False-negative ultrasound-guided vacuum-assisted biopsy of the breast: difference with US-detected and MRI-detected lesions

A validation of an entropy-based artificial intelligence for ultrasound data in breast tumors

BACKGROUND

The application of artificial intelligence (AI) in the ultrasound (US) diagnosis of breast cancer (BCa) is increasingly prevalent. However, the impact of US-probe frequencies on the diagnostic efficacy of AI models has not been clearly established.

OBJECTIVES

To explore the impact of using US-video of variable frequencies on the diagnostic efficacy of AI in breast US screening.

METHODS

This study utilized different frequency US-probes (L14: frequency range: 3.0-14.0 MHz, central frequency 9 MHz, L9: frequency range: 2.5-9.0 MHz, central frequency 6.5 MHz and L13: frequency range: 3.6-13.5 MHz, central frequency 8 MHz, L7: frequency range: 3-7 MHz, central frequency 4.0 MHz, linear arrays) to collect breast-video and applied an entropy-based deep learning approach for evaluation. We analyzed the average two-dimensional image entropy (2-DIE) of these videos and the performance of AI models in processing videos from these different frequencies to assess how probe frequency affects AI diagnostic performance.

RESULTS

The study found that in testing set 1, L9 was higher than L14 in average 2-DIE; in testing set 2, L13 was higher in average 2-DIE than L7. The diagnostic efficacy of US-data, utilized in AI model analysis, varied across different frequencies (AUC: L9 > L14: 0.849 vs. 0.784; L13 > L7: 0.920 vs. 0.887).

CONCLUSION

This study indicate that US-data acquired using probes with varying frequencies exhibit diverse average 2-DIE values, and datasets characterized by higher average 2-DIE demonstrate enhanced diagnostic outcomes in AI-driven BCa diagnosis. Unlike other studies, our research emphasizes the importance of US-probe frequency selection on AI model diagnostic performance, rather than focusing solely on the AI algorithms themselves. These insights offer a new perspective for early BCa screening and diagnosis and are of significant for future choices of US equipment and optimization of AI algorithms.

The effect of image resolution on convolutional neural networks in breast ultrasound

Purpose

The objective of this research was to investigate the efficacy of various parameter combinations of Convolutional Neural Networks (CNNs) models, namely MobileNet and DenseNet121, and different input image resolutions (REZs) ranging from 64×64 to 512×512 pixels, for diagnosing breast cancer.

Materials and methods

During the period of June 2015 to November 2020, two hospitals were involved in the collection of two-dimensional ultrasound breast images for this retrospective multicenter study. The diagnostic performance of the computer models MobileNet and DenseNet 121 was compared at different resolutions.

Results

The results showed that MobileNet had the best breast cancer diagnosis performance at 320×320pixel REZ and DenseNet121 had the best breast cancer diagnosis performance at 448×448pixel REZ.

Conclusion

Our study reveals a significant correlation between image resolution and breast cancer diagnosis accuracy. Through the comparison of MobileNet and DenseNet121, it is highlighted that lightweight neural networks (LW-CNNs) can achieve model performance similar to or even slightly better than large neural networks models (HW-CNNs) in ultrasound images, and LW-CNNs' prediction time per image is lower.

Current status of MRI-guided vacuum-assisted breast biopsy in Japan

In April 2018, the national health insurance coverage of MRI-guided vacuum-assisted breast biopsy (VAB) was instituted with the application of the Japan Breast Cancer Society. Although MRI-guided VAB has been considered as a special procedure for a long time, having an access to this procedure should be recommended for facilities performing breast MRI as in Western countries. From now on, relevant societies should make efforts in data collection and quality control of MRI-guided VAB in Japan. We must avoid the following. To delay the early diagnosis of breast cancer in the judgment of an inaccurate indication, perform unnecessary biopsy due to overestimation of diagnosis, and reduce the success rate of MRI-guided VAB with immature techniques. This review explains the current status of MRI-guided VAB in Japan and shares procedure and biopsy data as a future reference from an experienced facility.

The clinical impact of MRI screening for BRCA mutation carriers: the first report in Japan

BACKGROUND

There is no consensus on the appropriate surveillance for high-risk women with breast cancer in Japan. We investigated their imaging features and pathological characteristics to build a proper surveillance system for asymptomatic high-risk individuals in the future.

METHODS

We retrospectively reviewed 93 female (median age 43 years) BRCA1 and BRCA2 mutation carriers from our institutional clinical database from 2011 to 2017. The study population was composed of 112 breast cancers. Mammography and MRI were reviewed by examiners blinded to patients' clinical history. Final surgical or biopsy histopathology served as the reference standard in all the patients.

