A nonrigid registration of MR breast images using complex-valued wavelet transform.

In this paper, a fast, slice-by-slice, nonrigid registration algorithm of dynamic magnetic resonance breast images is presented. The method is based on a multiresolution motion estimation of the breast using complex discrete wavelet transform (CDWT): the pyramid of oriented complex subimages is used to implement a hierarchical phase-matching-based motion estimation algorithm. The resulting motion estimate is nonrigid and pixel-independent. To assess the method performance, we computed the correlation coefficient and the normalized mutual information between pre- and postcontrast images with and without realignment. The indices increased after using our approach and the improvement was superior to rigid or affine registration. A set of clinical scores was also evaluated. The clinical validation demonstrated an increased readability in the subtraction images. In particular, CDWT registration allowed a best definition of breast and lesion borders and greater detail detectability.

BMRI in early detection of breast cancer in patients with increased genetic risk: our preliminary results.

Because of the higher risk of developing breast cancer and the early onset of the disease in women proved or suspected to be carriers of a breast cancer susceptibility gene, a dedicated screening should be offered as a less invasive approach with respect to the otherwise suggested prophylactic mastectomy. This should be optimized in order to overcome the limitations of conventional breast imaging with the application of new technologies such as Breast Magnetic Resonance Imaging (BMRI). A diagnostic protocol for routine control in patients with high risk for developing breast cancer has been prepared. Within a 7 months period, 23 patients suspected or proved to carry a breast cancer susceptibility gene underwent BMRI. Four breast cancers were identified with BMRI. In these cases mammography was negative because of the density of the parenchyma or for its fibroglandular pattern. US was negative in two cases, not specific for malignancy in one case and considered as only possibly malignant but with biopsy recommendation on the basis of MR findings in the last one. Clinic analysis was positive for mass in two cases. The accuracy of BMRI is known to be higher than that of conventional imaging in the study of breast parenchyma. High spatial resolution and no breast density influence can give more detailed information about smaller lesions and the right extent of the disease.

The Italian multi-centre project on evaluation of MRI and other imaging modalities in early detection of breast cancer in subjects at high genetic risk.

This report presents the preliminary results of the first phase (21 months) of a multi-centre, non-randomised, prospective study, aimed at evaluating the effectiveness of contrast-enhanced magnetic resonance imaging (MRI), X-ray mammography (XM) and ultrasound (US) in early diagnosis of breast cancer (BC) in subjects at high genetic risk. This Italian national trial (coordinated by the Istituto Superiore di Sanità, Rome) so far recruited 105 women (mean age 46.0 years; median age 51.0; age range 25-77 years), who were either proven BRCA1 or BRCA2 mutation carriers or had a 1 in 2 probability of being carriers (40/105 with a previous personal history of BC). Eight cases of breast carcinomas were detected in the trial (mean age 55.3 years, median age 52.5; age range 35-70 years; five with previous personal history of BC). All trial-detected BC cases (8/8) were identified by MRI, while XM and US correctly classified only one. MRI had one false positive case, XM and US none. Seven "MRI-only" detected cancers (4 invasive, 3 in situ) occurred in both pre- (n = 2) and post-menopausal (n = 5) women. With respect to the current XM screening programmes addressed to women in the age range 50-69 years, the global incidence of BC in the trial (7.6%) was over ten-fold higher. The cost per "MRI-only" detected cancer in this particular category of subjects at high genetic risk was substantially lower than that of an XM-detected cancer in the general women population. These preliminary results confirmed that MRI is a very useful tool to screen subjects at high genetic risk for breast carcinoma, not only in pre-, but also in post-menopausal age, with a low probability of false positive cases.

Lobular breast cancer: how useful is breast magnetic resonance imaging?

