Screening mammography performance according to breast density: a comparison between radiologists versus standalone intelligence detection.

BACKGROUND: Artificial intelligence (AI) algorithms for the independent assessment of screening mammograms have not been well established in a large screening cohort of Asian women. We compared the performance of screening digital mammography considering breast density, between radiologists and AI standalone detection among Korean women. METHODS: We retrospectively included 89,855 Korean women who underwent their initial screening digital mammography from 2009 to 2020. Breast cancer within 12 months of the screening mammography was the reference standard, according to the National Cancer Registry. Lunit software was used to determine the probability of malignancy scores, with a cutoff of 10% for breast cancer detection. The AI's performance was compared with that of the final Breast Imaging Reporting and Data System category, as recorded by breast radiologists. Breast density was classified into four categories (A-D) based on the radiologist and AI-based assessments. The performance metrics (cancer detection rate [CDR], sensitivity, specificity, positive predictive value [PPV], recall rate, and area under the receiver operating characteristic curve [AUC]) were compared across breast density categories. RESULTS: Mean participant age was 43.5 ± 8.7 years; 143 breast cancer cases were identified within 12 months. The CDRs (1.1/1000 examination) and sensitivity values showed no significant differences between radiologist and AI-based results (69.9% [95% confidence interval [CI], 61.7-77.3] vs. 67.1% [95% CI, 58.8-74.8]). However, the AI algorithm showed better specificity (93.0% [95% CI, 92.9-93.2] vs. 77.6% [95% CI, 61.7-77.9]), PPV (1.5% [95% CI, 1.2-1.9] vs. 0.5% [95% CI, 0.4-0.6]), recall rate (7.1% [95% CI, 6.9-7.2] vs. 22.5% [95% CI, 22.2-22.7]), and AUC values (0.8 [95% CI, 0.76-0.84] vs. 0.74 [95% CI, 0.7-0.78]) (all P < 0.05). Radiologist and AI-based results showed the best performance in the non-dense category; the CDR and sensitivity were higher for radiologists in the heterogeneously dense category (P = 0.059). However, the specificity, PPV, and recall rate consistently favored AI-based results across all categories, including the extremely dense category. CONCLUSIONS: AI-based software showed slightly lower sensitivity, although the difference was not statistically significant. However, it outperformed radiologists in recall rate, specificity, PPV, and AUC, with disparities most prominent in extremely dense breast tissue.

Correlation of shear-wave elastography stiffness and apparent diffusion coefficient values with tumor characteristics in breast cancer.

We aimed to investigate the correlation between shear-wave elastography (SWE) and apparent diffusion coefficient (ADC) values in breast cancer and to identify the associated characteristics. We included 91 breast cancer patients who underwent SWE and breast MRI prior to surgery between January 2016 and November 2017. We measured the lesion's mean (Emean) and maximum (Emax) elasticities of SWE and ADC values. We evaluated the correlation between SWE, ADC values and tumor size. The mean SWE and ADC values were compared for categorical variable of the pathological/imaging characteristics. ADC values showed negative correlation with Emean (r = - 0.315, p = 0.002) and Emax (r = - 0.326, p = 0.002). SWE was positively correlated with tumor size (r = 0.343-0.366, p < 0.001). A higher SWE value indicated a tendency towards a higher T stage (p < 0.001). Triple-negative breast cancer showed the highest SWE values (p = 0.02). SWE were significantly higher in breast cancers with posterior enhancement, vascularity, and washout kinetics (p < 0.02). SWE stiffness and ADC values were negatively correlated in breast cancer. SWE values correlated significantly with tumor size, and were higher in triple-negative subtype and aggressive imaging characteristics.

Diagnostic Performance of Artificial Intelligence-Based Computer-Aided Detection Software for Automated Breast Ultrasound.

