Classification of breast ultrasound images using fractal feature

Fractal analyses have been applied successfully for the image compression, texture analysis, and texture image segmentation. The fractal dimension could be used to quantify the texture information. In this study, the differences of gray value of neighboring pixels are used to estimate the fractal dimension of an ultrasound image of breast lesion by using the fractal Brownian motion. Furthermore, a computer-aided diagnosis (CAD) system based on the fractal analysis is proposed to classify the breast lesions into two classes: benign and malignant. To improve the classification performances, the ultrasound images are preprocessed by using morphology operations and histogram equalization. Finally, the k-means classification method is used to classify benign tumors from malignant ones. The US breast image databases include only histologically confirmed cases: 110 malignant and 140 benign tumors, which were recorded. All the digital images were obtained prior to biopsy using by an ATL HDI 3000 system. The receiver operator characteristic (ROC) area index AZ is 0.9218, which represents the diagnostic performance.     

Papillary lesions of the breast: evaluation with stereotactic directional vacuum-assisted biopsy.

PURPOSE: To evaluate the use of stereotactic directional vacuum-assisted biopsy (SDVAB) in diagnosing and managing papillary lesions of the breast. MATERIALS AND METHODS: The authors retrospectively reviewed the mammographic and histopathologic findings of 26 cases in which papillary lesions were diagnosed at SDVAB. In all cases, subsequent surgical excision (n = 20) or long-term imaging follow-up (n = 6) was performed and correlated with findings at SDVAB. RESULTS: SDVAB of 26 lesions yielded tissue that was classified as benign in 12, atypical in six, and malignant in eight. Of the 12 lesions that were diagnosed as histologically benign at SDVAB, six were surgically excised. Of these six lesions, five yielded benign correlative results. The sixth lesion was thought to be discordant with the imaging findings, and was surgically excised and determined to be malignant. Of the six benign lesions that were not surgically sampled for biopsy, five decreased in size and one was not seen at radiographic follow-up. Of the six lesions diagnosed as atypical at SDVAB that were surgically excised, one was benign and five were atypical. None proved to be malignant. Of the eight lesions diagnosed as malignant at SDVAB, surgical excision demonstrated ductal carcinoma in situ in all eight; two also had foci of invasive carcinoma. CONCLUSION: Benign and malignant papillary lesions of the breast can be reliably diagnosed at SDVAB when the SDVAB results correlate with the imaging findings. However, the extent of malignant papillary disease may be underestimated at SDVAB; in our study, invasive carcinoma was later discovered in 25% of patients with this diagnosis.     

高频乳腺X线片对乳腺良恶性病变钙化的分析

目的研究高频乳腺X线片对鉴别乳腺良、恶性病变钙化的意义。方法152例经临床和病理证实的女性乳腺良、恶性病变患者（87例良性病变和65例恶性病变）均经高频乳腺X线摄影。回顾性分析了所有患者的乳腺X线表现，并着重鉴别了乳腺良、恶性病变钙化的X线特征。结果在乳腺X线片上，大多数良性病变的钙化表现为散在分布的粗颗粒状、环状或斑片状灶，形态规则，密度较高，而恶性病变的钙化多呈簇状分布的细颗粒与导管型，诸如泥沙状、短棒状、针尖状灶，形态不规则，密度较低。结论乳腺病变钙化的X线表现随病变性质而不同，因此，高频乳腺X线片对鉴别乳腺良、恶性病变起重要作用。     

Benign breast lesions that simulate malignancy: magnetic resonance imaging with radiologic-pathologic correlation

The typical appearance of benign breast conditions on magnetic resonance imaging (MRI) is well established and diagnosis is usually easy. However, cases of benign breast lesions that are extremely difficult to differentiate from malignant breast tumors are occasionally encountered in MRI of the breast because overlap between benign and malignant lesions characteristics is found. This article describes the MRI features of a variety of suspicious breast conditions that were confirmed to be benign in the histopathologic study. We evaluated both enhancement kinetics and lesion morphological information to differentiate malignant from benign lesions. We also correlated the MRI findings with clinical data, and mammographic, ultrasound, and pathologic findings. Lesions evaluated included benign proliferative breast disease, fibroadenoma, intraductal papilloma, granular cell tumor, pseudoangiomatous stromal hyperplasia, fat necrosis, mastitis, inflammatory granuloma, epidermal inclusion cyst, and benign intramammary lymph node.     

