scholarly journals Development of Novel Deep Multimodal Representation Learning-based Model for the Differentiation of Liver Tumors on B-Mode Ultrasound Images

2021 ◽  
Author(s):  
Masaya Sato ◽  
Tamaki Kobayashi ◽  
Yoko Soroida ◽  
Takashi Tanaka ◽  
Takuma Nakatsuka ◽  
...  
Keyword(s):  
Diagnostic Accuracy ◽  
Diagnostic Performance ◽  
Liver Tumors ◽  
Malignant Tumors ◽  
Representation Learning ◽  
Ultrasound Images ◽  
Blood Biomarkers ◽  
Liver Nodules ◽  
Representation Model ◽  
Multimodal Representation

Abstract Recently, multimodal representation learning for images and other information such as numbers or language has gained much attention due to the possibility of combining latent features using a single distribution. The aim of the current study was to analyze the diagnostic performance of deep multimodal representation model-based integration of tumor image, patient background, and blood biomarkers for the differentiation of liver tumors observed using B-mode ultrasonography (US). First, we applied supervised learning with a convolutional neural network (CNN) to 972 liver nodules in the training and development sets (479 benign and 493 malignant nodules), to develop a predictive model using segmented B-mode tumor images. Additionally, we also applied a deep multimodal representation model to integrate information about patient background or blood biomarkers to B-mode images. We then investigated the performance of the models in an independent test set of 108 liver nodules, including 53 benign and 55 malignant tumors. Using only the segmented B-mode images, the diagnostic accuracy and area under the curve (AUC) values were 68.52% and 0.721, respectively. As the information about patient background such as age or sex and blood biomarkers was integrated, the diagnostic performance increased in a stepwise manner. The diagnostic accuracy and AUC value of the multimodal DL model (which integrated B-mode tumor image, patient age, sex, AST, ALT, platelet count, and albumin data) reached 96.30% and 0.994, respectively. Integration of patient background and blood biomarkers in addition to US image using multimodal representation learning outperformed the CNN model using US images. We expect that the deep multimodal representation model could be a feasible and acceptable tool that can effectively support the definitive diagnosis of liver tumors using B-mode US in daily clinical practice.

scholarly journals Intraoperative frozen section analysis of ovarian tumors: a 11-year review of accuracy with clinicopathological correlation in a Hong Kong Regional hospital

2019 ◽  
Vol 29 (4) ◽  
pp. 772-778 ◽  
Author(s):  
Fred Yau-Lung Kung ◽  
Alex Koon-ho Tsang ◽  
Ellen Lok-man Yu
Keyword(s):  
Regression Analysis ◽  
Diagnostic Accuracy ◽  
Diagnostic Performance ◽  
Malignant Tumors ◽  
Early Stage ◽  
Frozen Section ◽  
Ovarian Tumors ◽  
Stage I ◽  
Clinicopathological Parameters ◽  
Staging Surgery

ObjectiveIntra-operative frozen section (IFS) can provide an instinct guide for treatment of ovarian tumors intra-operatively, though limitations exist. This study intended to evaluate the diagnostic performance of IFS and possible clinicopathological factors influencing the diagnostic accuracy of IFS.MethodsA retrospective review of IFS of ovarian lesions from 2006 to 2016 was done. The diagnostic performance of benign, borderline, and malignant IFS diagnosis was evaluated. Logistic regression analysis was used to assess the influence of clinicopathological parameters on the likelihood of underdiagnosis.ResultsThere were 1143 consecutive cases during the study period. The overall accuracy was 93.7%. For benign diagnoses, the IFS diagnostic accuracy, sensitivity, and specificity were 97.20%, 100%, and 92.51%, respectively. If borderline and malignant diagnoses were considered as a single group, the IFS diagnostic accuracy was 97.20%, with 92.51% sensitivity and 100% specificity. At univariate regression analysis, intact capsules at time of delivery (ORunadj = 1.9), stage I lesions (ORunadj = 3.76) and ultrasound (USG) score 0 (ORunadj = 2.52) were positively associated with underdiagnosis. Further multivariate analysis showed that only stage I lesions (OR = 3.62) and USG score 0 (OR = 2.32) were positively associated with underdiagnosis. For the cases with underdiagnosed IFS, 54% (34/63) received incomplete primary staging surgery.ConclusionsThe study demonstrated that IFS provided excellent specificity to differentiate borderline or malignant tumors from benign lesions. IFS in early-stage ovarian cancers needs to be interpreted with caution, though IFS is most important for this group of lesions. A reliable IFS diagnosis often requires efficient communication between surgeons and pathologists.

scholarly journals The value of S-Detect in improving the diagnostic performance of radiologists for the differential diagnosis of thyroid nodules