RATIONALE AND OBJECTIVES: This study aimed to evaluate the diagnostic performance of radiologists following the utilization of artificial intelligence (AI)-based computer-aided detection software (CAD) in detecting suspicious lesions in automated breast ultrasounds (ABUS). MATERIALS AND METHODS: ABUS-detected 262 breast lesions (histopathological verification; January 2020 to December 2022) were included. Two radiologists reviewed the images and assigned a Breast Imaging Reporting and Data System (BI-RADS) category. ABUS images were classified as positive or negative using AI-CAD. The BI-RADS category was readjusted in four ways: the radiologists modified the BI-RADS category using the AI results (AI-aided 1), upgraded or downgraded based on AI results (AI-aided 2), only upgraded for positive results (AI-aided 3), or only downgraded for negative results (AI-aided 4). The AI-aided diagnostic performances were compared to radiologists. The AI-CAD-positive and AI-CAD-negative cancer characteristics were compared. RESULTS: For 262 lesions (145 malignant and 117 benign) in 231 women (mean age, 52.2 years), the area under the receiver operator characteristic curve (AUC) of radiologists was 0.870 (95% confidence interval [CI], 0.832-0.908). The AUC significantly improved to 0.919 (95% CI, 0.890-0.947; P = 0.001) using AI-aided 1, whereas it improved without significance to 0.884 (95% CI, 0.844-0.923), 0.890 (95% CI, 0.852-0.929), and 0.890 (95% CI, 0.853-0.928) using AI-aided 2, 3, and 4, respectively. AI-CAD-negative cancers were smaller, less frequently exhibited retraction phenomenon, and had lower BI-RADS category. Among nonmass lesions, AI-CAD-negative cancers showed no posterior shadowing. CONCLUSION: AI-CAD implementation significantly improved the radiologists' diagnostic performance and may serve as a valuable diagnostic tool.

Mammography-Based Deep Learning Model for Coronary Artery Calcification.

AIMS: Mammography, commonly used for breast cancer screening in women, can also predict cardiovascular disease. We developed mammography-based deep learning models for predicting coronary artery calcium (CAC) scores, an established predictor of coronary events. METHODS AND RESULTS: We evaluated a subset of Korean adults who underwent image mammography and CAC computed tomography and randomly selected approximately 80% of the participants as the training dataset, used to develop a convolutional neural network (CNN) to predict detectable CAC. The sensitivity, specificity, area under the receiver operating characteristic curve (AUROC), and overall accuracy of the model's performance were evaluated. The training and validation datasets included 5235 and 1208 women, respectively, (mean age, 52.6 [±10.2] years), including non-zero cases (46.8%). The CNN-based deep learning prediction model based on the Resnet18 model showed the best performance. The model was further improved using contrastive learning strategies based on positive and negative samples: sensitivity, 0.764 (95% CI, 0.667-0.830); specificity, 0.652 (95% CI, 0.614-0.710); AUROC, 0.761 (95% CI, 0.742-0.780); and accuracy, 70.8% (95% CI, 68.8-72.4). Moreover, including age and menopausal status in the model further improved its performance (AUROC, 0.776; 95% CI, 0.762-0.790). The Framingham risk score yielded an AUROC of 0.736 (95% CI, 0.712-0.761). CONCLUSIONS: Mammography-based deep learning models showed promising results for predicting CAC, performing comparably to conventional risk models. This indicates mammography's potential for dual-risk assessment in breast cancer and cardiovascular disease. Further research is necessary to validate these findings in diverse populations, with a particular focus on representation from national breast screening programs.

Diagnostic performance of screening mammography according to menstrual cycle among Asian women.

PURPOSE: To investigate the performance metrics of screening mammography according to menstrual cycle week in premenopausal Asian women. METHODS: This retrospective study included 69,556 premenopausal Asian women who underwent their first screening mammography between 2011 and 2019. The presence or absence of a breast cancer diagnosis within 12 months after the index screening mammography served as the reference standard, determined by linking the study data to the national cancer registry data. Menstrual cycles were calculated, and participants were assigned to groups according to weeks 1-4. The performance metrics included cancer detection rate (CDR), sensitivity, specificity, and positive predictive value (PPV), with comparisons across menstrual cycles. RESULTS: Among menstrual cycles, the lowest CDR at 4.7 per 1000 women (95% confidence interval [CI], 3.8-5.8 per 1000 women) was observed in week 4 (all P < 0.05). The highest sensitivity of 72.7% (95% CI, 61.4-82.3) was observed in week 1, although the results failed to reach statistical significance. The highest specificity of 80.4% (95% CI, 79.5-81.3%) was observed in week 1 (P = 0.01). The lowest PPV of 2.2% (95% CI, 1.8-2.7) was observed in week 4 (all P < 0.05). CONCLUSION: Screening mammography tended to show a higher performance during week 1 and a lower performance during week 4 of the menstrual cycle among Asian women. These results emphasize the importance of timing recommendations that consider menstrual cycles to optimize the effectiveness of screening mammography for breast cancer detection.