Breast lesions: differential diagnosis using digital subtraction angiography

To evaluate its potential for differentiating benign from malignant breast lesions, digital subtraction angiography of the breast (DSAB) was performed in 23 women with mammographic evidence of disease, and the results were compared with surgical biopsy findings. The DSAB technique employed breast immobilization with modest compression and bolus injection; following the injection of contrast material, 30-40 sequential subtraction images were obtained over a 5-minute interval. The average technical settings were 50 k Vp and 10 mAs, resulting in an estimated radiation dose to the breast of 0.05 mrad (0.5 mu Gy) per exposure. DSAB consistently demonstrated retention of contrast material and abnormal vasculature in malignant lesions but not in benign lesions. In the 22 breast lesions for which there was histopathologic correlation, DSAB correctly categorized eight of nine malignant and 11 of 13 benign lesions. Although this series is small, the initial results of DSAB suggest its potential for differentiating benign from malignant lesions.     

Sonographic criteria for differentiation of benign and malignant solid breast lesions: size is of value

PURPOSE: The purpose of this study was to evaluate whether lesion size may influence the value of sonographic findings in the differential diagnosis between benign or malignant breast lesions. MATERIALS AND METHODS: Sonographic features of 256 histologically confirmed (148 benign, 108 malignant) breast lesions were retrospectively and independently reviewed by three radiologists unaware of mammographic findings and pathology results. Each lesion was assessed for several sonographic features and assigned a level of suspicion. Logistic modelling defined the predictive value of each sonographic feature per se and in relation to lesion size. The k statistic (k) evaluated interobserver agreement in lesion classification. The accuracy of breast sonography in characterising solid lesions was also evaluated. RESULTS: Analysis of the sonographic features predictive of malignant disease, taken as a whole, showed that only irregular margins and marked hypoechogenicity maintain their predictive value independent of lesion size. When lesion size is considered, the other features remain significant only for lesions larger than 7 mm. Interobserver agreement for sonographic suspicion, when calculated not taking into account lesion size, was good or excellent whereas it was reduced for lesions smaller than 7 mm. Accuracy of breast sonography improved when evaluating lesions larger than 7 mm. CONCLUSIONS: Lesion size influences the value of sonographic findings in distinguishing benign from malignant lesions. The usually adopted criteria in sonography have a significantly lower accuracy in characterisation of small lesions.     

Periodic mammographic follow-up of probably benign lesions: results in 3,184 consecutive cases

The author prospectively evaluated the value of periodic mammographic surveillance among 3,184 consecutive cases of nonpalpable, probably benign breast lesions detected with mammography. Follow-up consisted of four mammographic examinations during a 3- or 3.5-year period. Clinical outcome was ascertained in each case after the study period, whether or not patients complied with the protocol. Probably benign lesions were subsequently found to be malignant in 17 cases (positive predictive value for cancer, 0.5%). Fifteen of the 17 cancers were identified by means of interval mammographic change prior to development of a palpable mass; all 17 were stage 0 or stage 1 tumors. All 17 women who had cancer currently show no evidence of tumor recurrence (median duration of follow-up, 5 years). These results should help establish the validity of managing mammographically detected, probably benign lesions with periodic mammographic surveillance. By decreasing the number of biopsies of benign lesions and thereby substantially reducing costs, this approach may help overcome a major barrier to widespread use of mammographic screening.     

Mammography-guided stereotactic fine-needle aspiration cytology of nonpalpable breast lesions: prospective comparison with surgical biopsy results

We assessed the usefulness of fine-needle aspiration cytology (FNAC) in evaluating nonpalpable breast abnormalities by prospectively performing stereotactic mammography-guided FNAC on 100 women undergoing surgical excisional biopsy. Mammographic and cytologic diagnoses, on a scale of 1 (benign) to 4 (malignant), were assigned for each case and compared with the surgical pathologic diagnosis. Sensitivity and specificity were examined at different diagnostic cutoff points for regarding a mammographic or cytologic diagnosis as positive or negative. Of the 100 breast biopsy specimens, 70 were benign and 30 were malignant. For both mammography and FNAC, the optimal diagnostic cutoff point was between diagnosis 2 (mammography, probably benign; cytology, atypical) and 3 (mammography and cytology both suspicious for malignancy). At this cutoff point, FNAC had a sensitivity of 0.77 and specificity of 1.00, vs 0.73 sensitivity and 0.79 specificity for mammography. Pearson coefficient analysis revealed significant correlations between both mammographic and FNAC diagnoses and surgical pathology (p less than .001 for both). Our results suggest that use of mammography-guided FNAC may reduce the number of breast biopsies performed for benign lesions.     

