Ultrasound Classification Research Articles

Feature Enhancement in Medical Ultrasound Videos Using Contrast-Limited Adaptive Histogram Equalization.

Speckle noise reduction algorithms are extensively used in the field of ultrasound image analysis with the aim of improving image quality and diagnostic accuracy. However, significant speckle filtering induces blurring, and this requires the enhancement of features and fine details. We propose a novel framework for both multiplicative noise suppression and robust contrast enhancement and demonstrate its effectiveness using a wide range of clinical ultrasound scans. Our approach to noise suppression uses a novel algorithm based on a convolutional neural network that is first trained on synthetically modeled ultrasound images and then applied on real ultrasound videos. The feature improvement stage uses an improved contrast-limited adaptive histogram equalization (CLAHE) method for enhancing texture features, contrast, resolvable details, and image structures to which the human visual system is sensitive in ultrasound video frames. The proposed CLAHE algorithm also considers an automatic system for evaluating the grid size using entropy, and three different target distribution functions (uniform, Rayleigh, and exponential), and interpolation techniques (B-spline, cubic, and Lanczos-3). An extensive comparative study has been performed to find the most suitable distribution and interpolation techniques and also the optimal clip limit for ultrasound video feature enhancement after speckle suppression. Subjective assessments by four radiologists and experimental validation using three quality metrics clearly indicate that the proposed framework generates superior performance compared with other well-established methods. The processing pipeline reduces speckle effectively while preserving essential information and enhancing the overall visual quality and therefore could find immediate applications in real-time ultrasound video segmentation and classification algorithms.

Radiological–pathological correlation of the British Thyroid Association ultrasound classification of thyroid nodules: a real-world validation study

To evaluate the real-world performance of the British Thyroid Association (BTA) U classification, specifically focusing on radiology-pathology correlation and to glean learning points. Adults undergoing a neck ultrasound for thyroid nodules were reviewed over a period of 1-year. Data including demographics, nodule characteristics, BTA grading, and cytology/histopathology were retrieved with a minimum 24-month follow-up. Of 1,225 graded nodules in 964 patients, cytology and/or histology were available for 300 (24%). 57 cancers were detected. Of 24 (2%) U5 nodules, 14 were malignant, of 51 (4%) U4, 22 were malignant, of 256 (21%) U3, 20 were malignant, and from 894 (73%) U2 nodules, one cancer was discovered. BTA U grading with fine-needle aspiration (FNA)/core biopsy achieved 96.5% sensitivity, 93.7% specificity, and 93.9% accuracy compared to excision. There was no association between nodule size and rate of malignancy. This is the first study to validate the use of the BTA U-grading system in UK clinical practice. The BTA U-grading system is a robust and reliable method of evaluating the risk of malignancy in thyroid nodules with a high negative predictive value. Key learning points gleaned from the study were accurate assessment of nodule echogenicity, careful evaluation of solid-cystic nodules, optimising ultrasound technique, and the low-risk nature of U3 nodules.

Ultrasound Assessment of Muscle Injury Associated with Closed Limb Fracture.

Objective The aim of this study was to assess muscle injury associated with upper and lower closed limb fracture using ultrasound, and to develop ultrasound classification criteria for muscle injury. Patients and Methods Thirty patients with limb fracture and muscle injury participated in this study. Ultrasonography was used to assess muscle fibre, hematoma, vascular injury, and diameter growth rate. Injury was classified into three grades according to the ultrasound imaging: scores of less than, equal to, or greater than 9. Results Of 30 patients, focal fibre rupture was observed in 11 cases; in 9 cases, the injured area exceeded 30% of the muscle area. Six patients had muscle hematoma (the largest reaching 39 mm); in 4 patients, the hematoma showed a honeycombed pattern. Vascular rupture was observed in 6 patients, of which 2 had decreased main arterial diameter and blood flow. The greatest increase in muscle thickness was 17 mm. Of all patients, 11 showed an increase in the diameter growth rate of the muscle exceeding 50%. In addition, among the 30 patients, 11 patients with scores ranging from 4 to 8 received conservative treatment; 9 patients with scores ranging from 10 to 14 received operative treatment; and 10 patients with scores equal to 9 received either conservative or operative treatment. Conclusions Ultrasonography is useful for diagnosing muscle injury associated with closed limb fracture. The ultrasound classification criteria for muscle injury can be used to assess the severity of injury and guide the decision of treatment.

