Evaluation of mediastinal lymph node segmentation of heterogeneous CT data with full and weak supervision

Accurate lymph node size estimation is critical for staging cancer patients, initial therapeutic management, and assessing response to therapy. Current standard practice for quantifying lymph node size is based on a variety of criteria that use uni-directional or bi-directional measurements. Segmentation in 3D can provide more accurate evaluations of the lymph node size. Fully convolutional neural networks (FCNs) have achieved state-of-the-art results in segmentation for numerous medical imaging applications, including lymph node segmentation. Adoption of deep learning segmentation models in clinical trials often faces numerous challenges. These include lack of pixel-level ground truth annotations for training, generalizability of the models on unseen test domains due to the heterogeneity of test cases and variation of imaging parameters. In this paper, we studied and evaluated the performance of lymph node segmentation models on a dataset that was completely independent of the one used to create the models. We analyzed the generalizability of the models in the face of a heterogeneous dataset and assessed the potential effects of different disease conditions and imaging parameters. Furthermore, we systematically compared fully-supervised and weakly-supervised methods in this context. We evaluated the proposed methods using an independent dataset comprising 806 mediastinal lymph nodes from 540 unique patients. The results show that performance achieved on the independent test set is comparable to that on the training set. Furthermore, neither the underlying disease nor the heterogeneous imaging parameters impacted the performance of the models. Finally, the results indicate that our weakly-supervised method attains 90%- 91% of the performance achieved by the fully supervised training.

Overview of imaging findings associated with systemic therapies in advanced epithelial ovarian cancer

PURPOSE

To provide an overview for radiologists of the systemic agents used in the treatment of advanced epithelial ovarian cancer (EOC) and their associated toxicities.

RESULTS

EOC is a common gynecological malignancy, with the majority of patients presenting with advanced stage disease at the time of diagnosis. Although primary cytoreductive surgery and chemotherapy are the principal treatments for EOC, recurrence rates of disease remain high. As several molecular targeted therapies have been developed in the last decade, various novel agents have shown efficacy in the treatment of advanced EOC. Advanced EOC will be discussed by outlining the relevant radiological features of toxicities.

CONCLUSION

Knowledge of the systemic therapies utilized in the treatment of advanced EOC and their associated radiological features is critical in diagnostic image interpretation.

Overview of systemic treatment in recurrent and advanced cervical cancer: a primer for radiologists

Imaging has a central role in surveillance of cervical cancer, guiding decision on when to initiate treatment for recurrent disease and to guide management in advanced cervical cancer. Due to the increased availability of pelvic radiation therapy, the rate of atypical presentation of recurrent disease has increased. Simultaneously, the array of systemic therapies now available for advanced cervical cancer has considerably expanded in the last few years, with therapies now available in mid and low-income countries. While pelvic recurrences are amenable of loco-regional treatment, recurrent disease may present with metastases to the thoracoabdominal organs, lymph nodes, bones, skin and brain, for which systemic treatment represent the standard of care. Besides combined chemotherapy regimens, alternative chemotherapies, biosimilars and immune checkpoint inhibitors are now available, each associated with a definite pattern of response and toxicity. In this review, after describing the typical and atypical presentations of recurrent and advanced cervical carcinoma on cross-sectional imaging, we will discuss systemic treatment for recurrent or advanced disease and their associated radiographic sequelae, in light of the newly available therapies.

Review of targeted therapy in chronic lymphocytic leukemia: what a radiologist needs to know about CT interpretation

The last 5 years have been marked by profound innovation in the targeted treatment of chronic lymphocytic leukemia (CLL) and indolent lymphomas. Using CLL as a case study, we present a timeline and overview of the current treatment landscape for the radiologist, including an overview of clinical and radiological features of CLL, discussion of the targeted agents themselves, and the role of imaging in response and toxicity assessment. The goal is to familiarize the radiologist with multiple Food and Drug Administration (FDA)-approved targeted agents used in this setting and associated adverse events which are commonly observed in this patient population.

Immune-related tumour response assessment criteria: a comprehensive review

Growing emphasis on precision medicine in oncology has led to increasing use of targeted therapies that encompass a spectrum of drug classes including angiogenesis inhibitors, immune modulators, signal transduction inhibitors, DNA damage modulators, hormonal agents etc. Immune therapeutic drugs constitute a unique group among the novel therapeutic agents that are transforming cancer treatment, and their use is rising. The imaging manifestations in patients on immune therapies appear to be distinct from those typically seen with conventional cytotoxic therapies. Patients on immune therapies may demonstrate a delayed response, transient tumour enlargement followed by shrinkage, stable size, or initial appearance of new lesions followed by stability or response. These newer patterns of response to treatment have rendered conventional criteria such as World Health Organization and response evaluation criteria in solid tumours suboptimal in monitoring changes in tumour burden. As a consequence, newer imaging response criteria such as immune-related response evaluation criteria in solid tumours and immune-related response criteria are being implemented in many trials to effectively monitor patients on immune therapies. In this review, we discuss the traditional and new imaging response criteria for evaluation of solid tumours, review the outcomes of various articles which compared traditional criteria with the new immune-related criteria and discuss pseudo-progression and immune-related adverse events.