2020 ◽  
Vol 22 (4) ◽  
pp. 415
Author(s):  
Qi Wei ◽  
Shu-E Zeng ◽  
Li-Ping Wang ◽  
Yu-Jing Yan ◽  
Ting Wang ◽  
...  
Keyword(s):  
Diagnostic Accuracy ◽  
Predictive Value ◽  
Diagnostic Performance ◽  
Thyroid Nodules ◽  
Characteristic Curve ◽  
Diagnostic Value ◽  
Ultrasound Images ◽  
Experienced Radiologist ◽  
Before And After ◽  
Sensitivity Specificity

Aims: To compare the diagnostic value of S-Detect (a computer aided diagnosis system using deep learning) in differentiating thyroid nodules in radiologists with different experience and to assess if S-Detect can improve the diagnostic performance of radiologists.Materials and methods: Between February 2018 and October 2019, 204 thyroid nodules in 181 patients were included. An experienced radiologist performed ultrasound for thyroid nodules and obtained the result of S-Detect. Four radiologists with different experience on thyroid ultrasound (Radiologist 1, 2, 3, 4 with 1, 4, 9, 20 years, respectively) analyzed the conventional ultrasound images of each thyroid nodule and made a diagnosis of “benign” or “malignant” based on the TI-RADS category. After referring to S-Detect results, they re-evaluated the diagnoses. The diagnostic performance of radiologists was analyzed before and after referring to the results of S-Detect.Results: The accuracy, sensitivity, specificity, positive predictive value and negative predictive value of S-Detect were 77.0, 91.3, 65.2, 68.3 and 90.1%, respectively. In comparison with the less experienced radiologists (radiologist 1 and 2), S-Detect had a higher area under receiver operating characteristic curve (AUC), accuracy and specificity (p <0.05). In comparison with the most experienced radiologist, the diagnostic accuracy and AUC were lower (p<0.05). In the less experienced radiologists, the diagnostic accuracy, specificity and AUC were significantly improved when combined with S-Detect (p<0.05), but not for experienced radiologists (radiologist 3 and 4) (p>0.05).Conclusions: S-Detect may become an additional diagnostic method for the diagnosis of thyroid nodules and improve the diagnostic performance of less experienced radiologists. 

scholarly journals Feasibility and diagnostic reliability of quantitative flow ratio in the assessment of non-culprit lesions in acute coronary syndrome

2021 ◽  
Author(s):  
Aslihan Erbay ◽  
Lisa Penzel ◽  
Youssef S. Abdelwahed ◽  
Jens Klotsche ◽  
Anne-Sophie Schatz ◽  
...  
Keyword(s):  
Diagnostic Accuracy ◽  
Diagnostic Performance ◽  
Primary Pci ◽  
Flow Ratio ◽  
Diagnostic Reliability ◽  
Index Procedure ◽  
Coronary Syndromes ◽  
Quantitative Flow Ratio ◽  
Quantitative Flow ◽  
Good Agreement

AbstractSeveral studies have demonstrated the feasibility and safety of hemodynamic assessment of non-culprit coronary arteries in setting of acute coronary syndromes (ACS) using fractional flow reserve (FFR) measurements. Quantitative flow ratio (QFR), recently introduced as angiography-based fast FFR computation, has been validated with good agreement and diagnostic performance with FFR in chronic coronary syndromes. The aim of this study was to assess the feasibility and diagnostic reliability of QFR assessment during primary PCI. A total of 321 patients with ACS and multivessel disease, who underwent primary PCI and were planned for staged PCI of at least one non-culprit lesion were enrolled in the analysis. Within this patient cohort, serial post-hoc QFR analyses of 513 non-culprit vessels were performed. The median time interval between primary and staged PCI was 49 [42–58] days. QFR in non-culprit coronary arteries did not change between acute and staged measurements (0.86 vs 0.87, p = 0.114), with strong correlation (r = 0.94, p ≤ 0.001) and good agreement (mean difference -0.008, 95%CI -0.013–0.003) between measurements. Importantly, QFR as assessed at index procedure had sensitivity of 95.02%, specificity of 93.59% and diagnostic accuracy of 94.15% in prediction of QFR ≤ 0.80 at the time of staged PCI. The present study for the first time confirmed the feasibility and diagnostic accuracy of non-culprit coronary artery QFR during index procedure for ACS. These results support QFR as valuable tool in patients with ACS to detect further hemodynamic relevant lesions with excellent diagnostic performance and therefore to guide further revascularisation therapy.

scholarly journals Fully Data-Driven Pseudohealthy Synthesis for Planning Valve-Sparing Aortic Root Reconstruction using Conditional Variational Autoencoders

2020 ◽  
Vol 6 (3) ◽  
pp. 284-287
Author(s):  
Jannis Hagenah ◽  
Mohamad Mehdi ◽  
Floris Ernst
Keyword(s):  
Aortic Root ◽  
Similarity Index ◽  
Ground Truth ◽  
Representation Learning ◽  
Patient Specific ◽  
Ultrasound Images ◽  
Specific Geometry ◽  
The Individual ◽  
Native Root ◽  
Original Information

