Nodule Classification on Low-Dose Unenhanced CT and Standard-Dose Enhanced CT: Inter-Protocol Agreement and Analysis of Interchangeability

이경희,이경원,박지훈,한경화,김지항,이상민,박창민 대한영상의학회 2018 Korean Journal of Radiology Vol.19 No.3

Objective: To measure inter-protocol agreement and analyze interchangeability on nodule classification between low-dose unenhanced CT and standard-dose enhanced CT. Materials and Methods: From nodule libraries containing both low-dose unenhanced and standard-dose enhanced CT, 80 solid and 80 subsolid (40 part-solid, 40 non-solid) nodules of 135 patients were selected. Five thoracic radiologists categorized each nodule into solid, part-solid or non-solid. Inter-protocol agreement between low-dose unenhanced and standard-dose enhanced images was measured by pooling κ values for classification into two (solid, subsolid) and three (solid, part-solid, non-solid) categories. Interchangeability between low-dose unenhanced and standard-dose enhanced CT for the classification into two categories was assessed using a pre-defined equivalence limit of 8 percent. Results: Inter-protocol agreement for the classification into two categories {κ, 0.96 (95% confidence interval [CI], 0.94−0.98)} and that into three categories (κ, 0.88 [95% CI, 0.85−0.92]) was considerably high. The probability of agreement between readers with standard-dose enhanced CT was 95.6% (95% CI, 94.5−96.6%), and that between low-dose unenhanced and standard-dose enhanced CT was 95.4% (95% CI, 94.7−96.0%). The difference between the two proportions was 0.25% (95% CI, -0.85–1.5%), wherein the upper bound CI was markedly below 8 percent. Conclusion: Inter-protocol agreement for nodule classification was considerably high. Low-dose unenhanced CT can be used interchangeably with standard-dose enhanced CT for nodule classification.

Differentiation between Benign and Malignant Solid Thyroid Nodules Using an US Classification System

이영훈,김동욱,인현신,박지성,김상효,엄재욱,김보미,이은주,노명호 대한영상의학회 2011 Korean Journal of Radiology Vol.12 No.5

Objective: To evaluate the diagnostic accuracy of a new ultrasound (US) classification system for differentiating between benign and malignant solid thyroid nodules. Materials and Methods: In this study, we enrolled 191 consecutive patients who received real-time US and subsequent US diagnoses for solid thyroid nodules, and underwent US-guided fine-needle aspiration. Each thyroid nodule was prospectively classified into 1 of 5 diagnostic categories by real-time US: “malignant,” “suspicious for malignancy,” “borderline,” “probably benign,” and “benign”. We evaluated the diagnostic accuracy of thyroid US and the cut-off US criteria by comparing the US diagnoses of thyroid nodules with cytopathologic results. Results: Of the 191 solid nodules, 103 were subjected to thyroid surgery. US categories for these 191 nodules were malignant (n = 52), suspicious for malignancy (n = 16), borderline (n = 23), probably benign (n = 18), and benign (n =82). A receiver-operating characteristic curve analysis revealed that the US diagnosis for solid thyroid nodules using the 5-category US classification system was very good. The sensitivity, specificity, positive and negative predictive values, and accuracy of US diagnosis were 86%, 95%, 91%, 92%, and 92%, respectively, when benign, probably benign, and borderline categories were collectively classified as benign (negative). Conclusion: The diagnostic accuracy of thyroid US for solid thyroid nodules is high when the above-mentioned US classification system is applied.

흉부 CT 영상에서 폐 결절 검출을 위한 Log-polar Sampling기반 Voxel Classification 방법

최욱진(Choi, Wook-Jin),최태선(Choi, Tae-Sun) 한국정보전자통신기술학회 2013 한국정보전자통신기술학회논문지 Vol.6 No.1

본 논문에서는 voxel classification을 이용한 폐 결절 자동 검출 시스템을 제안한다. 제안하는 폐 영상 분석 방법은 크게 세 단계로 구성된다. 첫 번째 단계에서는 분석 대상 폐 영역을 분할한다. 그리고 두 번째 단계는 분할된 폐 영역 내에서 폐 구조물을 분할한다. 마지막으로 두 번째 과정에서 분할된 폐결절후보와 폐혈관 voxel을 대상으로 log-polar sampling을 이용한 특징 벡터를 만들고, 특징벡터를 입력 값으로 하여 support vector machine classifier를 이용하여 분석대상 voxel을 폐 결절 voxel과 비결절 voxel로 구분하여 폐 결절을 검출한다. In this paper, we propose the pulmonary nodule detection system based on voxel classification. The proposed system consists of three main steps. In the first step, we segment lung volume. In the second step, the lung structures are initially segmented. In the last step, we classify the nodules using voxel classification. To describe characteristics of each voxel, we extract the log-polar sampling based features. Support Vector Machine is applied to the extracted features to classify into nodules and non-nodules.