ITC-6102RO, a novel B7-H3 antibody-drug conjugate, exhibits potent therapeutic effects against B7-H3 expressing solid tumors.

BACKGROUND: The B7-H3 protein, encoded by the CD276 gene, is a member of the B7 family of proteins and a transmembrane glycoprotein. It is highly expressed in various solid tumors, such as lung and breast cancer, and has been associated with limited expression in normal tissues and poor clinical outcomes across different malignancies. Additionally, B7-H3 plays a crucial role in anticancer immune responses. Antibody-drug conjugates (ADCs) are a promising therapeutic modality, utilizing antibodies targeting tumor antigens to selectively and effectively deliver potent cytotoxic agents to tumors. METHODS: In this study, we demonstrate the potential of a novel B7-H3-targeting ADC, ITC-6102RO, for B7-H3-targeted therapy. ITC-6102RO was developed and conjugated with dHBD, a soluble derivative of pyrrolobenzodiazepine (PBD), using Ortho Hydroxy-Protected Aryl Sulfate (OHPAS) linkers with high biostability. We assessed the cytotoxicity and internalization of ITC-6102RO in B7-H3 overexpressing cell lines in vitro and evaluated its anticancer efficacy and mode of action in B7-H3 overexpressing cell-derived and patient-derived xenograft models in vivo. RESULTS: ITC-6102RO inhibited cell viability in B7-H3-positive lung and breast cancer cell lines, inducing cell cycle arrest in the S phase, DNA damage, and apoptosis in vitro. The binding activity and selectivity of ITC-6102RO with B7-H3 were comparable to those of the unconjugated anti-B7-H3 antibody. Furthermore, ITC-6102RO proved effective in B7-H3-positive JIMT-1 subcutaneously xenografted mice and exhibited a potent antitumor effect on B7-H3-positive lung cancer patient-derived xenograft (PDX) models. The mode of action, including S phase arrest and DNA damage induced by dHBD, was confirmed in JIMT-1 tumor tissues. CONCLUSIONS: Our preclinical data indicate that ITC-6102RO is a promising therapeutic agent for B7-H3-targeted therapy. Moreover, we anticipate that OHPAS linkers will serve as a valuable platform for developing novel ADCs targeting a wide range of targets.

Targeting phosphomevalonate kinase enhances radiosensitivity via ubiquitination of the replication protein A1 in lung cancer cells.

Radiotherapy (RT) plays an important role in localized lung cancer treatments. Although RT locally targets and controls malignant lesions, RT resistance prevents RT from being an effective treatment for lung cancer. In this study, we identified phosphomevalonate kinase (PMVK) as a novel radiosensitizing target and explored its underlying mechanism. We found that cell viability and survival fraction after RT were significantly decreased by PMVK knockdown in lung cancer cell lines. RT increased apoptosis, DNA damage, and G2/M phase arrest after PMVK knockdown. Also, after PMVK knockdown, radiosensitivity was increased by inhibiting the DNA repair pathway, homologous recombination, via downregulation of replication protein A1 (RPA1). RPA1 downregulation was induced through the ubiquitin-proteasome system. Moreover, a stable shRNA PMVK mouse xenograft model verified the radiosensitizing effects of PMVK in vivo. Furthermore, PMVK expression was increased in lung cancer tissues and significantly correlated with patient survival and recurrence. Our results demonstrate that PMVK knockdown enhances radiosensitivity through an impaired HR repair pathway by RPA1 ubiquitination in lung cancer, suggesting that PMVK knockdown may offer an effective therapeutic strategy to improve the therapeutic efficacy of RT.

Screening Outcomes of Supplemental Automated Breast US in Asian Women with Dense and Nondense Breasts.

