TTN: Topological Transformer Network for Automated Coronary Artery Branch Labeling in Cardiac CT Angiography.

OBJECTIVE: Existing methods for automated coronary artery branch labeling in cardiac CT angiography face two limitations: 1) inability to model overall correlation of branches, since differences between branches cannot be captured directly. 2) a serious class imbalance between main and side branches. METHODS AND PROCEDURES: Inspired by the application of Transformer in sequence data, we propose a topological Transformer network (TTN), which solves the vessel branch labeling from a novel perspective of sequence labeling learning. TTN detects differences between branches by establishing their overall correlation. A topological encoding that represents the positions of vessel segments in the artery tree, is proposed to assist the model in classifying branches. Also, a segment-depth loss is introduced to solve the class imbalance between main and side branches. RESULTS: On a dataset with 325 CCTA, our method obtains the best overall result on all branches, the best result on side branches, and a competitive result on main branches. CONCLUSION: TTN solves two limitations in existing methods perfectly, thus achieving the best result in coronary artery branch labeling task. It is the first Transformer based vessel branch labeling method and is notably different from previous methods. CLINICAL IMPACT: This Pre-Clinical Research can be integrated into a computer-aided diagnosis system to generate cardiovascular disease diagnosis report, assisting clinicians in locating the atherosclerotic plaques.

Dynamic evaluation of unruptured intracranial aneurysms by 4D-CT angiography: comparison with digital subtraction angiography (DSA) and surgical findings.

BACKGROUND: This study was to prospectively investigate the feasibility of four-dimensional computed tomography angiography (4D-CTA) with electrocardiogram-gated (ECG) reconstruction for preoperative evaluation of morphological parameters, and compared with digital subtraction angiography (DSA). We also aimed to detect pulsation in unruptured intracranial aneurysms (UIAs) by using 4D-CTA, as a potential predicting factor of growth or rupture. MATERIALS: 64 patients with 64 UIAs underwent ECG-gated dynamic 4D-CTA imaging before treatment, of which 46 patients additionally underwent DSA. Original scanning data were reconstructed to produce 20 data sets of cardiac cycles with 5%-time intervals. The extent of agreement on UIAs morphological features assessed with 4D-CTA and DSA was estimated using the k coefficient of the Kappa test. The radiation doses were also calculated and compared between 4D-CTA and DSA. In the aneurysmal surgically treated in our institution, we were able to compare the surgical findings of the aneurysm wall with 4D-CTA images. We performed long-term follow-up on untreated patients. RESULTS: The morphological characteristics detected by 4D-CTA and DSA were consistent in aneurysm location (k = 1.0), shape (k = 0.76), maximum diameter (k = 0.94), aneurysm neck (k = 0.79) and proximity to parent and branch vessels (k = 0.85). 4D-CTA required lower radiation doses (0.32 ± 0.11 mSv) than DSA (0.84 ± 0.37 mSv, P < 0.001). Pulsation was detected in 26 of the 64 unruptured aneurysms, and all underwent neurosurgical clipping or interventional embolization. In aneurysms surgically treated in our hospital, we observed a significant correlation between 4D-CTA findings and surgical evaluation of the aneurysmal wall, in particular the irregular pulsations detected on 4D-CTA have demonstrated to correspond to dark-reddish thinner wall at surgery. CONCLUSIONS: In this proof-of-concept study, 4D-CTA provided real-time, non-invasive preoperative assessments of UIAs comparable to DSA. Moreover, optimal correlation between the irregular pulsation detected by 4D-CTA and the surgical findings support a possible role of this technique to identify aneurysms with a higher risk of rupture.

[18F] AlF­NOTA­FAPI­04 PET/CT as a promising tool for imaging fibroblast activation protein in gastrointestinal system cancers: a prospective investigation of comparative analysis with 18F-FDG.