Recent Advances in Core Needle Biopsy for Thyroid Nodules

정찬권,백정환 대한내분비학회 2017 Endocrinology and metabolism Vol.32 No.4

Core needle biopsy (CNB) was introduced as an alternative diagnostic tool to fine-needle aspiration (FNA), and is increasingly being used in the preoperative assessment of thyroid nodules. CNB provides a definitive diagnosis in most cases, but it sometimes may be inconclusive. CNB has the advantage of enabling a histologic examination in relation to the surrounding thyroid tissue, immunohistochemistry, and molecular testing that can provide a more accurate assessment than FNA in selected cases. Nevertheless, CNB should be performed only by experienced experts in thyroid interventions to prevent complications because CNB needles are larger in caliber than FNA needles. As recent evidence has accumulated, and with improvements in the technique and devices for thyroid CNB, the Korean Society of Thyroid Radiology released its 2016 thyroid CNB guidelines and the Korean Endocrine Pathology Thyroid Core Needle Biopsy Study Group published a consensus statement on the pathology reporting system for thyroid CNB in 2015. This review presents the current consensus and recommendations regarding thyroid CNB, focusing on indications, complications, and pathologic classification and reporting.

2.5차원 다중뷰 기반 특징 및 데이터 확장을 통한 간유리음영 결절의 다중클래스 분류

이선영(Seon Young Lee),정주립(Julip Jung),홍헬렌(Helen Hong),송용섭(Yong Sub Song),김형진(Hyungjin Kim),박창민(Chang Min Park) 한국정보과학회 2018 정보과학회 컴퓨팅의 실제 논문지 Vol.24 No.10

간유리음영 결절은 내부 고형 성분의 포함 여부 및 크기에 따라 악성도가 달라지기 때문에 고형 성분의 크기에 따른 혼합 간유리음영 결절과 순수 간유리음영 결절을 구분하는 것이 중요하다. 그러나 고형 성분의 크기가 작은 혼합 간유리음영 결절은 순수 간유리음영 결절과 유사하게 나타나 구분이 어렵다. 본 논문에서는 다중뷰-슬랩 관심영역 기반 특징 및 데이터 확장을 통한 간유리음영 결절의 분류 방법을 제안한다. 첫째, 분류기의 훈련 데이터 수를 늘리고 과적합을 피하기 위해 데이터 확장을 수행한다. 둘째, 간유리음영 결절로부터 네 종류의 관심 영역을 생성한다. 셋째, 각 관심영역으로부터 고형 성분의 특성을 반영하는 특징을 추출하고 의미 있는 특징을 선별한다. 넷째, 랜덤 포레스트를 이용해 간유리음영 결절을 분류한다. 2.5차원 다중뷰-슬랩 관심영역을 이용한 분류 정확도는 84.05%로 3차원 관심영역과 2.5차원 다중뷰 관심영역을 이용한 정확도보다 각각 6.27%, 4.44% 높았다. Differentiation between part-solid ground-glass nodules (GGNs) with a variable sized solid component from pure GGNs is important because the malignancy rate of GGN is different according to the presence and size of solid components in chest CT images. However, because of their similar appearance, it is difficult to distinguish between part-solid GGN and pure GGN when the part-solid GGN includes small solid components. In this paper, we propose a multi-class classification method for GGN using multiview-slab region of interest (ROI)-based features and data augmentation. First, data augmentation is performed to enlarge the training dataset of the classifier and to avoid the overfitting problem. Second, four ROIs are generated from a GGN. Third, features that reflect the characteristics of the solid component are extracted from each ROI, and significant features are selected for classification. Finally, GGNs are classified using the random forest (RF). In the experiments, the classification accuracy with 2.5-dimensional multiview-slab ROI was 84.05%, which was 6.27% and 4.44% higher than the classification accuracy with 3-dimensional ROI and 2.5-dimensional ROI, respectively.