Rendimiento diagnóstico del sistema de evaluación de riesgo ecográfico del nódulo tiroideo de la American Thyroid Association en endocrinología (estudio ETIEN 3)

IntroductionNo study has been published to date in Spain about the value of the American Thyroid Association (ATA) ultrasound risk assessment of thyroid nodules applied by endocrinologists. ObjectivesTo assess the diagnostic performance of ATA thyroid nodule risk classification applied by endocrinologists with respect to histological results. MethodsA retrospective, observational study of 317 patients (mean age, 51.7±13.7 years; 83.3% women) with thyroid nodules (maximum diameter: 3.2±1.4cm) who underwent US-guided FNA in endocrinology between October 2015 and December 2018, were classified based on the ATA ultrasound risk assessment. Surgery was performed in all of them. Quality criteria of the ultrasound classification were assessed as compared to histological results. ResultsOverall, 61.2% of nodules assessed were classified as benign (n=3) and very low (n=60) or low suspicion (n=131), 11.7% as intermediate suspicion (n=37), and 27.1% as high suspicion (n=86). Benign nodular disease was found in 260 patients, and thyroid cancer in 57 patients. There were 14 incidental papillary microcarcinomas. Classification of thyroid nodules as high suspicion showed a high diagnostic performance to identify malignant nodules (87.7% sensitivity, 86.2% specificity, negative predictive value 97.0%), improving sensitivity (93.0%) and reducing specificity (73.1%) when considering high and intermediate suspicion nodules as a risk for thyroid cancer. ConclusionsApplication by endocrinologists of the ATA ultrasound risk assessment of thyroid nodules shows a high diagnostic performance to identify malignant thyroid nodules before surgery.

Superb microvascular imaging is a rational choice for accurate Bosniak classification of renal cystic masses.

Bosniak classification system provides a fine imaging evaluation for the malignant cystic renal masses. Superb microvascular imaging (SMI) is a new ultrasonic technique which apply advanced clutter suppression to reflect microflow information. The aim of this study was to evaluate the performance of ultrasound Bosniak classification aided by SMI in diagnosis of cystic renal masses. By comparing with contrast-enhanced ultrasound (CEUS) and pathology, we evaluated the sensitivity, specificity and accuracy of conventional ultrasound and SMI combination in the diagnosis of renal cystic masses. Color Doppler Flow image (CDFI) and SMI had significant difference in evaluation the blood flow and Bosniak classification for cystic renal mass of IIf, III and IV (p = 0.020). SMI was consistent with CEUS for cystic renal mass Bosniak classification of IIf, III and IV. Although, CEUS had highest sensitivity of 97.4%, SMI had the highest area under curve in differential diagnosis malignance from benign masses [area under curve = 0.869 (0.743 - 0.995), p < 0.001]. SMI can display better in observing blood flow in septa and solid structures of renal cystic lesions compared with CDFI, thus, it can perform more accurate Bosniak classification for renal cystic mass. SMI and CEUS are consistent in Bosniak classification. Bosniak classification aided by SMI maybe an accurate non-invasive ultrasonic examination in distinguishing benign and malignant renal cystic lesions. Conventional ultrasound had limited ability to distinguish malignance in Bosniak classifications IIf and III. SMI can display better in observing blood flow in septa and solid structures of renal cystic lesions compared with CDFI.

Weakly Supervised Estimation of Shadow Confidence Maps in Fetal Ultrasound Imaging.

Detecting acoustic shadows in ultrasound images is important in many clinical and engineering applications. Real-time feedback of acoustic shadows can guide sonographers to a standardized diagnostic viewing plane with minimal artifacts and can provide additional information for other automatic image analysis algorithms. However, automatically detecting shadow regions using learning-based algorithms is challenging because pixel-wise ground truth annotation of acoustic shadows is subjective and time consuming. In this paper, we propose a weakly supervised method for automatic confidence estimation of acoustic shadow regions. Our method is able to generate a dense shadow-focused confidence map. In our method, a shadow-seg module is built to learn general shadow features for shadow segmentation, based on global image-level annotations as well as a small number of coarse pixel-wise shadow annotations. A transfer function is introduced to extend the obtained binary shadow segmentation to a reference confidence map. In addition, a confidence estimation network is proposed to learn the mapping between input images and the reference confidence maps. This network is able to predict shadow confidence maps directly from input images during inference. We use evaluation metrics such as DICE, inter-class correlation, and so on, to verify the effectiveness of our method. Our method is more consistent than human annotation and outperforms the state-of-the-art quantitatively in shadow segmentation and qualitatively in confidence estimation of shadow regions. Furthermore, we demonstrate the applicability of our method by integrating shadow confidence maps into tasks such as ultrasound image classification, multi-view image fusion, and automated biometric measurements.