RESULTS

Fifty-nine breast cancers met selection criteria; of these, 30 were BRCA1-associated tumors, and 29 were BRCA2-associated tumors. Invasive ductal carcinoma was the most prevalent type in both BRCA1 and BRCA2. There were statistically significant differences in phenotype, nuclear grade, and Ki-67 labeling index between BRCA1 and BRCA2 mutation carriers. Additionally, imaging findings on mammography and MRI were statistically different. Tumors in BRCA2 carriers demonstrated mammographic calcifications more frequently, while those in BRCA1 carriers demonstrated a mass or architectural distortion (P < 0.001). Enhancement pattern on MRI also significantly differed between the two subgroups (P = 0.006). The size of MRI-detected lesions was statistically smaller than the size of those detected by other modalities (P = 0.004).

CONCLUSIONS

The imaging and histological characteristics of BRCA1/2 mutation carriers were consistent with other countries' studies. MRI-detected lesions were significantly smaller than lesions detected by non-MRI modality. All lesions in BRCA1 mutation carriers could be detected by MRI.

Differentiating vacuum-assisted breast biopsy from core needle biopsy: Is it necessary?

Needle biopsy has replaced excisional biopsy as a definitive diagnostic technique for breast tumours, although excisional biopsy is still used for complete tumour removal for therapeutic and/or diagnostic purposes. Many vacuum-assisted breast biopsy (VAB) systems have been made available by several manufacturers since the release of the Mammotome (MMT) by Johnson & Johnson in 1995. Several recent discussions have been conducted to identify whether core needle biopsy (CNB) or VAB, is more appropriate. However, currently available VAB systems differ from the conventional system (i.e. articulate arm type 11-gauge(G) MMT), and the characteristics of both CNB and VAB have been improved. In CNB, a 14-G needle is frequently used to obtain a larger sample. By contrast, VAB is considered easier to perform because it uses a thinner needle and a lighter, non-tethered system. When differentiating CNB from VAB, the type of VAB should also be defined. In this review, we discuss the characteristics of ultrasonography-guided VAB and CNB with a focus on practical issues such as the number of samples and volume of tissue obtained during ultrasonography-guided needle biopsy.

Ultrasound Imaging Technologies for Breast Cancer Detection and Management: A Review

Ultrasound imaging is a commonly used modality for breast cancer detection and diagnosis. In this review, we summarize ultrasound imaging technologies and their clinical applications for the management of breast cancer patients. The technologies include ultrasound elastography, contrast-enhanced ultrasound, 3-D ultrasound, automatic breast ultrasound and computer-aided detection of breast ultrasound. We summarize the study results seen in the literature and discuss their future directions. We also provide a review of ultrasound-guided, breast biopsy and the fusion of ultrasound with other imaging modalities, especially magnetic resonance imaging (MRI). For comparison, we also discuss the diagnostic performance of mammography, MRI, positron emission tomography and computed tomography for breast cancer diagnosis at the end of this review. New ultrasound imaging techniques, ultrasound-guided biopsy and the fusion of ultrasound with other modalities provide important tools for the management of breast patients.

MRI vacuum-assisted breast biopsies

The indications, technique, results and limitations of MRI vacuum-assisted breast biopsies are discussed from a review of the literature. This was initially a home-grown technique and its development was slowed down by several factors. As a result of major technical advances, it has become a reliable and very consistent procedure with a low rate of underestimation. It is now an undisputed technique when suspicious MRI enhancement is seen with no corresponding mammography or ultrasound features.

Breast ultrasonography: state of the art

Ultrasonography (US) is an indispensable tool in breast imaging and is complementary to both mammography and magnetic resonance (MR) imaging of the breast. Advances in US technology allow confident characterization of not only benign cysts but also benign and malignant solid masses. Knowledge and understanding of current and emerging US technology, along with the application of meticulous scanning technique, is imperative for image optimization and diagnosis. The ability to synthesize breast US findings with multiple imaging modalities and clinical information is also necessary to ensure the best patient care. US is routinely used to guide breast biopsies and is also emerging as a supplemental screening tool in women with dense breasts and a negative mammogram. This review provides a summary of current state-of-the-art US technology, including elastography, and applications of US in clinical practice as an adjuvant technique to mammography, MR imaging, and the clinical breast examination. The use of breast US for screening, preoperative staging for breast cancer, and breast intervention will also be discussed.