MRI 对乳腺 X 线摄影中“局灶性不对称”征象的诊断价值

目的：探讨M RI对X线摄影中乳腺“局灶性不对称”（FAB ）征象的诊断价值。方法回顾性分析经病理证实的46例FAB的M RI形态学特点及动态增强时间‐信号强度曲线（T IC ）的类型。结果在FAB相应区域,M RI检出46例病灶,病理证实良性病变35例,恶性病变11例。其中外形分叶及边缘毛刺的准确度、阳性预测值以及阴性预测值分别为82．61％、66．67％、86．49％和86．96％、85．71％、87．18％。T IC显示：Ⅰ型17例,均为良性;Ⅱ型6例,其中良性4例,恶性2例;Ⅲ型9例,均为恶性。M RI对其良恶性诊断的准确度为86．96％,阳性预测值和阴性预测值分别为81．82％与96．88％。结论 M RI能够准确判断FAB中是否存在潜在病灶,并能较准确地判断其良恶性。     

Discrimination of MR images of breast masses with fractal-interpolation function models

RATIONALE AND OBJECTIVES: The authors evaluated the feasibility of using statistical fractal-dimension features to improve discrimination between benign and malignant breast masses at magnetic resonance (MR) imaging. MATERIALS AND METHODS: The study evaluated MR images of 32 malignant and 20 benign breast masses from archived data at the University of Pennsylvania Medical Center. The test set included four cases that were difficult to evaluate on the basis of border characteristics. All diagnoses had been confirmed at excisional biopsy. The fractal-dimension feature was computed as the mean of a sample space of fractal-dimension estimates derived from fractal interpolation function models. To evaluate the performance of the fractal-dimension feature, the classification effectiveness of five expert-observer architectural features was compared with that of the fractal dimension combined with four expert-observer features. Feature sets were evaluated with receiver operating characteristic analysis. Discrimination analysis used artificial neural networks and logistic regression. Robustness of the fractal-dimension feature was evaluated by determining changes in discrimination when the algorithm parameters were perturbed. RESULTS: The combination of fractal-dimension and expert-observer features provided a statistically significant improvement in discrimination over that achieved with expert-observer features alone. Perturbing selected parameters in the fractal-dimension algorithm had little effect on discrimination. CONCLUSION: A statistical fractal-dimension feature appears to be useful in distinguishing MR images of benign and malignant breast masses in cases where expert radiologists may have difficulty. The statistical approach to estimating the fractal dimension appears to be more robust than other fractal measurements on data-limited medical images.     

Pleomorphic adenoma of the breast

Pleomorphic adenoma is a very rare benign tumor of the breast. Only 70 cases have been reported in the world literature. Recently, we encountered a case of pleomorphic adenoma of the breast and thus present here the mammographic and ultrasonographic findings with the pathology of this rare breast tumor. The patient was a 76-year-old Japanese woman with a right breast mass. The mammography showed a 1.5-cm, lobulated high-density mass with partially ill-defined margins. Ultrasonography revealed an irregularly shaped mass with partially ill defined borders, hypoechoic and heterogeneous internal echoes, and posterior acoustic enhancement. These findings suggested an invasive carcinoma. Awareness of this type of tumor will help in correct diagnosis, in spite of the rarity of this disease.     

Benign and malignant mucocele-like tumors of the breast: mammographic and sonographic appearances

OBJECTIVE: The purpose of this study was to analyze and compare the mammographic and sonographic appearances of benign and malignant mucocele-like tumors. CONCLUSION: The mammographic appearance of mucocele-like tumor of the breast is characterized as pleomorphic calcifications, often increasing in number. Microcalcifications in malignant mucocele-like tumors extended over a wider area than those in benign mucocele-like tumors. Sonography often shows cysts with calcified or noncalcified mural nodules.     