OR27-1 Interobserver Agreement in the Assessement of Thyroid Nodules Ultrasound Features: A Blinded Multicenter Study

Background: Single-center trials have demonstrated an acceptable level of inter-observer agreement in the evaluation of the ultrasound (US) features of thyroid nodules, but limited data are available about the consistency in the assessment of US findings among different thyroid centers. Aim of the study: to assess the inter-observer agreement between different thyroid centers and among different specialists in the evaluation of the main US features of thyroid nodules. Materials and methods A blinded retrospective analysis of 100 electronically-recorded US images was conducted in three large-volume thyroid centers by seven qualified thyroid imaging experts, two radiologist and five endocrinologists. The following US features were evaluated: composition (solid, predominantly solid, predominantly cystic, and cystic); echogenicity (hyperechoic, isoechoic, mildly and deeply hypoechoic); margins (well-defined, ill-defined, microlobulated, and spiculated); calcifications (absent, microscopic, macroscopic, eggshell); hyperechoic foci of uncertain significance; comet-tail artifacts; vascularity (no vascular signals, perinodular and/or slight intranodular flow, and marked intranodular flow). Thyroid nodules were also classified according to four major US classification systems: AACE/ACE/AME, EU-TIRADS, ATA and ACR. The inter-observer agreement was calculated using cross-tabulation expressed in Cohen's Kappa. Kappa values were evaluated, according to Landis and Koch, as follows: 0-0.20 poor, 0.21-0.40 fair, 0.41-0.60 moderate, 0.61-0.80 substantial, and 0.81-1.0 almost perfect agreement. A sub-analysis assessed how many times and how many operators evaluated as suspicious for malignancy the US features of each nodule. Results: The inter-observer agreement resulted in a K-correlation coefficient of 34.5%, 44.0%, 42.3% and 38.8% for the ATA, AACE/ACE/AME, ACR, and EU-TIRADS classification systems, respectively. The interobserver agreement for the main thyroid nodule US findings resulted as follows: composition 53.2%; echogenicity 46.9%; margins 33.2%; comet-tail 10.6%, microcalcifications 46.8%; macrocalcifications 37.7%; eggshell calcifications 64.9%; intranodular vascularity 45.9%. Conclusion: The level of agreement among different thyroid centers in the description of suspicious US features in thyroid nodules ranged from fair to moderate, with the lowest level of consistency for the characteristics of margins and the presence of comet-tails. A similar range of variability was demonstrated also for the four US classification systems. An universally accepted lexicon of thyroid US features and a dedicated training in thyroid US findings definition are needed to improve the inter-observer agreement and the predictive value of US classification systems in real world practice.

MON-580 Diagnostic Accuracy of Different Thyroid Ultrasound Classification Systems and the Added Value of Operator Subjective Impression in Stratifying Nodule Risk

Background and aim. Several ultrasound (US) classifications for thyroid nodules have been proposed. Since most of them are hardly applicable in clinical practice, we set up the Modena US Thyroid Classification (MUT) that stratifies the risk of malignancy based on knowledge derived from scientific literature and on clinician subjective impression. The aim of the present study was to test the diagnostic accuracy of different thyroid US classification systems, AACE/ACE-AME, American Thyroid Association (ATA), British Thyroid Association (BTA), and MUT, and to evaluate inter-classification agreement. Methods. We prospectively enrolled 111 patients (33M, 78F; age 19-75) with indeterminate, suspicious or malignant cytology. All the patients underwent neck US before surgery and a score according to MUT was assigned: 1 not certainly nodular; 2 not suspect; 3 indeterminate; 4 suspect; 5 very suspect. Then, we retrospectively classified nodules according to AACE/ACE-AME, ATA and BTA. US pattern was related to hystology. Sensitivity, specificity, diagnostic cut-off value and accuracy of each classification were calculated. The overall agreement between classifications was quantified by Bland-Altman test. The agreement between single nodule analysis by different classifications was evaluated considering Weighted Cohen's Kappa. Results. Fifteen patients had uninodular and 96 multinodular goiter, for a total of 457 nodules. MUT has the highest accuracy (AUC 0.808) and specificity (89%), followed by ATA and BTA, and finally by AACE/ACE-AME. ATA and BTA are highly interchangeable and MUT is comparable to both of them. AACE/ACE-AME is the least interchangeable with all the other classifications. Considering agreement between single nodule analysis by different classifications, ATA and BTA had the best (k=0.723); AACE/ACE-AME showed slight agreement with BTA (k=0.177) and MUT (k=0.183), and fair agreement with ATA (k=0.282); MUT had fair agreement with both ATA (k=0.291) and BTA (k=0.271). Conclusions. Our findings bring out the limit in specificity of the current reference classifications, which improves when the subjective impression of the clinician is considered.

Ultrasound classification and clinical analysis of ovarian pregnancy: A study of 12 cases