PURPOSE: The radiopharmaceutical [18F]AlF-NOTA-FAPI-04 presents a promising alternative to 68 Ga-FAPI owing to its relatively longer half-life. This study aimed to evaluate the clinical usefulness of [18F]AlF-NOTA-FAPI-04 PET/CT for the diagnosis of primary and metastatic lesions in various types of gastrointestinal system cancers, compared with 18F-FDG PET/CT. METHODS: Patients diagnosed with gastrointestinal system malignancies were prospectively enrolled. All patients underwent both 18F-FDG and 18F-FAPI-04 PET/CT scans within one week, with 44 (73.3%) for cancer staging and 16 (26.7%) for tumor restaging. Diagnostic efficacy of the primary tumor, as well as the presence and number of lymph nodes and distant metastases, were assessed. Tumor uptake was quantified by the maximum standard uptake value (SUVmax). RESULTS: For detection of primary tumor, the diagnostic sensitivity of 18F-FDG PET/CT was 72.7%, while it was 97.7% for 18F-FAPI-04 PET/CT. Based on per-lymph node analysis, the sensitivity, specificity, and accuracy of 18F-FAPI-04 PET/CT in diagnosing metastatic lymph nodes were 91.89%, 92.00%, and 91.96%, respectively. These values were notably higher than those 18F-FDG PET/CT (79.72%, 81.33% and 80.80%, respectively). The 18F-FAPI-04 PET/CT surpassed 18F-FDG PET/CT in detecting suspected metastases in the brain (7 vs. 3), liver (39 vs. 20), bone (79 vs. 51), lung (11 vs. 4), and peritoneal carcinoma (48 vs. 22). Based on per-patient analysis, differential diagnostic accuracies (18F-FAPI-04 vs. 18F-FDG PET/CT) were observed in all patients (91.7% vs. 76.7%), the initial staging group (90.9% vs. 79.5%), and the re-staging group (93.8% vs. 68.7%). Additionally, 18F-FAPI-04 PET/CT revised final diagnosis in 31.7% of patients, contrasting with 18F-FDG PET/CT, and prompted changes in clinical management for 21.7% of the patients. CONCLUSION: 18F-FAPI-04 PET/CT outperforms 18F-FDG PET/CT in delineating the primary gastrointestinal tumors and detecting suspected metastatic lesions due to a higher target-to-background ratio (TBR). Moreover, 18F-FAPI-04 PET/CT could provide valuable guidance for tumor staging, thereby having a potential impact on patient management.

PET/CT-based radiomics analysis may help to predict neoadjuvant chemotherapy outcomes in breast cancer.

Background: The aim of this study was to evaluate the clinical usefulness of radiomics signature-derived 18F-fluorodeoxyglucose (18F-FDG) positron emission tomography-computed tomography (PET-CT) for the early prediction of neoadjuvant chemotherapy (NAC) outcomes in patients with (BC). Methods: A total of 124 patients with BC who underwent pretreatment PET-CT scanning and received NAC between December 2016 and August 2019 were studied. The dataset was randomly assigned in a 7:3 ratio to either the training or validation cohort. Primary tumor segmentation was performed, and radiomics signatures were extracted from each PET-derived volume of interest (VOI) and CT-derived VOI. Radiomics signatures associated with pathological treatment response were selected from within a training cohort (n = 85), which were then applied to generate different classifiers to predict the probability of pathological complete response (pCR). Different models were then independently tested in the validation cohort (n = 39) regarding their accuracy, sensitivity, specificity, and area under the curve (AUC). Results: Thirty-five patients (28.2%) had pCR to NAC. Twelve features consisting of five PET-derived signatures, four CT-derived signatures, and three clinicopathological variables were candidates for the model's development. The random forest (RF), k-nearest neighbors (KNN), and decision tree (DT) classifiers were established, which could be utilized to predict pCR to NAC with AUC ranging from 0.819 to 0.849 in the validation cohort. Conclusions: The PET/CT-based radiomics analysis might provide efficient predictors of pCR in patients with BC, which could potentially be applied in clinical practice for individualized treatment strategy formulation.

Feasibility Study of Intelligent Three-Dimensional Accurate Liver Reconstruction Technology Based on MRI Data.