MRI-guided vacuum-assisted breast biopsy: comparison with stereotactically guided and ultrasound-guided techniques

Objectives

To analyse the development of MRI-guided vacuum-assisted biopsy (VAB) in Switzerland and to compare the procedure with stereotactically guided and ultrasound-guided VAB.

Methods

We performed a retrospective analysis of VABs between 2009 and 2011. A total of 9,113 VABs were performed. Of these, 557 were MRI guided.

Results

MRI-guided VAB showed the highest growth rate (97 %) of all three procedures. The technical success rates for MRI-guided, stereotactically guided and ultrasound-guided VAB were 98.4 % (548/557), 99.1 % (5,904/5,960) and 99.6 % (2,585/2,596), respectively. There were no significant differences (P = 0.12) between the MRI-guided and the stereotactically guided procedures. The technical success rate for ultrasound-guided VAB was significantly higher than that for MRI-guided VAB (P < 0.001). There were no complications using MRI-guided VAB requiring open surgery. The malignancy diagnosis rate for MRI-guided VAB was similar to that for stereotactically guided VAB (P = 0.35).

Conclusion

MRI-guided VAB is a safe and accurate procedure that provides insight into clinical breast findings.

Key points

• Three vacuum-assisted breast biopsy (VAB) procedures were compared.

• Technical success rates were high for all three VAB procedures.

• Medical complications were relatively low using all three VAB procedures.

• The use of MRI-guided vacuum-assisted breast biopsy is growing.

Magnetic Resonance Imaging-Guided Breast Biopsy in Lesions Not Visualized by Mammogram or Ultrasound

The use of magnetic resonance imaging (MRI) for the diagnosis and evaluation of breast lesions is still in evolution. The aim of this study was to evaluate the outcome of MR-guided biopsy for suspicious lesions seen on MRI but not visualized by mammography or ultrasound. A retrospective review was conducted on all patients undergoing MRI-guided core needle biopsy at a National Cancer Institute-designated comprehensive cancer center. Biopsies were performed using a 1.5-Tesla MR with a seven-channel breast coil taking six cores in a clock face configuration with a 10-gauge vacuum-assisted VACORA biopsy device. One hundred twenty-seven patients underwent 187 biopsies without major complication. The lesion size ranged from 4 to 12 mm. Pathology on MRI-guided core biopsy material revealed 126 (68%) benign lesions. Of these, 12 were intraductal papillomas and two were fibroadenomas. Sixty-one (32%) were deemed clinically significant and included the following pathology: invasive cancer 19 patients (10%), ductal carcinoma in situ (DCIS) in 25 patients (13%), atypical ductal hyperplasia (ADH) 15 patients (8%), and lobular carcinoma in situ in one patient (1%). There were two specimens upgraded from ADH to DCIS and one specimen that was biopsied was called ADH but there was no residual atypia on final pathology. With a median follow-up of 24 months, there were no patients with benign pathology returning with a clinically significant lesion later. MRI-guided biopsy provides an accurate and safe method for sampling suspicious lesions when no other reasonable means of biopsy is available. MRI-guided biopsy yielded clinically significant findings in approximately one-third of the sampled specimens. This technique can provide a good representative sample of clinically significant pathology. MRI-guided biopsy is a relatively new modality, which can provide a nonsurgical diagnostic specimen in the absence of mammographic or ultrasound findings.

Ultrasound-guided diagnostic breast biopsy methodology: retrospective comparison of the 8-gauge vacuum-assisted biopsy approach versus the spring-loaded 14-gauge core biopsy approach

Background

Ultrasound-guided diagnostic breast biopsy technology represents the current standard of care for the evaluation of indeterminate and suspicious lesions seen on diagnostic breast ultrasound. Yet, there remains much debate as to which particular method of ultrasound-guided diagnostic breast biopsy provides the most accurate and optimal diagnostic information. The aim of the current study was to compare and contrast the 8-gauge vacuum-assisted biopsy approach and the spring-loaded 14-gauge core biopsy approach.

Methods

A retrospective analysis was done of all ultrasound-guided diagnostic breast biopsy procedures performed by either the 8-gauge vacuum-assisted biopsy approach or the spring-loaded 14-gauge core biopsy approach by a single surgeon from July 2001 through June 2009.