ObjectiveTo evaluate the clinical significance of preoperative ultrasound diagnosis and classification of ovarian pregnancy (OP). MethodsThe ultrasonographic reports, medical records, and operative summaries were reviewed for twelve women with a confirmed diagnosis of OP. According to the ultrasonographic appearance, OP was classified into two types. For every type, we analyzed the possibility and clinical significance of preoperative ultrasound diagnosis, combining with its clinical manifestations and treatments. ResultsAccording to sonographic features, twelve cases of OP can be classified into ruptured type (n = 6) and unruptured type (n = 6), the latter was subclassified into embryo sac type (n = 2) and non-homogeneous mass type (n = 4). All the unruptured OP have the characteristic solid hyperechoic rings or masses, and 66.7% (4/6) were correctvly diagnosed by preoperative ultrasound. No characteristic ultrasonogram was detected in ruptured OP which were all diagnosed as ruptured ectopic pregnancy (EP) or corpus luteum by preoperative ultrasound.The diagnostic and surgical procedures of four cases diagnosed by preoperative ultrasound were all laparoscopy and removal of the gestational products, and their average hemoperitoneum, operation time, hospital days was 313 ± 278 ml, 57 ± 9 min, and 4.25 ± 0.5days, respectively. While ruptured OP cases (6 cases diagnosed at first visit and 1 underdiagnosed but ruptured after 6 days), the diagnostic procedures was laparoscopy(3/7) or laparotomy(4/7), the surgical procedure was removal of the gestational products(2/7), wedge resection(3/7), or adenexectomy(2/7), and their average hemoperitoneum, operation time, and hospital days was 1914 ± 1059 ml, 93 ± 17 min, and 5.9 ± 1.3days, respectively. ConclusionsUltrasound plays a significant role in diagnosis of OP, and part of the unruptured OP based on the typical sonographic characteristics could be correctly diagnosed by an experienced sonographer, this is beneficial and effective in terms of the risk of patient's operation and life. Whereas ruptured OP were generally diagnosed as ruptured EP or corpus luteum because there were no characteristic ultrasound manifestations.

&lt;p&gt;A multicenter study of a contrast-enhanced ultrasound diagnostic classification of breast lesions&lt;/p&gt;

PurposeTo evaluate a classification model of contrast-enhanced ultrasound (CEUS) and examine the characteristics of patients with false-negative diagnosis.Patients and methodsA retrospective secondary analysis of a multicenter trial of CEUS for breast cancer diagnosis (from August 2015 to April 2017) was undertaken. Patients (n=1,023) with Breast Imaging Reporting and Data System 4–5 lesions on B-mode ultrasound underwent CEUS. Pathological diagnoses were available from surgical or biopsy specimens for correlation. Lesion maximum diameter (LMD), distance to the papilla (DtP), distance from the superficial edge of the lesion to the skin (DtS), distance from the deep edge of the lesion to the pectoralis muscle (DtPM), and body mass index (BMI) were evaluated.ResultsMedian age and BMI were 48.0 and 41.2 years and 23.2 and 22.4 kg/m2 for patients with malignant and benign lesions, respectively. Overall sensitivity, specificity, and accuracy of CEUS for malignancy were 89.4%, 65.3%, and 75.8%, respectively. The patients with true-positive and false-negative diagnosis (ie, with malignant lesion) were older than those with false-positive and true-negative diagnosis (ie, with benign lesion). Patients with true-positive and false-positive diagnoses had higher BMI than patients with true-negative and false-negative diagnoses (P=0.004). Patients with true-positive and false-negative diagnoses had larger LMD and DtP, as well as smaller DtS and DtPM.ConclusionOlder age, higher BMI, larger LMD and DtP, and smaller DtS and DtPM were associated with malignant lesions on CEUS. Patients with these characteristics should undergo further imaging.

Abstract P6-02-07: Is real-time rescanning by radiologists necessary for all ACR BI-RADS US category 3 to 5 lesions in the diagnostic setting?

Abstract Purpose: To assess the value of radiologists' rescanning before final interpretation for American College of Radiology (ACR) Breast Imaging Reporting and Data System ultrasonography (US) category 3–5 lesions. Materials and Methods: Image data on 1070 patients with 1070 category 3–5 breast lesions with a pathological diagnosis scanned between January 2016–June 2017 were included. Both real-time and static images were acquired for each lesion. The diagnostic performance of assessing static images and rescanning by the radiologist were each calculated. The positive predictive values (PPVs) of each category in the two groups were calculated and compared. Results: The sensitivity, specificity, PPV, and negative predictive value for real-time US were 98.9%, 58.2%, 44.8% and 99.4%, and for static images were 98.9%, 57.1%, 44.1% and 99.3%, respectively. Areas under the curves were not significantly different if final assessment was only dichotomized as negative or positive (≥ category 3: P = .566). Significant differences were observed if more detailed classification was performed (category 3–5 without subcategory: 0.969 vs. 0.955, P = .0113; category 3–5 with subcategory 4A–4C: 0.915 vs. 0.855, P &amp;lt; .0001). All PPVs of each category for each assessment were within the reference range provided by the ACR in 2013 except subcategory 4B of static image evaluation, which was also significantly higher than that of real-time assessment (54.8% vs. 40.7%, P = .037),BI-RADS US of real-time and static images evaluation correlate with pathology are listed in Table 1. The overall agreement of the two approaches was moderate (κ = 0.43 if lesions were assigned to category/subcategory 3, 4A–4C, 5). Table1BI-RADS US category/subcategoryPPV provided by ACRReal-time US n (%)Static image US n (%)P3&amp;gt;0 &amp;≤2%4674581.000Benign 464(99.4%)455(99.3%) Malignant 3(0.6%)3(0.7%) 4A&amp;gt;2% &amp;≤10%2883010.257Benign 271(94.1%)276(91.7%) Malignant 17(5.9%)25(8.3%) 4B&amp;gt;10% &amp;≤50%911350.037Benign 54(59.3%)61(45.2%) Malignant 37(40.7%)74(54.8%) 4C&amp;gt;50% &amp;&amp;lt;95%50790.473Benign 5(10.0%)4(5.1%) Malignant 45(90.0%)75(94.9%) 5≥95%174971.000Benign 3(1.7%)1(1.0%) Malignant 171(98.3%)96(99.0%) BI-RADS US Breast Imaging Reporting and Data System for Ultrasonography, PPV positive predictive value, ACR American College of Radiology Conclusion: Both static image evaluation and real-time assessment had similar diagnostic performance for most lesions; however, the diagnostic performance of static image evaluation for subcategory 4B lesions was lower. Real-time rescanning by the interpreter is strongly recommended for lesions of subcategory 4B after assessing static images. YH and JM contributed equally to this study. Citation Format: Hu Y, Mei J, Jiang X, Liu Q, Gong C. Is real-time rescanning by radiologists necessary for all ACR BI-RADS US category 3 to 5 lesions in the diagnostic setting? [abstract]. In: Proceedings of the 2018 San Antonio Breast Cancer Symposium; 2018 Dec 4-8; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2019;79(4 Suppl):Abstract nr P6-02-07.