Ovarian Cancer: An Evidence-Based, Easy-to-Use Prediction Rule to Optimize the Use of Follow-up Chest CT

PURPOSE

To create and validate an evidence-based prediction rule to optimize use of follow-up chest CT for ovarian cancer.

METHODS

In this Institutional Review Board-approved retrospective study performed at two academic medical centers, electronic medical records from January through December 2013 at center 1 (USA) and January 2012 through December 2013 at center 2 (South Korea) were searched to identify consecutive chest CTs performed within 5 years of initial cytoreductive surgery in patients with pathologically proven ovarian cancer. Three separate study cohorts were created: cohort 1, 316 CTs (in 150 patients) with high-grade serous ovarian cancer (HGSC) from center 1; cohort 2, 374 CTs (81 patients) with HGSC from center 2; and cohort 3, 87 CTs (56 patients) with non-HGSC histologies from center 1. A radiologist blinded to outcome of CT, using a prediction rule that utilized previously available information, categorized each CT into "high-risk" (stage 4 at presentation and/or preexisting abdominal disease [disease below diaphragmatic dome, visualized on abdominal CT]) or "low-risk" (neither of above). A blinded radiologist then reviewed chest CTs in random order to record thoracic metastases above the diaphragmatic dome, and outcome was compared with prediction rule risk category.

RESULTS

Among the three cohorts and in the total population, the prediction rule identified 94 of 316 (30%), 170 of 374 (45%), 53 of 87 (61%), and 317 of 777 (41%) CTs as "low-risk," respectively. The sensitivity, specificity, positive predictive value, negative predictive value, positive likelihood ratio, and negative likelihood ratio were as follows: cohort 1: 95%, 35%, 24%, 97%, 1.46, 0.14; cohort 2: 88%, 53%, 29%, 95%, 1.87, 0.22; cohort 3: 88%, 66%, 21%, 98%, 2.59, 0.18; total population: 91%, 47%, 26%, 96%, 1.72, 0.19. False-negative rate in the three cohorts and in total population was 3 of 94 (3%), 8 of 170 (5%), 1 of 53 (2%), and 12 of 317 (4%); however, in each of these cases there was concurrent new abdominal disease.

CONCLUSIONS

The easy-to-use prediction rule helps avoid unnecessary chest CTs in patients with ovarian cancer with high sensitivity and negative predictive value, and with minimal risk of missing thoracoabdominal metastases.

Patterns and Prognostic Importance of Hepatic Involvement in Patients with Serous Ovarian Cancer: A Single-Institution Experience with 244 Patients

Purpose To evaluate the frequency, patterns, and prognostic importance of metastatic hepatic involvement in serous ovarian cancer. Materials and Methods This institutional review board-approved retrospective study, with waived informed consent, included 244 patients with pathologically proven serous ovarian cancer (mean age ± standard deviation, 59 years ± 10.7; range 19-93 years). Electronic medical records and all available imaging studies over a median follow-up of 44 months (interquartile range [IQR], 27-70) were reviewed to identify the frequency of liver parenchymal invasion (LPI) from perihepatic peritoneal metastasis and hematogenous liver metastases. The associations and prognostic importance of LPI and hematogenous metastases were studied by using univariate and multivariate Cox proportional analysis. Results Eighty-four of 244 patients (34%) developed perihepatic metastases, of whom 55 (23%) developed LPI after median of 43 months (IQR, 25-63). Hematogenous hepatic metastases developed in 38 of 244 patients (16%) after median of 42 months (IQR, 26-64). At multivariate analysis, age (P = .008; hazard ratio [HR]: 1.03; 95% confidence interval [CI]: 1.009, 1.07) and suboptimal cytoreduction (P = .03; HR, 2.13; 95% CI: 1.12, 4.07) were associated with LPI. Increasing age (P = .01; HR, 1.04; 95% CI: 1.008, 1.08), high-grade tumor (P = .01; HR, 6.75; 95% CI: 1.44, 120.5), and advanced stage (P = .03; HR, 3.16; 95% CI: 1.94, 4.56) were associated with hematogenous metastases. Overall survival with and without LPI was similar (median, 80 months; IQR, 50-not reached vs 123 months; IQR, 49-279; P = .6). Hematogenous metastases were associated with significantly shorter survival at univariate (median 63 months, IQR 43-139 vs 145 months, IQR 50-not reached; P = .006) and multivariate analyses (P = .03; HR, 1.88; 95% CI: 1.14, 3.28). Conclusion Differentiating hematogenous metastases and LPI is important for radiologists; hematogenous metastases are associated with shorter survival, while LPI does not adversely affect survival and prognostically behaves like peritoneal disease. ^© RSNA, 2016.