Background Automated breast (AB) US effectively depicts mammographically occult breast cancers in Western women. However, few studies have focused on the outcome of supplemental AB US in Asian women who have denser breasts than Western women. Purpose To evaluate the performance of supplemental AB US on mammography-based breast cancer screening in Asian women with dense breasts and those with nondense breasts. Materials and Methods A retrospective database search identified asymptomatic Korean women who underwent digital mammography (DM) and supplemental AB US screening for breast cancer between January 2018 and December 2019. We excluded women without sufficient follow-up, established final diagnosis, or histopathologic results. Performance measures of DM alone and AB US combined with DM (hereafter AB US plus DM) were compared. The primary outcome was cancer detection rate (CDR), and the secondary outcomes were sensitivity and specificity. Subgroup analyses were performed based on mammography density. Results From 2785 screening examinations in 2301 women (mean age, 52 years ± 9 [SD]), 28 cancers were diagnosed (26 screening-detected cancers, two interval cancers). When compared with DM alone, AB US plus DM resulted in a higher CDR of 9.3 per 1000 examinations (95% CI: 7.7, 10.3) versus 6.5 per 1000 examinations (95% CI: 5.2, 7.2; P < .001) and a higher sensitivity of 90.9% (95% CI: 77.3, 100.0) versus 63.6% (95% CI: 40.9, 81.8; P < .001) but a lower specificity of 86.8% (95% CI: 85.2, 88.2) versus 94.6% (95% CI: 93.6, 95.5; P < .001) in women with dense breasts. In women with nondense breasts, AB US plus DM resulted in a higher CDR of 9.5 per 1000 examinations (95% CI: 7.1, 10.6) versus 6.3 per 1000 examinations (95% CI: 3.5, 7.1; P < .001), whereas specificity was lower at 95.2% (95% CI: 93.4, 96.8) versus 97.1% (95% CI: 95.8, 98.4; P < .001). Conclusion In Asian women, the addition of automated breast US to digital mammography showed higher cancer detection rates but lower specificities in both dense and nondense breasts. © RSNA, 2023 Supplemental material is available for this article.

NCK-associated protein 1 regulates metastasis and is a novel prognostic marker for colorectal cancer.

Metastatic colorectal cancer (CRC) remains a substantial problem for mortality and requires screening and early detection efforts to increase survival. Epithelial-mesenchymal transition (EMT) and circulation of tumor cells in the blood play important roles in metastasis. To identify a novel target for metastasis of CRC, we conducted a gene microarray analysis using extracted RNA from the blood of preclinical models. We found that NCK-associated protein 1 (NCKAP1) was significantly increased in the blood RNA of patient-derived xenograft (PDX) models of colon cancer. In the NCKAP1 gene knockdown-induced human colon cancer cell lines HCT116 and HT29, there was a reduced wound healing area and significant inhibition of migration and invasion. As the result of marker screening for cytoskeleton and cellular interactions, CRC treated with siRNA of NCKAP1 exhibited significant induction of CDH1 and phalloidin expression, which indicates enhanced adherent cell junctions and cytoskeleton. In HCT116 cells with a mesenchymal state induced by TGFß1, metastasis was inhibited by NCKAP1 gene knockdown through the inhibition of migration, and there was increased CTNNB1 expression and decreased FN expression. We established metastasis models for colon cancer to liver transition by intrasplenic injection shRNA of NCKAP1-transfected HCT116 cells or by implanting tumor tissue generated with the cells on cecal pouch. In metastasis xenograft models, tumor growth and liver metastasis were markedly reduced. Taken together, these data demonstrate that NCKAP1 is a novel gene regulating EMT that can contribute to developing a diagnostic marker for the progression of metastasis and new therapeutics for metastatic CRC treatment.

Performance analysis of screening mammography in Asian women under 40 years.

BACKGROUND: Screening mammography performance among young women remains uncertain in East Asia, where the proportion of young breast cancer patients is higher than that in Western countries. Thus, we analyzed the performance of screening mammography in women under 40 years in comparison with older age groups. METHODS: This retrospective study comprised 95,431 Asian women with 197,525 screening mammograms. The reference standard was determined by linkage to the national cancer registry data and the 12-month follow-up outcomes after the index mammogram. The performance metrics included sensitivity, specificity, cancer detection rate (CDR), positive predictive value (PPV), recall rate, and areas under the receiver operating characteristic curve (AUCs), with comparisons across age groups (30 s, 40 s, and ≥ 50 s). RESULTS: For young women aged < 40 years, sensitivity and AUC (95% confidence interval [CI]) of screening mammography were 60.4% (50.4-69.7) and 0.73 (0.68-0.77), respectively, with no significant difference compared to women in their 40 s (sensitivity: 64.0% [95% CI: 57.8-69.8], P = 0.52; AUC: 0.75 [95% CI: 0.73-0.78], P = 0.35). The CDR (95% CI) was 0.8 (0.6-1.1) per 1,000 mammograms for young women, poorer than 1.8 (1.6-2.1) per 1,000 mammograms for women in their 40 s (P < 0.001). The PPV and recall rate (95% CI) for young women were 0.6% (0.4-0.7) and 14.9% (14.6-15.1), poorer than 1.4% (1.2-1.6) and 13.3% (13.1-13.5) for women in their 40 s (P < 0.001), respectively. CONCLUSION: The accuracy of screening mammography for young women in their 30 s was not significantly different from that for women in their 40 s, but the cancer detection and recall rates were poorer.