Thyroid Nodule Location on Ultrasonography as a Predictor of Malignancy

The diagnostic capacity of ultrasonography (US) for differentiating between malignant and benign thyroid nodules is crucial in preventing unnecessary invasive procedures. This is the first study to evaluate whether thyroid nodule location on US has predictive value for malignancy. We retrospectively reviewed data from 219 patients with thyroid nodules who underwent fine-needle aspiration biopsy in 1 year. Patients' demographics as well as nodule's laterality, polarity, morphology, and multinodularity were analyzed. All malignant lesions were confirmed by surgical pathology. The majority of the patients were female (86.2%). Nodules were evenly distributed between the right lobe (46.3%) and left lobe (49.5%). Eight nodules (4.2%) were located in the isthmus. Most nodules (79.3%) were located in the lower pole, while 9.6% were located in the upper pole and 6.9% in the middle pole. Seventy-five patients (39.9%) had multiple nodules. Fourteen nodules were malignant, representing a prevalence of 7.4%. A significantly higher frequency of malignancy was observed in upper pole (22.2%) compared to lower pole (4.7%) and middle pole (15.4%). A multiple logistic regression model confirmed such association after adjusting for age, body mass index, multinodularity, and laterality. The odds of malignancy in the upper pole were 4 times higher than other locations (odds ratio, 4.6; P = .03). Our study is the first showing that thyroid nodules located in the upper pole can be considered as having higher risk for malignancy. It may enhance the predictive value of malignancy if it is included in thyroid nodule ultrasound classification guidelines. AACE = American Association of Clinical Endocrinologists; ATA = American Thyroid Association; BMI = body mass index; FNA = fine-needle aspiration; TMS = total malignancy score; TTW = taller than wide; US = ultrasonography.

Retrospective Application of the 2015 American Thyroid Association Guidelines for Ultrasound Classification, Biopsy Indications, and Follow-up Imaging of Thyroid Nodules: Can Improved Reporting Decrease Testing?

IntroductionThyroid ultrasound has been widely used to determine which nodules need further investigation. The goal of this study is to determine if using an ultrasonographic features checklist based on 2015 American Thyroid Association (ATA) guidelines can improve reporting and decrease unnecessary further testing. MethodsIn this retrospective study, ultrasonographic images of all nodules biopsied at our institution in 2014 and 2015 were reviewed by radiologists blinded to fine needle aspiration (FNA) biopsy result using a checklist. The checklist was prepared based on 2015 ATA guidelines. The ultrasonographic characteristics of thyroid nodules were compared with the result of biopsy to determine positive predictive value (PPV), negative predictive value (NPV), sensitivity, and specificity for predicting malignancy. Radiologists also made an overall recommendation on need for FNA. ResultsA total of 425 thyroid nodule ultrasound scans were reviewed by radiologists. Biopsy results of 31 nodules were malignant and 394 were non-malignant. Malignant nodules showed higher frequency of solid composition, hypoechoechogenicity, and cervical lymph node involvement compared to benign nodules. Solid nodule composition had the highest PPV (13%) and NPV (94.7%). Extra-thyroid extension had the highest specificity (90.1%). Lesion vascularity had the highest sensitivity (83.8%), followed by hypoechogenicity (65.6%). Overall, the checklist had a positive predictive value of 9%, negative predictive value of 97.5%, sensitivity of 96.8%, and specificity of 11.14%. Radiologists determined that 10% of the nodules were very low-risk and did not require FNA. ConclusionUsing a checklist based on 2015 ATA guideline thyroid nodule ultrasonographic features is a sensitive tool with high NPV to predict benign thyroid nodule, thereby preventing unnecessary FNAs.