Intelligent three-dimensional (3D) reconstruction technology plays an important role in the diagnosis and treatment of diseases. It has been widely used in assisted liver surgery. At present, the 3D reconstruction information of liver is mainly obtained based on CT enhancement data. It has also been commercialized. However, there are few reports on the display of 3D reconstruction information of the liver based on MRI. The purpose of this study is to propose a new idea of intelligent 3D liver reconstruction based on MRI technology and verify its feasibility. Two different liver scanning data (CT and MRI) were selected from the same batch of patients at the same time (patients with a time interval of no more than two weeks and without surgery). The results of liver volume, segmentation, tumor, and simulated surgery based on MRI volume data were compared with those based on CT data. The results show that the results of 3D reconstruction based on MRI data are highly consistent with those based on CT 3D reconstruction. At the same time, in addition to providing the information provided by CT 3D reconstruction, it also has its irreplaceable advantages. For example, multi-phase (early, middle and late arterial, hepatobiliary, etc.) scanning of MRI technology can provide more disease information and display of biliary diseases. In a word, MRI technology can be used for 3D reconstruction of the liver. Hence, a new feasible and effective method to show the liver itself and its disease characteristics is proposed.

A Giant Borderline Phyllodes Tumor of Breast With Skin Ulceration Leading to Non-Insular Tumorigenic Hypoglycemia: A Case Report and Literature Review.

Phyllodes tumor (PT) is a special type of breast tumors, including three types: malignant, borderline, and benign. Most of these tumors form unilateral disease and can rapidly increase in size. The occurrence of axillary lymph node metastasis is rare. Tumor-associated hypoglycemia can be divided into non-islet cell tumor and insulinoma. In non-islet cell tumor hypoglycemia (NICTH), a considerable high molecular weight form of insulin like growth factor 2 (IGF-2) is formed, which abnormally binds to insulin receptors in the tissues and causes hypoglycemia. Breast phyllodes tumors with NICTH are rare and first reported in 1983. Surgical resection is the main treatment and hypoglycemia symptoms usually resolve after surgery. Nevertheless, prior to surgery, intravenous glucose infusion is used to maintain blood glucose levels. A female patient presented with a rapidly growing breast mass and was diagnosed with a phyllodes tumor with NICTH at our hospital in August 2020; she was successfully treated through surgical resection. We reviewed the relevant literature to investigate and analyze the relationship between NICTH and phyllodes tumors, as well as optimize its diagnosis and treatment.

Prediction Model Combining Clinical and MR Data for Diagnosis of Lymph Node Metastasis in Patients With Rectal Cancer.

BACKGROUND: Determining the status of lymph node (LN) metastasis in rectal cancer patients preoperatively is crucial for the treatment option. However, the diagnostic accuracy of current imaging methods is low. PURPOSE: To develop and test a model for predicting metastatic LNs of rectal cancer patients based on clinical data and MR images to improve the diagnosis of metastatic LNs. STUDY TYPE: Retrospective. SUBJECTS: In all, 341 patients with histologically confirmed rectal cancer were divided into one training set (120 cases) and three validation sets (69, 103, 49 cases). FIELD STRENGTH/SEQUENCE: 3.0T, axial and sagittal T2 -weighted turbo spin echo and diffusion-weighted imaging (b = 0 s/mm2 , 800 s/mm2 ) ASSESSMENT: In the training dataset, univariate logistic regression was used to identify the clinical factors (age, gender, and tumor markers) and MR data that correlated with LN metastasis. Then we developed a prediction model with these factors by multiple logistic regression analysis. The accuracy of the model was verified using three validation sets and compared with the traditional MRI method. STATISTICAL TESTS: Univariate and multivariate logistic regression. The area under the curve (AUC) value was used to quantify the diagnostic accuracy of the model. RESULTS: Eight factors (CEA, CA199, ADCmean, mriT stage, mriN stage, CRM, EMVI, and differentiation degree) were significantly associated with LN metastasis in rectal cancer patients (P<0.1). In the training set (120) and the three validation sets (69, 103, 49), the AUC values of the model were much higher than the diagnosis by MR alone (training set, 0.902 vs. 0.580; first validation set, 0.789 vs. 0.743; second validation set, 0.774 vs. 0.573; third validation set, 0.761 vs. 0.524). DATA CONCLUSION: For the diagnosis of metastatic LNs in rectal cancer patients, our proposed logistic regression model, combining clinical and MR data, demonstrated higher diagnostic efficiency than MRI alone. LEVEL OF EVIDENCE: 4 TECHNICAL EFFICACY STAGE: 2.

