurring at 700 nm to that occurring at 800 nm was calculated for each PA focus and was called the PAI ratio. Results Based on the change of PA signal amplitude between 700 nm and 800 nm, 10 out of 11 specimens containing microcalcifications and 8 out of 10 specimens without calcifications were correctly identified on blind review; the sensitivity, specificity, accuracy, positive predictive and negative predictive values of our blind review were 90.91%, 80.0%, 85.71%, 83.33% and 88.89%. The PAI ratio in the microcalcification group was significantly higher than that in the control group (the median PAI ratio, 2.46 versus 1.11, respectively, P = .001). On subgroup analysis in the microcalcification group, neither malignant diagnosis nor the number or size of calcification-foci was proven to contribute to PAI ratios. Conclusion Breast microcalcifications generated distinguishable PA signals unlike breast tissue without calcifications. So, PAI, a non-ionizing and non-invasive hybrid imaging technique, can be an alternative in overcoming the limitations of conventional US imaging.
    PLoS ONE 08/2014; 9(8):e105878. DOI:10.1371/journal.pone.0105878 · 3.23 Impact Factor
Show more