Proposal for Endoscopic Ultrasonography Classification for Small Pancreatic Cancer.

Backgrounds: Endoscopic ultrasonography (EUS) is used to observe the stricture of the main pancreatic duct (MPD) and in diagnosing pancreatic cancer (PC). We investigate the findings on EUS by referring to the histopathological findings of resected specimens. Materials and Methods: Six patients with carcinoma in situ (CIS) and 30 patients with invasive carcinoma of 20 mm or less were included. The preoperative EUS findings were classified as follows. A1: Simple stricture type—no findings around the stricture; A2: Hypoecho stricture type—localized hypoechoic area without demarcation around the stricture; A3: Tumor stricture type—tumor on the stricture; B: Dilation type—the dilation of the pancreatic duct without a downstream stricture; C: Parenchymal tumor type—tumor located apart from the MPD. Results: Classes A1 and A2 consisted of 2 CISs, and 4 invasive carcinomas included two cases smaller than 5 mm in diameter. Most of the cancers classified as A3 or C were of invasive carcinoma larger than 5 mm in diameter. All cancers classified as B involved CIS. Serial pancreatic-juice aspiration cytologic examination (SPACE) was selected for all types of cases, with a sensitivity of 92.0%, while EUS-guided fine needle aspiration cytology (EUS-FNA) was only useful for invasive carcinoma, and its sensitivity was 66.7%. Conclusions: Stricture without a tumor could be a finding for invasive PC and pancreatic duct dilation without a downstream stricture could be a finding indicative of CIS. Carcinoma smaller than 5 mm in diameter could not be recognized by EUS. SPACE had a high sensitivity for diagnosing small PC.

Clinical Value of Shear Wave Elastography Added to Targeted Ultrasound (Second-Look Ultrasound) in the Evaluation of Breast Lesions Suspicious of Malignancy Detected on Magnetic Resonance Imaging.

To determine the value of shear wave elastography (SWE) added to targeted ultrasound (US) after breast magnetic resonance imaging (MRI). From July 2015 to October 2017, 40 patients who underwent targeted US evaluations of suspicious MRI-detected American College of Radiology Breast Imaging Reporting and Data System category 4 lesions (mass or nonmass enhancement) were enrolled in this prospective study. B-mode US and SWE examinations were performed to detect US correlates to MRI-detected lesions; their Breast Imaging Reporting and Data System categories were recorded; lesions that were dark blue on a 6-point color scale or had maximum elasticity of 30 kPa or less were categorized as soft. Biopsy was performed with US or MRI guidance, with the pathologic findings correlated with MRI, US, and SWE findings. The value of SWE for lesion detection and identification of benign lesions was determined. The mean age of the 40 patients was 51.1 years. There were 48 MRI-detected lesions (20 cancers, 3 high-risk lesions, and 25 benign lesions). Ultrasound correlates (8 category 3 and 25 category 4) were shown for 33 lesions (69%; P < .0001), with 16 cancers (80%; P < .0001) and 17 benign lesions. Shear wave elastography assisted detection of 3 (19%) cancers on US imaging. All 7 soft US category 3 lesions were benign (7 of 33 [21%]; P = .0014). Shear wave elastography was useful with targeted US after breast MRI to increase cancer detection by US. A significant number of US correlates to MRI-detected lesions could have been identified as benign (category 3 and soft) before biopsy, with the potential of short-interval follow-up of MRI-detected lesions with benign US correlates instead of biopsy.

High Body Mass Index and Thyroid Stimulating Hormone Levels Do Not Affect Thyroid Nodule Selection for Fine-Needle Aspiration Biopsy after Ultrasound Evaluation

Background and Objectives: This study was to evaluate whether high body mass index (BMI) or thyroid stimulating hormone (TSH) level would affect selecting thyroid nodule for fine-needle aspiration biopsy (FNA) after ultrasound (US) evaluation. Materials and Methods: A total of 3155 thyroid nodules (2159 benign and 996 malignant nodules) were included. Four grades of BMI and three levels of TSH were applied for grouping. US features of the thyroid nodules were divided into ‘probably benign’ and ‘suspicious for malignancy’ categories. Patients were grouped according to gender and univariate and multivariate logistic regression analysis were used to find the association between variables and malignancy. Results: TSH levels were significantly higher in the malignant group (p＜0.001). The grades of BMI did not show difference between the malignant and benign groups (females, p=0.074 and males, p=0.157). Younger age and ‘suspicious for malignancy’ US category were independent risk factors for malignancy in both genders. In females, a high TSH level (odds ratio=1.010, p＜0.001) had significant association with malignancy. Except for younger age (odds ratio=0.998, p＜0.001), no variable in nodules with ‘probably benign’ US category was significantly associated with malignancy. Conclusion: High TSH levels were more frequent in thyroid malignancy group, but neither high BMI nor high TSH level give additional information for FNA selection after US.