Cardiac Magnetic Resonance Feature Tracking: A Novel Method to Assess Left Ventricular Three-Dimensional Strain Mechanics After Chronic Myocardial Infarction.

RATIONALE AND OBJECTIVES: This study was designed to assess left ventricular deformation after chronic myocardial infarction (CMI) using cardiac magnetic resonance feature tracking (CMR-FT) technology, and analyze its relationship with left ventricular ejection fraction (LVEF) and infarcted transmurality. MATERIALS AND METHODS: Ninety-six patients with CMI and 72 controls underwent 3.0 T CMR scanning. Strain parameters were measured by dedicated software, including global peak longitudinal strain (GPLS), global peak circumferential strain (GPCS), global peak radial strain (GPRS), segmental peak longitudinal strain (PLS), peak circumferential strain (PCS), and peak radial strain (PRS). All enhanced myocardium segments were divided into subendocardial infarction (SI) and transmural infarction (TI) group. Pearson, intraclass correlation coefficient and receiver operating characteristic analysis were performed to compare the parameters' mean values between SI and TI groups. RESULTS: GPLS, GPRS, and GPCS in CMI group were significantly decreased comparing with control group. PRS and PCS in TI group were significantly lower than those in SI group, whereas no statistical difference was observed in PLS. In Pearson correlation analysis, LVEF was strongly correlated with GPLS, GPRS, and GPCS in CMI patients. Additionally, excellent reproducibility of all strain parameters was observed. In receiver operating characteristic analysis, segmental PRS and PCS might differentiate SI from TI with higher diagnostic efficiency (p < 0.05), while PLS was less valuable (p > 0.05). CONCLUSION: CMR-FT could noninvasively and quantitatively assess global and regional myocardial strain in CMI patients with excellent reproducibility and strong correlation with LVEF. Additionally, segmental myocardial strain parameters indicate potential clinical value in differentiating myocardial infarction subtype.

Dynamically constructed network with error correction for accurate ventricle volume estimation.

Automated ventricle volume estimation (AVVE) on cardiac magnetic resonance (CMR) images is very important for clinical cardiac disease diagnosis. However, current AVVE methods ignore the error correction for the estimated volume. This results in clinically intolerable ventricle volume estimation error and further leads to wrong ejection fraction (EF) assessment, which significantly limits the application potential of AVVE methods. The objective of this paper is to address this problem with AVVE and further make it more clinically applicable. We proposed a dynamically constructed network to achieve accurate AVVE. First, we introduced a novel dynamically constructed deep learning framework, that evolves a single model into a bi-model volume estimation network. In this way, the EF correlation can be built directly based on the bi-model network. Second, we proposed an error correction strategy using dynamically created residual nodes, which is based on stochastic configurations with an EF correlation constraint. Finally, we formulated the proposed method into an end-to-end joint optimization framework for accurate ventricle volume estimation with effective error correction. Experiments and comparisons on large-scale cardiac magnetic resonance datasets were carried out. Results show that the proposed method outperforms state-of-the-art methods, and has good potential for clinical application. Besides, the proposed method is the first work to achieve error correction for AVVE and also has the potential to be extended to other medical index estimation tasks.

Advanced myocardial characterization in hypertrophic cardiomyopathy: feasibility of CMR-based feature tracking strain analysis in a case-control study.