Factors associated with radiologic-pathologic discordance in magnetic resonance imaging after neoadjuvant chemotherapy for breast cancer.

BACKGROUND: Although breast MRI is known to be the best imaging modality for assessing the response after neoadjuvant chemotherapy (NAC), discordance still remains between MRI findings and final pathology findings. PURPOSE: To evaluate imaging and clinicopathologic factors associated with radiologic-pathologic discordance in breast cancer patients after NAC. MATERIAL AND METHODS: This retrospective study included 104 breast cancer patients (mean age: 50.2 years) who underwent breast MRI examinations before and after NAC between June 2015 and December 2019. Radiologic complete response (rCR) was defined as equal or lesser enhancement compared with breast tissue in post-NAC MRI. Pathologic CR (pCR) was defined as absence of invasive cancer in final pathology. Imaging and clinicopathologic factors associated with radiologic-pathologic discordance were analyzed with logistic regression analysis. RESULTS: Overall rCR and pCR rates were 37.5% (39/107) and 40.2% (43/107), respectively. Multivariate analysis revealed that the presence of non-mass enhancement (odds ratio [OR], 3.6; 95% confidence interval [CI], 1.1-11.2; P = 0.03) and multicentric lesions on pre-NAC MRI (OR, 4.2; 95% CI, 1.2-14.9; P = 0.03) were independently associated with radiologic-pathologic discordance. False-positive rate for predicting residual tumor was the most prevalent in HER2-positive cancers (86.7%). CONCLUSION: When determining rCR, the presence of non-mass enhancement and multicentric lesions on pre-NAC MRI, and HER2-positive cancers should be interpreted with caution.

Reproducibility of Automated Breast Ultrasonography and Handheld Ultrasonography for Breast Lesion Size Measurement.

ABSTRACT: The purpose of our study was to evaluate the reproducibility of size measurement of breast lesions using automated breast ultrasonography (ABUS) compared with that with handheld ultrasonography (HHUS). Three breast radiologists performed HHUS and measured the lesions size in 2 different phantoms: lesions with various shape, size, and same stiffness (phantom 1) and lesions with same shape, size, and various stiffness (phantom 2). After 1 month, the same radiologists measured the lesion size of the same breast phantoms in the images obtained using ABUS. We evaluated interobserver variability between 3 radiologists in ABUS and HHUS, and intraobserver variability of radiologists between ABUS and HHUS. Intraclass correlation coefficient (ICC) was used in statistical analysis. The measured size of lesions on HHUS was slightly larger than that on ABUS in both phantom 1 and 2, although not statistically significant (P = 0.314, P = 0.858). There were no significant differences in size measurements between the radiologists' measurements and the reference size in phantom 2 (P = 0.862). The ICCs for the interobserver agreement between the 3 radiologists were 0.98 to 0.99 on ABUS and 0.99 to 1.00 on HHUS, respectively. The ICCs for the intraobserver agreement between ABUS and HHUS were 0.97 to 0.97 in phantom 1 and 0.98 to 0.99 in phantom 2. In conclusion, ABUS showed excellent interobserver and intraobserver agreement with HHUS in measuring size of the lesions, regardless of shape, size, and stiffness. Therefore, ABUS mixed with HHUS can be used reliably in following up breast lesions size.

Impact of Skeletal Muscle Loss and Visceral Obesity Measured Using Serial CT on the Prognosis of Operable Breast Cancers in Asian Patients.