RATIONALE AND OBJECTIVES: To review magnetic resonance imaging (MRI) findings in lobular breast carcinoma, the in situ or infiltrating subtype, with special attention to the dynamic curves with the aim to evaluate possible differences with ductal carcinoma. METHODS: In 2 years, 27 patients with lobular and one with tubular carcinoma underwent MRI at the Istituto Nazionale Tumori of Milan. RESULTS: All lobular carcinomas demonstrated early or late enhancement (100% sensitivity), without significant differences in morphology compared with ductal carcinoma, but frequently with a different shape of the dynamic curves. CONCLUSIONS: Due to its infiltrative growth associated to only limited connective tissue reaction, lobular carcinoma often encounters difficulties in mammographic diagnosis. In contrast, MRI can be very helpful in evaluating the true extension of the disease, especially when breast conservation is considered. Due to a more consistent fibrotic stroma, these lesions sometimes show a delayed enhancement, which suggests that more than one set of subtracted images should be evaluated during MRI analysis.

[Peripheral dysostosis, hypoplasia of the nose, low stature. Description of a case]. / Disostosi periferica, ipoplasia del naso, bassa statura. Descrizione di un caso.

The authors describe the case of a five-year old girl presenting short stature. The evidence of small hypoplasic nose, of abnormally small hands and feet with typical radiological alterations and psychomotor retardation, leads to the diagnosis of acrodysostosis. Moreover, the endocrinological investigations disclosed a dysequilibrium in the pituitary-thyroid axis, which, to our knowledge, has never been described to date in the syndrome of acrodysostosis.

A Fuzzy-C-means clustering algorithm for a volumetric analysis of paranasal sinus and nasal cavity cancers.

In this paper, a semi-automatic segmentation algorithm for volumetric analysis of paranasal sinus and nasal cavity cancers is presented and validated. The algorithm, based on a semi-supervised Fuzzy-C-means method, was applied to a Magnetic Resonance data sets (each of them composed by T1-weighted, Contrast Enhanced T1-weighted and T2-weighted images) for a total of 64 tumor-contained slices. Method performances are tested by both a numerical and a clinical validation. Results show that the proposed method has a higher accuracy in quantifying lesion area than a region growing algorithm and it can be applied in the evaluation of tumor response to therapy.

[The current state of the diagnostics of breast diseases in Lombardy]. / Indagine conoscitiva sulla situazione attuale della diagnostica senologica nella Regione Lombardia.

PURPOSE: The SIRM study group for senology in Lombardy set the following goals: to quantify the number of radiologist working in public radiology centers and in private credited hospitals in Lombardy; to quantify the number of radiologists working in diagnostic senology and the mean time they devote to this field; to survey the diagnostic equipment used for senology purposes and check its adequacy. MATERIAL AND METHODS: We considered 58 centers of senology diagnostic imaging in the radiology departments of general hospitals in 8 Lombardy provinces. First we evaluated the screening programs for early breast cancer detection carried out with clinical mammography. All data were evaluated with reference to female residents aged 40 to 69, to the average adherence rates reported in the literature, to the protocols adopted in different centers relative to the examination frequency. Mammography was always associated with a physical examination and then the patient was submitted to radiological investigations to make the final diagnosis. The form we used for data report is summarized in Table I. RESULTS: In the centers for diagnostic senology we surveyed in 1998, in all 182,724 mammograms, 58,686 breast US examinations and 7,097 needle biopsies for cytology or microhistology were performed; 35.5% of the female population which should have been screened actually underwent the examination and 32.6% of them was also submitted to breast US, while 3.88% underwent FNAB. Sixty-three mammographic units and 62 US units were operated by 152 radiologists. An average 2,900 radiographic investigations per mammographic unit were performed yearly. Each radiologist reported an average 1,202 mammograms a year. Refer to the tables for detailed data reporting by province. Thirty-six per cent of radiologists presently reads more than 1,500 breast examinations a year. The radiologists working in diagnostic imaging in senology devote an average 20% of their working time to this field. As for the technical adequacy of mammographic and US equipment relative to some reference levels reported in detail in the paper, only 42% of the former and 48% of the latter were up to standards. DISCUSSION AND CONCLUSIONS: Despite the massive work done in the field of senology, the estimated needs of the female population have not been met yet, given the increased demand for senologic examinations and women's growing awareness of the need of early breast cancer diagnosis. The answer to this problem lies in a political health care policy that should promote the current clinical programs increasing both instrumental and human resources. Alternatively screening programs could be planned on a regional basis, with the pro of reaching a larger female population and that of a more homogeneous methodological approach. Our data indicate that breast cancer detection is best achieved with clinical and screening tools: the former can be easily promoted in a short time and could thus make our first goal, while the latter, which is more complex, could be implemented later. Thus, a growing clinical activity could make a very good basis for high quality breast screening programs. Finally, it is necessary to guarantee high quality standards for equipment, methods and training of medical and nonmedical staff.