OBJECTIVES: This study aimed to evaluate the feasibility and reproducibility of using cardiovascular magnetic resonance feature tracking (CMR-FT) for analysis of bi-ventricular strain and strain rate (SR) in hypertrophic cardiomyopathy (HCM) patients as well as to explore the correlation between right ventricular (RV) and left ventricular (LV) deformation. METHODS: A total of 60 HCM patients and 48 controls were studied. Global and segmental peak values of bi-ventricular longitudinal, circumferential, radial strain, and systolic SR were analyzed. Pearson analysis was performed to investigate the correlation of RV and LV deformation. Intra-observer and inter-observer reproducibility were also assessed. RESULTS: LV mass in the HCM group was significantly higher than that in the control group. LV end-systolic and end-diastolic volume and RV end-systolic and end-diastolic volume in the HCM group were all significantly lower than the correlated parameters in the control group (p < 0.001, respectively), whereas no statistical difference was found in ejection fraction (p > 0.05). Global longitudinal strain (GLS), global longitudinal strain rate (GLSR), global circumferential strain (GCS), global circumferential strain rate (GCSR), global radial strain (GRS), and global radial strain rate (GRSR) of the LV and RV were all significantly lower than the control group, and segmental strain and SR were also true (p < 0.001, respectively). Bi-ventricular strain and SR measurements were highly reproducible at both intra- and inter-observer levels. Additionally, Pearson analysis showed RV GCS, GLS, and GRS positively correlated with LV GCS, GLS, and GRS (r = 0.713, p < 0.001; r = 0.728, p < 0.001; r = 0.730, p < 0.001, respectively). CONCLUSIONS: CMR-FT is a promising approach to analyze impairment of global and segmental myocardium deformation in HCM patients non-invasively and quantitatively. KEY POINTS: • CMR-FT allows for advanced myocardial characterization with high reproducibility. • As compared with controls, HCM patients have significant differences in CMR-FT strain analysis while ejection fraction was similar. • CMR-FT may serve as an early biomarker of HCM in subjects at risk.

Deep Atlas Network for Efficient 3D Left Ventricle Segmentation on Echocardiography.

We proposed a novel efficient method for 3D left ventricle (LV) segmentation on echocardiography, which is important for cardiac disease diagnosis. The proposed method effectively overcame the 3D echocardiography's challenges: high dimensional data, complex anatomical environments, and limited annotation data. First, we proposed a deep atlas network, which integrated LV atlas into the deep learning framework to address the 3D LV segmentation problem on echocardiography for the first time, and improved the performance based on limited annotation data. Second, we proposed a novel information consistency constraint to enhance the model's performance from different levels simultaneously, and finally achieved effective optimization for 3D LV segmentation on complex anatomical environments. Finally, the proposed method was optimized in an end-to-end back propagation manner and it achieved high inference efficiency even with high dimensional data, which satisfies the efficiency requirement of clinical practice. The experiments proved that the proposed method achieved better segmentation results and a higher inference speed compared with state-of-the-art methods. The mean surface distance, mean hausdorff surface distance, and mean dice index were 1.52 mm, 5.6 mm and 0.97 respectively. What's more, the method is efficient and its inference time is 0.02s. The experimental results proved that the proposed method has a potential clinical application for 3D LV segmentation on echocardiography.

Commensal correlation network between segmentation and direct area estimation for bi-ventricle quantification.

Accurate and automated cardiac bi-ventricle quantification based on cardiac magnetic resonance (CMR) image is a very crucial procedure for clinical cardiac disease diagnosis. Two traditional and commensal tasks, i.e., bi-ventricle segmentation and direct ventricle function index estimation, are always independently devoting to address ventricle quantification problem. However, because of inherent difficulties from the variable CMR imaging conditions, these two tasks are still open challenging. In this paper, we proposed a unified bi-ventricle quantification framework based on commensal correlation between the bi-ventricle segmentation and direct area estimation. Firstly, we proposed the area commensal correlation between the two traditional cardiac quantification tasks for the first time, and designed a novel deep commensal network (DCN) to join these two commensal tasks into a unified framework based on the proposed commensal correlation loss. Secondly, we proposed an differentiable area operator to model the proposed area commensal correlation and made the proposed model continuously differentiable. Thirdly, we proposed a high-efficiency and novel uncertainty estimation method through one-time inference based on cross-task output variability. And finally DCN achieved end-to-end optimization and fast convergence as well as uncertainty estimation with one-time inference. Experiments on the four open accessible short-axis CMR benchmark datasets (i.e., Sunnybrook, STACOM 2011, RVSC, and ACDC) showed that the proposed method achieves best bi-ventricle quantification accuracy and optimization performance. Hence, the proposed method has big potential to be extended to other medical image analysis tasks and has clinical application value.

A Combined Fully Convolutional Networks and Deformable Model for Automatic Left Ventricle Segmentation Based on 3D Echocardiography.