False-negative MR imaging of malignant breast tumors

In this study we analyze MR-negative malignant lesions of the breast. A total of 204 patients with palpable and/or mammographic lesions were studied. The MR technique consisted of the turbo FLASH and MP-RAGE subtraction techniques. All patients underwent surgical biopsy and/or mastectomy and all specimens were examined by the correlative radiologic-histologic mapping technique. A total of 208 lesions were evaluated; 145 turned out to be malignant and 63 proved to be benign. Six malignant lesions were misinterpreted as benign on MR imaging; thus, suspicious contrast enhancement was present in 96 % of the lesions detected by mammography, US, or clinical examination. Especially 4 of the 17 ductal carcinoma in situ (DCIS) lesions were misinterpreted (23.5 %). Despite optimal technique, 6 malignant lesions were not identified by MR imaging. The highest prevalence of these MR occult lesions was in the group of DCIS. Although MR imaging has an important role in the evaluation of breast lesions and, primarily, in ruling out malignancy, one should be aware of the fact that false-negative MR findings do occur. Received 8 July 1996; Revision received 7 October 1996; Accepted 6 February 1997     

Lesion miss rates and false-negative rates for 1115 consecutive cases of stereotactically guided needle-localized open breast biopsy with long-term follow-up

PURPOSE: To retrospectively determine the lesion miss rate and false-negative rate of needle-localized open breast biopsy (NLOBB) with stereotactic guidance in a large study population. MATERIALS AND METHODS: The ethical review board approved the study; the need for informed consent was waived. A total of 1115 stereotactic NLOBBs performed in 1068 women aged 22-90 years (mean age, 54 years) were tracked to determine outcomes. In cases of malignancy, NLOBB was considered to be diagnostically successful. The mammographic outcomes in all patients with benign results at NLOBB were tracked for at least 2 years. Cases without such mammographic follow-up were cross-referenced with a tumor registry after at least 54 months. The lesion miss rate was based on all malignant lesions and all lesions with long-term mammographic follow-up. Results from all 1115 NLOBBs were used to report the false-negative rate. RESULTS: Of 1115 NLOBBs, 472 (42%) had malignant results. Mammographic follow-up data were available for 535 (83%) of 643 NLOBBs with benign results. Mammographic follow-up revealed 11 lesions that were missed with NLOBB, of which five were malignant and six benign. Thus, the lesion miss rate with NLOBB was 1.1% (11/[472 + 535]). Among the 643 cases, in 108 of them without mammographic follow-up (17%), cross-referencing with a tumor registry did not reveal missed cases of breast carcinoma. The false-negative rate was therefore 1.0% (5/[472 + 5]). CONCLUSION: On the basis of the results of long-term follow-up, the authors noted a lesion miss rate of 1.1% and a false-negative rate of 1.0%, which indicate that NLOBB with stereotactic guidance is an accurate method for diagnosing breast lesions.     

Contribution of ultrasonography and fine-needle aspiration cytology to the differential diagnosis of palpable solid breast lesions.

PURPOSE: To define the role of ultrasonography (US) and fine-needle aspiration biopsy (FNAB) relative to mammography in differentiating between benign and malignant palpable solid breast lesions, and to assess the contribution of FNAB cytology to the delay between referral and the definitive diagnosis of breast cancer. MATERIAL AND METHODS: We retrospectively reviewed the mammograms and US images of 84 palpable breast lesions, 63 of them also blindly. The lesions were classified as benign, indeterminate or malignant by both modalities. The results were compared with histologic diagnoses. The cytologic reports of 57 lesions were reviewed and compared to the final diagnoses. The delay from referral until diagnosis was calculated for each case. RESULTS: Eighty-one of the 84 lesions (96%) were visible as a local abnormality at US. Fifty-two of the 53 cancers were seen as a tumor (n=51) or an architectural distortion (n=1). In the blinded analysis, the sensitivity of FNAB cytology was 92%, specificity 83%, and overall accuracy 88%. There were no false-negative malignancies in the three modalities combined. The delay until the date of the final diagnosis was shorter in the group with a cytologic diagnosis positive for malignancy. CONCLUSION: Malignancy is unlikely if the US, mammographic and cytologic findings of a palpable breast lesion are all benign. Active and critical use of these three modalities could cut down the number of surgical biopsies of benign breast lesions and provide prompt surgical treatment for malignant lesions.     

US of mammographically detected clustered microcalcifications

PURPOSE: To determine whether ultrasonography (US) can depict breast masses associated with mammographically detected clustered microcalcifications and whether the visibility at US is different between benign and malignant lesions. MATERIALS AND METHODS: Ninety-four patients with 100 mammographically detected microcalcification clusters prospectively underwent US with a 10- or 12-MHz transducer before mammographically guided presurgical hook-wire localization. The visibility of breast masses at US was correlated with histologic and mammographic findings. RESULTS: Surgical biopsy revealed 62 benign lesions, 30 intraductal cancers, and eight invasive cancers. At US, breast masses associated with microcalcifications were seen in 45 (45%) of 100 cases. US depicted more breast masses associated with malignant (31 [82%] of 38) than with benign (14 [23%] of 62) microcalcifications (P: <.001). In malignant microcalcification clusters larger than 10 mm, US depicted associated breast masses in all 25 cases. There was no statistically significant difference in shape and distribution of calcific particles, as well as in breast composition, at mammography between US visible and invisible groups. CONCLUSION: Given a known mammographic location, US with a high-frequency transducer can depict breast masses associated with malignant microcalcifications, particularly clusters larger than 10 mm. US can be used to visualize large clusters of microcalcifications that have a very high suspicion of malignancy.     