Results

Among 1443 ultrasound-guided diagnostic breast biopsy procedures performed, 724 (50.2%) were by the 8-gauge vacuum-assisted biopsy technique and 719 (49.8%) were by the spring-loaded 14-gauge core biopsy technique. The total number of false negative cases (i.e., benign findings instead of invasive breast carcinoma) was significantly greater (P = 0.008) in the spring-loaded 14-gauge core biopsy group (8/681, 1.2%) as compared to in the 8-gauge vacuum-assisted biopsy group (0/652, 0%), with an overall false negative rate of 2.1% (8/386) for the spring-loaded 14-gauge core biopsy group as compared to 0% (0/148) for the 8-gauge vacuum-assisted biopsy group. Significantly more (P < 0.001) patients in the spring-loaded 14-gauge core biopsy group (81/719, 11.3%) than in the 8-gauge vacuum-assisted biopsy group (18/724, 2.5%) were recommended for further diagnostic surgical removal of additional tissue from the same anatomical site of the affected breast in an immediate fashion for indeterminate/inconclusive findings seen on the original ultrasound-guided diagnostic breast biopsy procedure. Significantly more (P < 0.001) patients in the spring-loaded 14-gauge core biopsy group (54/719, 7.5%) than in the 8-gauge vacuum-assisted biopsy group (9/724, 1.2%) personally requested further diagnostic surgical removal of additional tissue from the same anatomical site of the affected breast in an immediate fashion for a benign finding seen on the original ultrasound-guided diagnostic breast biopsy procedure.

Conclusions

In appropriately selected cases, the 8-gauge vacuum-assisted biopsy approach appears to be advantageous to the spring-loaded 14-gauge core biopsy approach for providing the most accurate and optimal diagnostic information.

Magnetic resonance-guided vacuum-assisted breast biopsy: results in 100 Japanese women

PURPOSE

The aim of this study was to clarify the frequency of malignancy and the histopathological characteristics of the lesions in patients undergoing magnetic resonance imaging (MRI)-guided vacuum-assisted biopsy (VAB).

MATERIALS AND METHODS

A retrospective review of 100 consecutive patients with 102 lesions who had undergone MRI-guided VAB was performed. The biopsies were performed on a 1.5-T MR scanner using a commercially available biopsy system. None of the lesions seen with MRI could be detected by mammography or second-look ultrasonography.

RESULTS

The average lesion sizes of the focus, mass, and nonmass lesions before the biopsy were 4.5, 8.2, and 21 mm, respectively. Twelve patients (12%) had lesions located in the deep portion of the breast, close to the pectoral muscle. The biopsy was successfully performed without important side effects in all patients. Histopathological findings were invasive ductal carcinoma in 6 (6%), in situ carcinoma in 28 (27%), and high-risk and benign in 68 (67%). Two high-risk lesions were upgraded to ductal carcinoma in situ (DCIS), and three DCIS lesions were upgraded to invasive ductal carcinoma at surgical excision.

CONCLUSION

The high rate of DCIS might be a unique feature among Japanese women. However, MRI-guided VAB is necessary for MRI-only visible suspicious lesions in Japan.

Screen-detected malignant breast lesions diagnosed following benign (B2) or normal (B1) needle core biopsy diagnoses

Breast needle core biopsy (NCB) is now a standard diagnostic procedure in the triple assessment of screen detected breast lesions. However, unlike fine needle aspiration (FNA) cytology, information on the miss rate including false-negative diagnoses (FN) of malignancy (benign 'B2' or normal 'B1' NCB with a malignant outcome) is limited.

METHODS

A large series of NCBs (121,742) performed over an 8-year period has been studied to assess the frequency and causes of missing a malignant diagnosis on NCB and to evaluate their impact on patients' management in the screening service.

RESULTS

During the period of this study, 50,691 were diagnosed as B2 and 9599 were diagnosed as B1. Of those, 779 B2 and 919 B1 were diagnosed as malignant on the subsequent surgical specimens, respectively, giving a FN rate of 3.0%. However when year of diagnosis was taken into consideration, we found that during the period 1999-2001, the FN rate for B2 was 2.7% while the miss rate for B1 was 4.0%. This showed marked improvement over time to reach a figure of 0.5% and 0.5% for B2 and B1, respectively, during the period 2005-2007. On detailed review of cases from a single screening region diagnosed during the last 3 years (2005-2008), 14 cases (0.17% of all NCBs) with malignant surgery were diagnosed as B2 (seven cases; FN rate 0.19%) and B1 (seven cases; B1 biopsy rate from cancer 0.19%). In these cases, NCB was unsatisfactory, there was a discrepancy between radiological abnormalities and histological findings with recommendation for excision or suspicious/malignant cytological diagnosis on concurrent FNA material. Therefore, our results indicate that the malignancy miss rate on NCB is rare and FN NCB diagnoses had no impact on patient management.