Segmentation of the left ventricle (LV) from three-dimensional echocardiography (3DE) plays a key role in the clinical diagnosis of the LV function. In this work, we proposed a new automatic method for the segmentation of LV, based on the fully convolutional networks (FCN) and deformable model. This method implemented a coarse-to-fine framework. Firstly, a new deep fusion network based on feature fusion and transfer learning, combining the residual modules, was proposed to achieve coarse segmentation of LV on 3DE. Secondly, we proposed a method of geometrical model initialization for a deformable model based on the results of coarse segmentation. Thirdly, the deformable model was implemented to further optimize the segmentation results with a regularization item to avoid the leakage between left atria and left ventricle to achieve the goal of fine segmentation of LV. Numerical experiments have demonstrated that the proposed method outperforms the state-of-the-art methods on the challenging CETUS benchmark in the segmentation accuracy and has a potential for practical applications.

Imaging c-Met expression using 18F-labeled binding peptide in human cancer xenografts.

OBJECTIVES: c-Met is a receptor tyrosine kinase shown inappropriate expression and actively involved in progression and metastasis in most types of human cancer. Development of c-Met-targeted imaging and therapeutic agents would be extremely useful. Previous studies reported that c-Met-binding peptide (Met-pep1, YLFSVHWPPLKA) specifically targets c-Met receptor. Here, we evaluated 18F-labeled Met-pep1 for PET imaging of c-Met positive tumor in human head and neck squamous cell carcinoma (HNSCC) xenografted mice. METHODS: c-Met-binding peptide, Met-pep1, was synthesized and labeled with 4-nitrophenyl [18F]-2-fluoropropionate ([18F]-NPFP) ([18F]FP-Met-pep1). The cell uptake, internalization and efflux of [18F]FP-Met-pep1 were assessed in UM-SCC-22B cells. In vivo pharmacokinetics, blocking and biodistribution of the radiotracers were investigated in tumor-bearing nude mice by microPET imaging. RESULTS: The radiolabeling yield for [18F]FP-Met-pep1 was over 55% with 97% purity. [18F]FP-Met-pep1 showed high tumor uptake in UM-SCC-22B tumor-bearing mice with clear visualization. The specificity of the imaging tracer was confirmed by significantly decreased tumor uptake after co-administration of unlabeled Met-pep1 peptides. Prominent uptake and rapid excretion of [18F]FP-Met-pep1 was also observed in the kidney, suggesting this tracer is mainly excreted through the renal-urinary routes. Ex vivo biodistribution showed similar results that were consistent with microPET imaging data. CONCLUSIONS: These results suggest that 18F-labeled c-Met peptide may potentially be used for imaging c-Met positive HNSCC cancer in vivo and for c-Met-targeted cancer therapy.

Multi-Views Fusion CNN for Left Ventricular Volumes Estimation on Cardiac MR Images.

OBJECTIVE: Left ventricular (LV) volume estimation is a critical procedure for cardiac disease diagnosis. The objective of this paper is to address a direct LV volume prediction task. METHODS: In this paper, we propose a direct volume prediction method based on the end-to-end deep convolutional neural networks. We study the end-to-end LV volume prediction method in items of the data preprocessing, network structure, and multiview fusion strategy. The main contributions of this paper are the following aspects. First, we propose a new data preprocessing method on cardiac magnetic resonance (CMR). Second, we propose a new network structure for end-to-end LV volume estimation. Third, we explore the representational capacity of different slices and propose a fusion strategy to improve the prediction accuracy. RESULTS: The evaluation results show that the proposed method outperforms other state-of-the-art LV volume estimation methods on the open accessible benchmark datasets. The clinical indexes derived from the predicted volumes agree well with the ground truth ( ${\rm{EDV:R}}^{{\rm 2}}={\text{0.974}}$, ${\rm{RMSE\,}}= {\text{9.6}}{\rm{\,ml}}$; ${\rm{ESV:R}}^{{\rm 2}}={\text{0.976}}$, ${\rm{RMSE}}= {\text{7.1}}\,{\text{ml}}$; ${\rm{EF:R}}^{{\rm 2}} ={\text{0.828}}$, ${\rm{RMSE}}= {\text{4.71}}\% $). CONCLUSION: Experimental results prove that the proposed method may be useful for the LV volume prediction task. SIGNIFICANCE: The proposed method not only has application potential for cardiac diseases screening for large-scale CMR data, but also can be extended to other medical image research fields.