[Hepatic and hepatocarcinoma magnetic resonance: comparison of the results obtained with paramagnetic (gadolinium) and superparamagnetic (iron oxide particles) contrast media]. / Risonanza Magnetica epatica ed epatocarcinoma: confronto fra risultati ottenuti con mezzo di contrasto paramagnetico (gadolinio) e superparamagnetico (particelle di ossido di ferro).

PURPOSE: To compare prospectively dynamic gadolinium (Gd)-enhanced with superparamagnetic iron oxide (SPIO)-enhanced MRI for the detection of hepatocellular carcinoma (HCC). MATERIAL AND METHODS: Twenty-five patients with histologically proven HCC and liver cirrhosis (28% of them in B or C Child class) underwent dynamic Gd-enhanced MRI and, a few days later, (mean interval: three days) SPIO-enhanced MRI. Only patients with availability of clinical and imaging follow-up for at least seven months were enrolled in this prospective study. Axial dynamic Gd-enhanced imaging was performed with T1 gradient-recalled echo (GRE) sequences. Both axial and sagittal SPIO-enhanced imaging were performed with respiratory triggered T2-weighted turbo spin-echo (TSE) and T1-T2*-weighted GRE sequences. MR images were reviewed by two independent radiologists. The readers scored each lesion for the presence of HCC and assigned confidence levels based on a five-grade scale: 1, definitely or almost definitely absent; 2, possibly present; 3, probably present; 4, definitely present; 5, definitely present with optimal liver/lesion contrast or good liver/lesion contrast and morphological signs (intact capsule, intranodular septa, extra-capsular infiltration), useful for locoregional treatment planning. A positive diagnostic value was assessed for scores of 3 or higher. RESULTS: Gd-enhanced and SPIO-enhanced MRI found 44 lesions. The combined use of TSE and GRE SPIO-enhanced sequences detected 11 more lesions (25% improvement in sensitivity) than Gd-enhanced MRI. One lesion (2.27%) was detected only with Gd-enhanced MRI. Eight of twelve lesions visible with a single contrast agent measured less than 1 cm in diameter. HCC detectability was 75% with Gd-enhanced MRI and 97.7% with SPIO-enhanced MRI. SPIO-enhanced T2-weighted TSE images showed significantly higher diagnostic value than SPIO-enhanced T1-T2*GRE images only in three cases, while nodule morphological characteristics (capsule, septa, different cell differentiation components) were better depicted by TSE images. DISCUSSION AND CONCLUSIONS: In our study the combined use of SPIO-enhanced T2-weighted TSE and T1-T2*-weighted GRE sequences showed higher sensitivity than gadolinium-enhanced GRE dynamic imaging (97.7% versus 75%). These results are at least partly related to our study conditions, that is: 1) MRI was performed with a 1T system, 2) both axial and sagittal SPIO-enhanced imaging were performed with respiratory triggered T2-weighted TSE and T1-T2*-weighted GRE, 3) there was a low freaquency of severe cirrhosis.

Validation of an elastic matching algorithm based on complex wavelets for the realignment of dynamic MR breast images.

In this paper, a novel approach for the registration of dynamic magnetic resonance (MR) breast images is validated. The approach is fully automatic and it performs a multi-resolution motion field estimation based on complex discrete wavelets transform (CDWT). The validation was designed to assess the registration quality in patient breast data. A set of diagnostic features of MR breast image was defined and two expert radiologists were asked to score the capability to detect these features in the subtraction images. Two registration methods were compared: a rigid registration algorithm and the proposed approach. Results show that the novel approach is superior in terms of both lesion detectability and lesion definition.

A use of a neural network to evaluate contrast enhancement curves in breast magnetic resonance images.