기식성 음성 장애 환자들의 음향학적 특성과 분류 변인—결절, 용종환자를 중심으로—

이인애(Inae Lee),성철재(Cheoljae Seong) 사단법인 한국언어학회 2020 언어학 Vol.0 No.88

Incomplete phonation of vocal cords seems to cause breathy voice. Vocal nodules and vocal polyps are typical functional voice disorders. This study aimed to find the acoustic parameters useful for discriminating these two groups. In this study, 56 subjects diagnosed with speech disorders (27 vocal nodules, 29 vocal polyps) were recruited, and the MDVP (Multidimensional Voice Program) played within CSL (Computerized Speech Lab. Kay Elemetrics Co., Model No. 4300) was used for recording sustained vowel /a/. Praat (ver. 6.0.48) was used for acoustic analysis. Mann-Whitney U test showed that the MDVP parameters (excluding frequency-relevant variables) had no effect on discriminating two groups (vocal nodules vs. vocal polyps). However, when it comes to the cepstral parameters measured by Praat, there was statistically significant differences between groups: RNR (p<.o5), cepsPeak (p<.o5), quefrency (p<.o5), and CPP(p) <.o5) and cepslntercept (p<.o5). With respect to the relative rank of variable importance, quefrency has the highest rank, and rnr and CPP followed in order. As a result of classification by the logistic regression model using these three variables, the overall classification accuracy for the training data was 81.4%, 94.4% for the vocal nodule group, and 68.4% for the vocal polyp group. In test data, nodules were 100% and vocal fold polyps were 30%. Although the overall classification accuracy was 65%, the classification accuracy of the polyps was too low, making it difficult to trust the model. In the case of the svm (support vector machine) model, the overall classification ratio of the training data was 94.6% (nodules (17/18)=94.4%, polyps (18/19)=94.7%). The overall classification accuracy of the test data was found to be 68.4% (nodule (6/9) = 66.7%. Polyp (7/10) = 70%). It was found that, therefore, the performance of the svm was higher and more stable.

Update from the 2022 World Health Organization Classification of Thyroid Tumors: A Standardized Diagnostic Approach

정찬권,Andrey Bychkov,Kennichi Kakudo 대한내분비학회 2022 Endocrinology and metabolism Vol.37 No.5

The fifth edition of the World Health Organization (WHO) histologic classification of thyroid neoplasms released in 2022 includes newly recognized tumor types, subtypes, and a grading system. Follicular cell-derived neoplasms are categorized into three families (classes): benign tumors, low-risk neoplasms, and malignant neoplasms. The terms “follicular nodular disease” and “differentiated high-grade thyroid carcinoma” are introduced to account for multifocal hyperplastic/neoplastic lesions and differentiated thyroid carcinomas with high-grade features, respectively. The term “Hürthle cells” is replaced with “oncocytic cells.” Invasive encapsulated follicular and cribriform morular variants of papillary thyroid carcinoma (PTC) are now redefined as distinct tumor types, given their different genetic alterations and clinicopathologic characteristics from other PTC subtypes. The term “variant” to describe a subclass of tumor has been replaced with the term “subtype.” Instead, the term “variant” is reserved to describe genetic alterations. A histologic grading system based on the mitotic count, necrosis, and/or the Ki67 index is used to identify high-grade follicular-cell derived carcinomas and medullary thyroid carcinomas. The 2022 WHO classification introduces the following new categories: “salivary gland-type carcinomas of the thyroid” and “thyroid tumors of uncertain histogenesis.” This review summarizes the major changes in the 2022 WHO classification and their clinical relevance.

흉부 CT 영상에서 결절의 밝기값, 재질 및 형상 증강 영상 기반의 GGN-Net을 이용한 간유리음영 결절 자동 분류

변소현(So Hyun Byun),정주립(Julip Jung),홍헬렌(Helen Hong),송용섭(Yong Sub Song),김형진(Hyungjin Kim),박창민(Chang Min Park) 한국컴퓨터그래픽스학회 2018 컴퓨터그래픽스학회논문지 Vol.24 No.5