OBJECTIVE: This study aimed to investigate the impact of baseline values and temporal changes in body composition parameters, including skeletal muscle index (SMI) and visceral adipose tissue area (VAT), measured using serial computed tomography (CT) imaging on the prognosis of operable breast cancers in Asian patients. MATERIALS AND METHODS: This study retrospectively included 627 Asian female (mean age ± standard deviation [SD], 53.6 ± 8.3 years) who underwent surgery for stage I-III breast cancer between January 2011 and September 2012. Body composition parameters, including SMI and VAT, were semi-automatically calculated on baseline abdominal CT at the time of diagnosis and follow-up CT for post-treatment surveillance. Serial changes in SMI and VAT were calculated as the delta values. Multivariable Cox regression analysis was used to evaluate the association of baseline and delta SMI and VAT values with disease-free survival. RESULTS: Among 627 patients, 56 patients (9.2%) had breast cancer recurrence after a median of 40.5 months. The mean value ± SD of the baseline SMI and baseline VAT were 43.7 ± 5.8 cm²/m² and 72.0 ± 46.0 cm², respectively. The mean value of the delta SMI was -0.9 cm²/m² and the delta VAT was 0.5 cm². The baseline SMI and VAT were not significantly associated with disease-free survival (adjusted hazard ratio [HR], 0.983; 95% confidence interval [CI], 0.937-1.031; p = 0.475 and adjusted HR, 1.001; 95% CI, 0.995-1.006; p = 0.751, respectively). The delta SMI and VAT were also not significantly associated with disease-free survival (adjusted HR, 0.894; 95% CI, 0.766-1.043; p = 0.155 and adjusted HR, 1.001; 95% CI, 0.989-1.014; p = 0.848, respectively). CONCLUSION: Our study revealed that baseline and early temporal changes in SMI and VAT were not independent prognostic factors regarding disease-free survival in Asian patients undergoing surgery for breast cancer.

Diagnosing thyroid nodules with atypia of undetermined significance/follicular lesion of undetermined significance cytology with the deep convolutional neural network.

To compare the diagnostic performances of physicians and a deep convolutional neural network (CNN) predicting malignancy with ultrasonography images of thyroid nodules with atypia of undetermined significance (AUS)/follicular lesion of undetermined significance (FLUS) results on fine-needle aspiration (FNA). This study included 202 patients with 202 nodules ≥ 1 cm AUS/FLUS on FNA, and underwent surgery in one of 3 different institutions. Diagnostic performances were compared between 8 physicians (4 radiologists, 4 endocrinologists) with varying experience levels and CNN, and AUS/FLUS subgroups were analyzed. Interobserver variability was assessed among the 8 physicians. Of the 202 nodules, 158 were AUS, and 44 were FLUS; 86 were benign, and 116 were malignant. The area under the curves (AUCs) of the 8 physicians and CNN were 0.680-0.722 and 0.666, without significant differences (P > 0.05). In the subgroup analysis, the AUCs for the 8 physicians and CNN were 0.657-0.768 and 0.652 for AUS, 0.469-0.674 and 0.622 for FLUS. Interobserver agreements were moderate (k = 0.543), substantial (k = 0.652), and moderate (k = 0.455) among the 8 physicians, 4 radiologists, and 4 endocrinologists. For thyroid nodules with AUS/FLUS cytology, the diagnostic performance of CNN to differentiate malignancy with US images was comparable to that of physicians with variable experience levels.

Breast Cancer Screening with Abbreviated Breast MRI: 3-year Outcome Analysis.

Background Data are limited regarding the performance of abbreviated screening breast MRI during consecutive years and the characteristics of breast cancers missed and detected with it. Purpose To assess the longitudinal diagnostic performance of abbreviated screening MRI and to determine whether the screening outcomes of abbreviated MRI differed between yearly time periods for 3 consecutive years. Materials and Methods This retrospective study included 1975 consecutive women who underwent abbreviated screening MRI between September 2015 and August 2018. Breast Imaging Reporting and Data System (BI-RADS) categories 3-5 defined positive results, and BI-RADS categories 1-2 defined negative results. Cancer detection rate (CDR), sensitivity, specificity, positive predictive value (PPV), abnormal interpretation rate (AIR), and interval cancer rate were assessed annually. Yearly performance measures were compared with the Fisher exact test by using the permutation method. Clinical-pathologic and imaging characteristics of the missed and detected cancers were compared by using the Fisher exact test and the Wilcoxon rank sum test. Results A total of 1975 women (median age, 49 years; interquartile range, 44-56 years) underwent 3037 abbreviated MRI examinations over 3 years. CDR (year 1 to year 3, 6.9-10.7 per 1000 examinations), positive predictive value for recall (9.7% [six of 62] to 15.6% [12 of 77]), positive predictive value for biopsy (31.6% [six of 19] to 63.2% [12 of 19]), sensitivity (75.0% [six of eight] to 80.0% [12 of 15]), and specificity (93.5% [807 of 863] to 94.1% [1041 of 1106]) were highest in year 3, and AIR (7.1% [62 of 871] to 6.9% [77 of 1121]) was lowest in year 3. However, all outcome measures did not differ statistically between years 1, 2, and 3 (all P > .05). The interval cancer rate was 0.66 per 1000 examinations (two of 3037). Thirty-eight breast cancers were identified in 36 women; 29 were detected with abbreviated MRI, but nine were missed. Of these, seven were detected with other imaging modalities after negative results at the last screening MRI examination, and two were interval cancers. All missed cancers were node-negative early-stage invasive cancers and were smaller (median size, 0.8 cm vs 1.2 cm; P = .01) than detected cancers. Conclusion Screening outcome measures of abbreviated MRI were sustained without significant differences between 3 consecutive years. All cancers missed at abbreviated MRI were node-negative invasive cancers and tended to be smaller than detected cancers. © RSNA, 2021 See also the editorial by Lee in this issue. Online supplemental material is available for this article.