For the diagnosis of breast cancer using magnetic resonance imaging (MRI), one of the most important parameters is the analysis of contrast enhancement. A three-dimensional MR sequence is applied before and five times after bolus injection of paramagnetic contrast medium (Gd-DTPA). The dynamics of absorption are described by a time/intensity enhancement curve, which reports the mean intensity of the MR signal in a small region of interest (ROI) for about 8 minutes after contrast injection. The aim of our study was to use an artificial neural network to automatically classify the enhancement curves as "benign" or "malignant." We used a classic feed-forward back-propagation neural network, with three layers: five input nodes, two hidden nodes, and one output node. The network has been trained with 26 pathologic curves (10 invasive carcinoma [K], two carcinoma-in-situ [DCIS], and 14 benign lesion [B]). The trained network has been tested with 58 curves (36 K, one DCIS, 21 B). The network was able to correctly identify the test curves with a sensitivity of 76% and a specificity of 90%. For comparison, the same set of curves was analyzed separately by two radiologists (a breast MR expert and a resident radiologist). The first correctly interpreted the curves with a sensitivity of 76% and a specificity of 90%, while the second scored 59% for sensitivity and 90% for specificity. These results demonstrate that a trained neural network recognizes the pathologic curves at least as well as an expert radiologist. This algorithm can help the radiologist attain rapid and affordable screening of a large number of ROIs. A complete automatic computer-aided diagnosis support system should find a number of potentially interesting ROIs and automatically analyze the enhancement curves for each ROI by neural networks, reporting to the radiologist only the potentially pathologic ROIs for a more accurate, manual, repeated evaluation.

Tubular nephrotoxicity after intravenous urography with ionic high-osmolal and nonionic low-osmolal contrast media in patients with chronic renal insufficiency.

Nephrotoxicity of intravenous contrast media is more frequent and striking in patients with risk factors, the major one being preexisting chronic renal insufficiency. New nonionic low-osmolal contrast media allegedly have less nephrotoxicity than the traditional ionic high-osmolal ones. This was tested for two contrast media in a group of 18 patients with stable chronic renal insufficiency. The urinary excretion of two brush-border enzymes (alanine aminopeptidase, AAP, and gamma-glutamyl transpeptidase, gamma-GT) and of a lysosomal enzyme (N-acetyl-beta glucosaminidase, NAG), functional markers of tubular injury, were measured before and after intravenous urography with an ionic high-osmolal radiocontrast medium, meglumine sodium diatrizoate, or with a non ionic low-osmolal one, iopamidol. Urinary NAG excretion did not change significantly after administration of either contrast media. Urinary AAP and gamma-GT excretion increased significantly (p less than 0.01) after diatrizoate. After iopamidol, only gamma-GT excretion increased significantly (p less than 0.05). Our data suggest that the nonionic low-osmolal radiocontrast medium iopamidol is less toxic to tubules than the ionic high-osmolal medium diatrizoate and that the brush-border enzymes AAP and gamma-GT are sensitive markers for this toxicity.

[Nephrotoxicity of contrast media. Comparison of iopamidol and diatrizoate by measurement of urinary enzymes after urography]. / Nefrotossicità da mezzi di contrasto. Confronto tra iopamidolo e diatrizoato mediante dosaggio di enzimi urinari dopo urografia.

An ionic (diatrizoate) and a non ionic (iopamidole) radiocontrast medium were compared as to their nephrotoxicity in a cohort of 21 patients with renal disease, 18 of which with normal renal function. The elevation of the urinary excretion of enzymes from renal tubular cells has been considered as a good index for renal tubular damage. We measured two lysosomal enzymes (NAG and beta glucuronidase) and a brush-border enzyme (gamma GT). In all patients we demonstrated an elevation of urinary excretion of the three enzymes already 12 hours after contrast medium administration. However, the elevation was statistically significant only after employment of the ionic contrast medium and concerned gamma GT and beta-glucuronidase; the urinary excretion of NAG did not vary significantly. Urinary enzymes levels returned to basal values 36 hours after intravenous pyelography. In conclusion, iopamidole showed a lower nephrotoxicity with respect to diatrizoate, as demonstrated by the lower levels of urinary enzymes excretion reached after its use.