본 논문에서는 흉부 CT 영상에서 결절의 밝기값, 재질 및 형상 증강 영상 기반의 GGN-Net을 이용해 간유리음영 결절 자동 분류 방법을 제안한다, 첫째, 입력 영상에 결절 내부의 고형 성분의 유무 및 크기 정보가 포함될 수 있도록 밝기값, 재질 및 형상 증강 영상의 활용을 제안한다. 둘째, 다양한 입력 영상을 여러 개의 컨볼루션 모듈을 통해 획득한 특징맵을 내부 네트워크에서 통합하여 훈련하는 GGN-Net를 제안한다. 제안 방법의 분류 정확성 평가를 위해 순수 간유리음영 결절 90개와 고형 성분의 크기가 5mm 미만인 혼합 간유리음영 결절 38개, 5mm 이상 고형 성분의 크기를 가지는 혼합 간유리음영 결절 23개의 데이터를 사용하였으며, 입력 영상이 간유리음영 결절 분류 결과에 미치는 영향을 비교하기 위해 다양한 입력 영상을 구성하여 결과를 비교하였다. 실험 결과, 밝기값, 재질 및 형상 정보가 함께 고려된 입력 영상을 사용한 제안 방법이 정확도가 82.75%로 가장 좋은 결과를 보였다. In this paper, we propose an automated method for the ground-glass nodule(GGN) classification using GGN-Net based on intensity, texture, and shape-enhanced images in chest CT images. First, we propose the utilization of image that enhances the intensity, texture, and shape information so that the input image includes the presence and size information of the solid component in GGN. Second, we propose GGN-Net which integrates and trains feature maps obtained from various input images through multiple convolution modules on the internal network. To evaluate the classification accuracy of the proposed method, we used 90 pure GGNs, 38 part-solid GGNs less than 5mm with solid component, and 23 part-solid GGNs larger than 5mm with solid component. To evaluate the effect of input image, various input image set is composed and classification results were compared. The results showed that the proposed method using the composition of intensity, texture and shape-enhanced images showed the best result with 82.75% accuracy.

흉부 CT 영상에서 다중 뷰 영상과 텍스처 분석을 통한 고형 성분이 작은 폐 간유리음영 결절 분류

이선영(Seon Young Lee),정주립(Julip Jung),이한상(Han Sang Lee),홍헬렌(Helen Hong) 한국멀티미디어학회 2017 멀티미디어학회논문지 Vol.20 No.7

Cu-Ni-Co-Fe-S 매트의 분쇄/분급 공정 설계

정민욱,한성수,남철우,박경호,박재구 한국자원공학회 2017 한국자원공학회지 Vol.54 No.1

Bond work index of Cu-Ni-Co-Fe-S matte having different sulfur contents was measured, followed by closed grinding circuit design which consisted of ball mill and spiral classifier. At first, the mineralogical and chemical contents of the matte was analyzed using XRF and SEM-EDS method. Afterward, Grinding work index of matte was estimated using the Bond work index test. The work index measured was used to determine grinding circuit specification lastly. It was found that the more the sulfur concentration was in the matte, the less Bond work index value was as sulfur contents increased from S 15.9 wt.% - S 24.9 wt.%. As a result, the grinding energy was reduced up to 45% in the range. For grinding process design, Closed grinding circuit was compared with open grinding circuit, in which hydrocyclone and spiral classifier were also investigated. Cu-Fe-Ni-Co-S 매트의 분쇄일지수를 측정하였으며, 이어서 볼밀과 스파이럴 분급기로 구성된 매트의 폐회로 분쇄 공정을 설계하였다. 우선 XRF와 SEM-EDS 분석을 이용하여 매트의 광물학적, 화학적 조성 변화를 조사하였다. 이어서 본드 분쇄 일지수 시험 방법을 이용하여 본드 분쇄 일지수를 측정하였으며, 측정된 본드 분쇄 일지수를 바탕으로 분쇄 공정에서 사용되는 기기의 종류와 규격을 결정하였다. 실험 결과, 매트에 황 함량이 많을수록 본드 분쇄일지수 값이 낮게 나타났으며, 황 함량이 15.9 wt.%에서 24.9 wt.%으로 증가함에 따라 최대 45%의 분쇄일지수가 감소함을 확인할 수 있었다. 분쇄 공정 설계 시, 개회로와 폐회로 공정을 비교하였으며, 폐회로 공정 운용 시 분급 공정으로습식 싸이클론과 스파이럴 분급기의 사용을 검토하였다.