Thyroid Ultrasound Texture Classification Using Autoregressive Features in Conjunction With Machine Learning Approaches

The thyroid is one of the largest endocrine glands in the human body, which is involved in several body mechanisms like controlling protein synthesis, use of energy sources, and controlling the body's sensitivity to other hormones. Thyroid segmentation and volume reconstruction are hence essential to diagnose thyroid related diseases as most of these diseases involve a change in the shape and size of the thyroid over time. Classification of thyroid texture is the first step toward the segmentation of the thyroid. The classification of texture in thyroid Ultrasound (US) images is not an easy task as it suffers from low image contrast, presence of speckle noise, and non-homogeneous texture distribution inside the thyroid region. Hence, a robust algorithmic approach is required to accurately classify thyroid texture. In this paper, we propose three machine learning based approaches: Support Vector Machine; Artificial Neural Network; and Random Forest Classifier to classify thyroid texture. The computation of features for training these classifiers is based on a novel approach recently proposed by our team, where autoregressive modeling was applied on a signal version of the 2D thyroid US images to compute 30 spectral energy-based features for classifying the thyroid and non-thyroid textures. Our approach differs from the methods proposed in the literature as they use image-based features to characterize thyroid tissues. We obtained an accuracy of around 90% with all the three methods.

Systemic immune-inflammation index and ultrasonographic classification of breast imaging-reporting and data system predict outcomes of triple-negative breast cancer.

IntroductionThis research was conducted to explore the relationship between the systemic immune-inflammation index (SII) and breast imaging-reporting and data system (BI-RADS) classification using ultrasonography and the survival of patients with triple-negative breast cancer (TNBC) in a cohort of Chinese.MethodsA total of 215 TNBC patients treated at our hospital between November 2008 and March 2016 were enrolled in this study. We used the log-rank test and Kaplan–Meier curves to assess the overall survival (OS) and disease-free survival (DFS) differences between groups. The prognostic role of SII and other clinicopathological characteristics in TNBC patients were identified using the Cox regression model.ResultsPatients with low and high SII had median OS of 60.9 and 40.3 months, respectively, (HR=3.78, 95% CI: 2.16–4.15, P<0.001); while the median DFS was 22.4 months and 14.4 months for TNBC patients with low and high SII, respectively (HR =3.16, 95% CI: 1.82–4.02, P<0.001). For patients with grade 5 ultrasonographic BI-RADS classification, the median DFS and OS were 41.2 and 16.5 months, respectively, whereas, it was 57.7 and 21.3 months, respectively, for those with BI-RADS grades 3–4 (P<0.01). According to multivariable analyses, increased SII was a risk factor that independently predicted poor OS (HR =2.96, 95% CI: 2.18–3.98, P<0.001) and DFS (HR = 2.85, 95% CI: 1.62–3.81, P=0.005). In addition, tumor stage, BI-RADS, and histological grade also independently predicted poor OS (P=0.002, <0.001, 0.004).ConclusionPretreatment SII and BI-RADS 5 were independent indicators for prognosis in TNBC patients. It is imperative to conduct prospective studies to evaluate the potential role of SII in patient selection, treatment guidance, and design of clinical trials.

Ultrasonographic prediction of the location of the proximal submandibular stones relative to the lingual nerve for the strategic transoral surgery

Introduction Transoral surgical removal of the proximal submandibular stones has been performed to preserve the gland. Application of ultrasound could extend the indication of transoral surgery to the impalpable proximal stones. However, technical unfamiliarity of ultrasound scanning and anatomical complexity of proximal submandibular area can be hurdles for the general application of ultrasound guided transoral surgery. The objectives of this study were to assess the feasibility of preoperative ultrasonographic determination of the location of stone relative to the lingual nerve (LN) and to find optimal surgical approach for the proximal submandibular stones. Methods We retrospectively reviewed the medical records of 170 patients who were diagnosed as proximal submandibular stones and treated ultrasonography-guided transoral surgery from Mar 2011 to May 2018. Three patients had bilateral stones. The posterior edge of mylohyoid (PM) was used as the preoperative ultrasonographic criterion to classify the proximal submandibular stones into 3 types: stones anterior to PM (type I), stones crossing PM (type II), and stones posterior to PM (type III). The correlation between ultrasonographic and operative findings was investigated. Results Preoperative ultrasonography determined the proximal stones as type I in 40 (23.8%) cases, as type II in 69 (41.1%) cases, and as type III in 59 (35.1%) cases. According to the operative records, the stones were approached medially to the LN in 58 (34.5%) cases and laterally to the LN in 110 (65.5%) cases. In type 1 stones, 90% of stones were approached medially to the LN, while in type 2 and 3 stones, 78.3% and 88.1% of stones were approached laterally to the LN during transoral surgery, respectively. The surgical approaches were significantly different in relation to preoperative ultrasonographic classification of the proximal submandibular stones (P Conclusion Optimal transoral surgical approach relative to the lingual nerve can be selected based upon preoperative ultrasonographic findings of the proximal submandibular stones.