Clustering of breast microcalcifications: revisited

AIM: To verify the diagnostic value of the traditional definition of 'clustering' of microcalcifications (more than five in the area of 1 cm(2)or 1 cm(3)) on mammography in the differential diagnosis of benign and malignant breast disease.METHODS AND MATERIALS: Three radiologists without knowledge of the final pathology retrospectively counted the number of microcalcifications per 0.25 cm(2) (0.5 x 0.5 cm) unit area on mammography in 57 pathologically proven non-palpable lesions including 26 cancers and 31 benign diseases. Pleomorphism of the microcalcifications, associated architectural distortion or mass or increased density and distribution of microcalcifications were also evaluated.RESULTS: The mean numbers of microcalcifications per 0.25 cm(2) were 16.4 in malignant and 16.7 in benign diseases (no statistically significant difference between the two groups). Pleomorphism of the microcalcifications, associated architectural distortion or mass or increased density were, however, important determining parameters. Clustering was more frequently observed in benign diseases. CONCLUSION: In this study, the mean number of microcalcifications per unit area is much larger than the traditional definition of 'clustering' and clustering itself is not effective in the differential diagnosis of benign and malignant breast lesions. Imaging features other than numbers of calcification per unit area are more important in assessing the significance of mammographic clustered microcalcifications.     

Tubular carcinoma of the breast: mammographic and sonographic features

OBJECTIVE: The purpose of this study was to define specific mammographic and sonographic features of tubular carcinoma of the breast. MATERIALS AND METHODS: Seventeen pathologically confirmed cases of tubular carcinoma were characterized retrospectively by two radiologists. Mammograms and sonograms were available for all patients. RESULTS: Fifteen of the 17 tubular carcinomas appeared as irregularly shaped masses with spiculated margins on mammography. Sixteen of the 17 masses had central densities. Spicules longer than the diameter of the central lesion were noted in eight (53%) of 15 tubular carcinomas. Eight tubular carcinomas had associated calcifications, with calcifications suspected of being malignant in four cases. On sonography, 15 hypoechoic masses were seen. The margins of the masses on sonography were described as ill-defined in 14 (93%) of the 15 cases. Posterior acoustic shadowing was present in 14 of the 15 cases. CONCLUSION: Tubular carcinomas are usually seen on mammography as irregularly shaped masses with central densities and spiculated margins, and most tubular carcinomas can be identified on sonography as hypoechoic masses with ill-defined margins and posterior acoustic shadowing. Although the mammographic and sonographic features of tubular carcinoma are not sufficiently specific to differentiate tubular carcinomas from radial scars, sonography can be useful for guiding biopsies and evaluating for multifocal and multicentric disease.     

Epidermal cyst of the breast: detection in a screening programme

Two cases of epidermal cyst of the breast, a rare benign condition, were detected during a 3-year period in a mammographic screening programme, from 57,954 screening examinations. It is not uncommon for epidermal cysts to be initially misdiagnosed. The mammographic, ultrasound and histological features are presented. It is recommended that these lesions be resected because they possibly have malignant potential.     

乳腺良恶性肿块的钼靶X线征象分析

目的提高对乳腺疾病的钼靶X线征象认识。材料和方法良性病变60例(69个病灶),恶性病变51例(55个病灶),年龄13～72岁。患者均行乳腺钼靶X线摄影并经手术病理证实。结果良性病灶多为圆形、卵圆形,边缘规则,等密度、低密度或混合密度,80％的病灶可见完全或不完全的晕环。恶性病灶多为分叶形或不规则形,80%为高密度,20%为等密度,没有低密度和混合密度病灶。病灶边缘模糊,61％伴有钙化,39％患侧血管增粗扭曲紊乱。结论钼靶乳腺摄影是乳腺影象学检查的首选方法,95%以上的病变可望术前作出正确诊断。