AbstractAortic root aneurysm is treated by replacing the dilated root by a grafted prosthesis which mimics the native root morphology of the individual patient. The challenge in predicting the optimal prosthesis size rises from the highly patient-specific geometry as well as the absence of the original information on the healthy root. Therefore, the estimation is only possible based on the available pathological data. In this paper, we show that representation learning with Conditional Variational Autoencoders is capable of turning the distorted geometry of the aortic root into smoother shapes while the information on the individual anatomy is preserved. We evaluated this method using ultrasound images of the porcine aortic root alongside their labels. The observed results show highly realistic resemblance in shape and size to the ground truth images. Furthermore, the similarity index has noticeably improved compared to the pathological images. This provides a promising technique in planning individual aortic root replacement.

scholarly journals Correctable biliary atresia and cholangiocarcinoma: a case report of a 63-year-old patient

2019 ◽  
Vol 5 (1) ◽  
Author(s):  
Masaki Nio ◽  
Motoshi Wada ◽  
Hideyuki Sasaki ◽  
Hiromu Tanaka ◽  
Masatoshi Hashimoto ◽  
...  
Keyword(s):  
Biliary Atresia ◽  
Liver Tumors ◽  
Malignant Tumors ◽  
Porta Hepatis ◽  
Case Presentation ◽  
Cancer Occurrence ◽  
Long Term Survivors ◽  
Near Future

Abstract Background Although cancer occurrence following surgery for biliary atresia has gradually increased, the development of cholangiocarcinoma in a native liver survivor of biliary atresia is extremely rare. Case presentation A 3-month-old female patient with the correctable type of biliary atresia underwent a cystoduodenostomy. At 16 years of age, she underwent multiple surgeries including lysis of intestinal adhesions, ileostomy, and gastrojejunostomy at another hospital. At 54 years of age, she underwent lithotomy at the porta hepatis, resection of the residual cystic bile duct with gallbladder, and hepaticojejunostomy in Roux-en-Y fashion. As she approached the age of 63, her computed tomography scan showed no liver tumors. In the following year, she developed cholangiocarcinoma at the porta hepatis and underwent chemotherapy. However, the cancer progressed, and she died before she reached the age of 64 years. Conclusions Cholangiocarcinoma is extremely rare in patients with biliary atresia. However, physicians should follow up patients with biliary atresia as closely as possible, as malignant tumors secondary to biliary atresia may increase in number in the near future because of the growing number of long-term survivors with biliary atresia.

Assessing the Performance of a Novel Point-of-Care Qualitative Assay for Early Diagnosis of Acute Coronary Syndrome

Cardiology ◽  
2020 ◽  
pp. 1-8
Author(s):  
Ronny Alcalai ◽  
Boris Varshisky ◽  
Ahmad Marhig ◽  
David Leibowitz ◽  
Larissa Kogan-Boguslavsky ◽  
...  
Keyword(s):  
Acute Coronary Syndrome ◽  
Chest Pain ◽  
Diagnostic Accuracy ◽  
Diagnostic Performance ◽  
Troponin I ◽  
Point Of Care ◽  
Final Diagnosis ◽  
Cardiac Biomarkers ◽  
Fatty Acid Binding ◽  
Coronary Syndrome

<b><i>Background:</i></b> Early and accurate diagnosis of acute coronary syndrome (ACS) is essential for initiating lifesaving interventions. In this article, the diagnostic performance of a novel point-of-care rapid assay (SensAheart<sup>©</sup>) is analyzed. This assay qualitatively determines the presence of 2 cardiac biomarkers troponin I and heart-type fatty acid-binding protein that are present soon after onset of myocardial injury. <b><i>Methods:</i></b> We conducted a prospective observational study of consecutive patients who presented to the emergency department with typical chest pain. Simultaneous high-sensitive cardiac troponin T (hs-cTnT) and SensAheart testing was performed upon hospital admission. Diagnostic accuracy was computed using SensAheart or hs-cTnT levels versus the final diagnosis defined as positive/negative. <b><i>Results:</i></b> Of 225 patients analyzed, a final diagnosis of ACS was established in 138 patients, 87 individuals diagnosed with nonischemic chest pain. In the overall population, as compared to hs-cTnT, the sensitivity of the initial SensAheart assay was significantly higher (80.4 vs. 63.8%, <i>p</i> = 0.002) whereas specificity was lower (78.6 vs. 95.4%, <i>p</i> = 0.036). The overall diagnostic accuracy of SensAheart assay was similar to the hs-cTnT (82.7% compared to 76.0%, <i>p</i> = 0.08). <b><i>Conclusions:</i></b> Upon first medical contact, the novel point-of-care rapid SensAheart assay shows a diagnostic performance similar to hs-cTnT. The combination of 2 cardiac biomarkers in the same kit allows for very early detection of myocardial damage. The SensAheart assay is a reliable and practical tool for ruling-in the diagnosis of ACS.

Sign in / Sign up

Export Citation Format

Share Document