Correction to: Diagnostic Performance of Ultrasound-Based Risk-Stratification Systems for Thyroid Nodules: Comparison of the 2015 American Thyroid Association Guidelines with the 2016 Korean Thyroid Association/Korean Society of Thyroid Radiology and 2017 American College of Radiology Guidelines Ha EJ, Na DG, Moon W-J, Lee YH, and Choi N Thyroid 2018;28:1532-1537. DOI: 10.1089/thy.2018.0094.

ThyroidVol. 29, No. 1 CorrectionFree AccessCorrection to: Diagnostic Performance of Ultrasound-Based Risk-Stratification Systems for Thyroid Nodules: Comparison of the 2015 American Thyroid Association Guidelines with the 2016 Korean Thyroid Association/Korean Society of Thyroid Radiology and 2017 American College of Radiology Guidelines by Ha EJ, Na DG, Moon W-J, Lee YH, and Choi N. Thyroid 2018;28:1532–1537. DOI: 10.1089/thy.2018.0094is erratum ofDiagnostic Performance of Ultrasound-Based Risk-Stratification Systems for Thyroid Nodules: Comparison of the 2015 American Thyroid Association Guidelines with the 2016 Korean Thyroid Association/Korean Society of Thyroid Radiology and 2017 American College of Radiology GuidelinesPublished Online:23 Jan 2019https://doi.org/10.1089/thy.2018.0094.correxAboutSectionsPDF/EPUB Permissions & CitationsPermissionsDownload CitationsTrack CitationsAdd to favorites Back To Publication ShareShare onFacebookTwitterLinked InRedditEmail In the November 2018 issue of Thyroid (vol. 28, no. 11; 1532–1537) the article entitled “Diagnostic Performance of Ultrasound-Based Risk-Stratification Systems for Thyroid Nodules: Comparison of the 2015 American Thyroid Association Guidelines with the 2016 Korean Thyroid Association/Korean Society of Thyroid Radiology and 2017 American College of Radiology Guidelines” requires correction.In the title and abstract of the article, the American College of Radiology was incorrectly named the American Congress of Radiology.The original title and abstract read: Diagnostic Performance of Ultrasound-Based Risk-Stratification Systems for Thyroid Nodules: Comparison of the 2015 American Thyroid Association Guidelines with the 2016 Korean Thyroid Association/Korean Society of Thyroid Radiology and 2017 American Congress of Radiology GuidelinesAbstractPurpose:The aim of this study was to compare the diagnostic performance of ultrasound (US)-based risk stratification systems for thyroid nodules in the 2015 American Thyroid Association (ATA) guidelines with those of the 2016 Korean Thyroid Association (KTA)/Korean Society of Thyroid Radiology (KSThR) and 2017 American Congress of Radiology (ACR) guidelines.These have been revised to read:Diagnostic Performance of Ultrasound-Based Risk-Stratification Systems for Thyroid Nodules: Comparison of the 2015 American Thyroid Association Guidelines with the 2016 Korean Thyroid Association/Korean Society of Thyroid Radiology and 2017 American College of Radiology GuidelinesAbstractPurpose: The aim of this study was to compare the diagnostic performance of ultrasound (US)-based risk stratification systems for thyroid nodules in the 2015 American Thyroid Association (ATA) guidelines with those of the 2016 Korean Thyroid Association (KTA)/Korean Society of Thyroid Radiology (KSThR) and 2017 American College of Radiology (ACR) guidelines.The online version has been corrected to reflect this. The authors apologize for these errors.FiguresReferencesRelatedDetailsCited byComparison of Different Ultrasound Classification Systems of Thyroid Nodules for Identifying Malignant Potential: A Cross-sectional StudyClinics, Vol. 76Related articlesDiagnostic Performance of Ultrasound-Based Risk-Stratification Systems for Thyroid Nodules: Comparison of the 2015 American Thyroid Association Guidelines with the 2016 Korean Thyroid Association/Korean Society of Thyroid Radiology and 2017 American College of Radiology Guidelines17 Dec 2018Thyroid Volume 29Issue 1Jan 2019 InformationCopyright 2019, Mary Ann Liebert, Inc., publishersTo cite this article:Correction to: Diagnostic Performance of Ultrasound-Based Risk-Stratification Systems for Thyroid Nodules: Comparison of the 2015 American Thyroid Association Guidelines with the 2016 Korean Thyroid Association/Korean Society of Thyroid Radiology and 2017 American College of Radiology Guidelines by Ha EJ, Na DG, Moon W-J, Lee YH, and Choi N. Thyroid 2018;28:1532–1537. DOI: 10.1089/thy.2018.0094.Thyroid.Jan 2019.159-159.http://doi.org/10.1089/thy.2018.0094.correxPublished in Volume: 29 Issue 1: January 23, 2019PDF download