Image-guided ovarian mass biopsy: efficacy and safety

PURPOSE

Image-guided needle biopsy represents a minimally invasive method for pathologic diagnosis of a mass. This study evaluates the diagnostic yield, accuracy, and safety of ovarian mass biopsy with combined core and fine-needle technique.

MATERIALS AND METHODS

Medical records of all women at least 18 years of age, referred from gynecologic oncology, who underwent image-guided ovarian mass biopsy from 2001 through 2011 were reviewed. Among 27 patients, ultrasound guidance was used in 13 (48%), six transabdominal and seven transvaginal; computed tomography guidance was used in 14 (52%), nine transabdominal and five transgluteal. Biopsy indications were suspected metastasis (n = 15; 56%), suspected ovarian cancer to be treated with neoadjuvant chemotherapy (n = 10; 37%), and relative contraindication to surgery (n = 2; 7%). Mean maximum lesion dimension was 9.9 cm (range, 2-23 cm), with solid composition in nine (33%), cystic in six (22%), and mixed in 12 (44%). Biopsy pathologic findings were compared versus those of the surgical specimen or, for masses that were not resected, versus the stability of benign masses and response to chemotherapy of malignant masses on follow-up.

RESULTS

All biopsies yielded a diagnosis. No biopsy-related complications were noted. Eleven patients (41%) did not undergo lesion resection and were followed for an average of 28.8 months (range, 0.3-118.4 mo). In no patient did malignancy develop during clinical follow-up after a benign biopsy diagnosis. Sensitivity and specificity for diagnosis of malignancy were 100% ± 0 (19 of 19) and 88% ± 26 (seven of eight), respectively, for cancer detection. In nine patients (33%) with final pathologic diagnosis of epithelial ovarian cancer, tumor seeding was not observed during a mean follow-up of 44.6 months (range, 1.3-110.2 mo).

CONCLUSIONS

Image-guided ovarian mass core needle biopsy results in a pathologic diagnosis of benign and malignant masses with high yield, accuracy, and safety.

Inferior vena cava filters for the cerebrovascular patient

Acute cerebrovascular disease is often complicated by deep venous thrombosis and pulmonary embolism. Many of these patients are at high risk of intracranial hemorrhage with therapeutic anticoagulation. These patients may benefit from insertion of inferior vena cava filters. Studies specifically dealing with stroke patients are lacking, but it is the authors' opinion that filters reduce the incidence of pulmonary embolism. There is little evidence to support the use of these devices prophylactically in patients who do not have venous thromboembolism. Retrievable filters are an attractive option but there are concerns about their safety; and if regularly used, a system for successful filter retrieval in all patients should be instituted. The role of concurrent anticoagulation with filters is not clear. However, we believe anticoagulation, in the absence of a contraindication, is beneficial in patients with active venous thromboembolism.

Stent-Assisted Coil Embolization of an Intraparenchymal Renal Artery Aneurysm in a Patient With Neurofibromatosis

True renal artery aneurysms are rare. They are generally asymptomatic, however, a few may present with hypertension, rupture, or renal dysfunction secondary to distal embolization. Indications for intervention include aneurysm of ≥2.0 cm in diameter, renovascular hypertension, enlarging aneurysm, associated dissection /rupture, and aneurysms in women of child-bearing age/ pregnancy. Endovascular therapy through coil embolization or stent graft exclusion is the recommended management. Coil embolization of the first and second order branch aneurysms is often associated with distal parenchymal loss and current stent graft technology prohibits use of these endoprostheses in the branch renal arteries. In this report, we describe successful stent-assisted coil embolization of an intraparenchymal aneurysm while preserving the distal parenchyma in a young woman with Neurofibromatosis type 1.

Diagnosing deep vein thrombosis

In this article, we discuss the approach for diagnosing deep vein thrombosis (DVT) in different patient populations. Clinical features and probability assessment guide further diagnostic tests. D-dimer testing is used as screening test; however, duplex ultrasound remains the primary confirmatory test. Computed tomography and magnetic resonance imaging are used only in select patient populations, such as when ultrasound results are equivocal, in patients suspected of central venous DVT, or as a part of combined protocol for diagnosis of pulmonary embolism. Contrast phlebography and plethysmography do not have much of a role during routine diagnosis of DVT.