Non-mass lesions detected by breast US: stratification of cancer risk for clinical management.

OBJECTIVE: To develop a classification system using imaging features to interpret breast non-mass lesions (NMLs) detected on US and to stratify their cancer risk. METHODS: This retrospective study included 715 patients with 715 breast NMLs detected on breast US from 2012 to 2016. Each patient underwent mammography at the time of diagnosis. Radiologists assessed US and mammographic features and final BI-RADS categories. Multivariable logistic regression was used to find imaging features associated with malignancy in a development dataset (n = 460). A system to classify BI-RADS categories (3 to 5) was developed based on the odds ratios (ORs) of imaging features significantly associated with malignancy and validated in a distinct validation dataset (n = 255). RESULTS: Among 715 NMLs, 385 (53.8%) were benign and 330 (46.2%) were malignant. In the development dataset, the following B-mode US features were associated with malignancy (all p < 0.001): segmental distribution (OR = 3.03; 95% confidence interval [CI], 1.50-6.15), associated calcifications (OR = 4.26; 95% CI, 1.62-11.18), abnormal ductal change (OR = 4.91; 95% CI, 2.07-11.68), and posterior shadowing (OR = 20.20; 95% CI, 6.46-63.23). The following mammographic features were also associated with malignancy (all p < 0.001): calcifications (OR = 7.98; 95% CI, 3.06-20.81) and focal asymmetry (OR = 4.75; 95% CI, 1.90-11.88). In the validation dataset, our classification system using US and mammography showed a higher area under the curve (0.951-0.956) compared to when it was not applied (0.908-0911) to predict malignancy with BI-RADS categories (p < 0.05). CONCLUSION: Our classification system which incorporates US and mammographic features of breast NMLs can help interpret and manage all NMLs detected on breast US by stratifying cancer risk according to BI-RADS categories. KEY POINTS: • When diagnosing breast NMLs detected on US, suspicious US features are segmental distribution, associated abnormal ductal change, calcifications, and posterior shadowing within or around the NML on B-mode US, while a probably benign US feature is the presence of multiple small cysts. • Corresponding suspicious mammographic features of breast NMLs detected on US are associated calcifications and focal asymmetry. • Our classification system which incorporates US features with and without mammography can potentially be used to interpret and manage any NMLs detected on breast US in clinical practice.

Diagnosis of thyroid nodules on ultrasonography by a deep convolutional neural network.

The purpose of this study was to evaluate and compare the diagnostic performances of the deep convolutional neural network (CNN) and expert radiologists for differentiating thyroid nodules on ultrasonography (US), and to validate the results in multicenter data sets. This multicenter retrospective study collected 15,375 US images of thyroid nodules for algorithm development (n = 13,560, Severance Hospital, SH training set), the internal test (n = 634, SH test set), and the external test (n = 781, Samsung Medical Center, SMC set; n = 200, CHA Bundang Medical Center, CBMC set; n = 200, Kyung Hee University Hospital, KUH set). Two individual CNNs and two classification ensembles (CNNE1 and CNNE2) were tested to differentiate malignant and benign thyroid nodules. CNNs demonstrated high area under the curves (AUCs) to diagnose malignant thyroid nodules (0.898-0.937 for the internal test set and 0.821-0.885 for the external test sets). AUC was significantly higher for CNNE2 than radiologists in the SH test set (0.932 vs. 0.840, P < 0.001). AUC was not significantly different between CNNE2 and radiologists in the external test sets (P = 0.113, 0.126, and 0.690). CNN showed diagnostic performances comparable to expert radiologists for differentiating thyroid nodules on US in both the internal and external test sets.

Radiomics Based on Thyroid Ultrasound Can Predict Distant Metastasis of Follicular Thyroid Carcinoma.

We aimed to evaluate whether radiomics analysis based on gray-scale ultrasound (US) can predict distant metastasis of follicular thyroid cancer (FTC). We retrospectively included 35 consecutive FTCs with distant metastases and 134 FTCs without distant metastasis. We extracted a total of 60 radiomics features derived from the first order, shape, gray-level cooccurrence matrix, and gray-level size zone matrix features using US imaging. A radiomics signature was generated using the least absolute shrinkage and selection operator and was used to train a support vector machine (SVM) classifier in five-fold cross-validation. The SVM classifier showed an area under the curve (AUC) of 0.90 on average on the test folds. Age, size, widely invasive histology, extrathyroidal extension, lymph node metastases on pathology, nodule-in-nodule appearance, marked hypoechogenicity, and rim calcification on the US were significantly more frequent among FTCs with distant metastasis compared to those without metastasis (p < 0.05). Radiomics signature and widely invasive histology were significantly associated with distant metastasis on multivariate analysis (p < 0.01 and p = 0.003). The classifier using the results of the multivariate analysis showed an AUC of 0.93. The radiomics signature from thyroid ultrasound is an independent biomarker for noninvasively predicting distant metastasis of FTC.

Pilot-scale subcritical-water extraction of nodakenin and decursin from Angelica gigas Nakai.

Active components were extracted from Angelica gigas Nakai by subcritical-water extraction (SWE) with the purpose of determining how the extraction conditions affect the SWE of antioxidant properties and active components (nodakenin and decursin), and to compare pilot-scale SWE (8 L) and conventional extraction methods. The extraction yields of nodakenin and decursin in the pilot-scale system were highest at 150 °C for 10 min and 190 °C for 15 min, respectively. The extraction yield of decursin increased as the stirring speed was increased to 200-250 rpm. Pearson's correlation indicated that the radical-scavenging activities using DPPH and ABTS assays were more sensitive to the Maillard reaction (R2 = 0.822 and 0.933, respectively) than to the total phenolic contents (R2 = 0.486 and 0.724, respectively). The extraction yield of decursin was higher when using conventional extraction methods than for SWE.

Diagnostic performance of abbreviated breast MRI for screening of women with previously treated breast cancer.

RATIONALE AND OBJECTIVES: To evaluate the diagnostic performance of abbreviated screening breast magnetic resonance imaging (AB-MRI) for screening in women with previously treated breast cancer. MATERIALS AND METHODS: This retrospective study included consecutive AB-MRI from September 2015 to December 2016 in patients with previously treated breast cancer. Longitudinal medical record of patients' demographics, outcomes of imaging surveillance and results of biopsy was reviewed. Protocol consisted of T2-weighted scanning and dynamic contrast-enhanced imaging including one pre-contrast and two post-contrast scans. A positive examination was defined as final assessment of BI-RADS 4 or 5 and negative was defined as BI-RADS 1, 2, or 3. Abnormal interpretation rate, cancer detection rate (CDR), sensitivity, specificity, accuracy, positive predictive value (PPV), and negative predictive value (NPV) were analyzed. RESULTS: Among total 1043 AB-MRI, 29 (2.8%) AB-MRI had suspicious findings including 26 (2.5%) BI-RADS 4 and 3 (0.3%) BI-RADS 5 assessments. CDR was 9.59 per 1000. Performance outcomes were as follows: sensitivity, 71.4%; specificity, 98.2%; accuracy, 97.8%; PPV 1, 35.7%; PPV3 50%; and NPV 99.6%. Four cancers with false negative MRI were all early cancers of <1.0 cm with node negative. One was palpable interval cancer while the others were alternative screening modality-detected asymptomatic cancers before the next MRI screening. CONCLUSION: AB-MRI showed high accuracy and NPV for detecting cancer recurrence in women with previously treated breast cancer. Missed cancers were all minimal cancers with node negative.