Deep Learning Subtraction Angiography: Improved Generalizability with Transfer Learning.

PURPOSE: To investigate the utility and generalizability of deep learning subtraction angiography (DLSA) for generating synthetic digital subtraction angiography (DSA) images without misalignment artifacts. MATERIALS AND METHODS: DSA images and native digital angiograms of the cerebral, hepatic, and splenic vasculature, both with and without motion artifacts, were retrospectively collected. Images were divided into a motion-free training set (n = 66 patients, 9,161 images) and a motion artifact-containing test set (n = 22 patients, 3,322 images). Using the motion-free set, the deep neural network pix2pix was trained to produce synthetic DSA images without misalignment artifacts directly from native digital angiograms. After training, the algorithm was tested on digital angiograms of hepatic and splenic vasculature with substantial motion. Four board-certified radiologists evaluated performance via visual assessment using a 5-grade Likert scale. Subgroup analyses were performed to analyze the impact of transfer learning and generalizability to novel vasculature. RESULTS: Compared with the traditional DSA method, the proposed approach was found to generate synthetic DSA images with significantly fewer background artifacts (a mean rating of 1.9 [95% CI, 1.1-2.6] vs 3.5 [3.5-4.4]; P = .01) without a significant difference in foreground vascular detail (mean rating of 3.1 [2.6-3.5] vs 3.3 [2.8-3.8], P = .19) in both the hepatic and splenic vasculature. Transfer learning significantly improved the quality of generated images (P < .001). CONCLUSIONS: DLSA successfully generates synthetic angiograms without misalignment artifacts, is improved through transfer learning, and generalizes reliably to novel vasculature that was not included in the training data.

Paracentesis: Faster and easier using the RenovaRP® pump.

PURPOSE Paracentesis is commonly performed in interventional radiology practice, and large volume paracentesis (LVP) using wall suction can take up to an hour to complete, placing significant stress on room and resource time. As the number of LVP procedures performed by Interventional Radiologists continue to increase, this study was undertaken to analyze the impact of the RenovaRP® Paracentesis Management System (GI Supply) on procedure time and patient satisfaction. METHODS Between March 9, 2020 and May 29, 2020, procedural data and patient satisfaction was collected as part of a practice quality improvement project and retrospectively analyzed on 39 sequential paracenteses performed with wall suction prior to acquiring the RenovaRP® system and subsequently on 42 paracenteses performed with use of the device. RESULTS A substantially higher fluid flow rate was found using the RenovaRP® system compared to wall suction, 237.2 mL/min vs. 108.6 mL/min (P < .001). This resulted in a significant decrease in procedure room time from 53 min to 31 min (P < .001). There was associated improvement in the patient experience during paracentesis. CONCLUSION The RenovaRP® decreases procedure time for LVP with improvement in the patient experience during paracentesis.

Deep Vein Thrombosis Due to Fecal Impaction.

BACKGROUND: Deep vein thrombosis (DVT) can occasionally be secondary to mechanical compressions, such as in May-Thurner syndrome. To our knowledge, no cases of DVT associated with mechanical compression by fecal impaction in a pediatric patient have been previously reported. CASE PRESENTATION: An 18-year-old developmentally delayed female presented to the emergency department with swelling involving her left lower extremity. Cross-sectional imaging revealed a significant stool burden in the rectosigmoid colon compressing the left external iliac vein and causing secondary thrombosis. Thrombectomy successfully alleviated the clot burden. CONCLUSION: This case underscores the potentially serious implications of severe fecal impaction in the pediatric and adolescent populations.

Radiographic Local Tumor Control and Pain Palliation of Sarcoma Metastases within the Musculoskeletal System with Percutaneous Thermal Ablation.

PURPOSE: To evaluate the effectiveness of percutaneous image-guided thermal ablation in achieving local tumor control and pain palliation of sarcoma metastases within the musculoskeletal system. MATERIALS AND METHODS: Retrospective review of 64 sarcoma metastases within the musculoskeletal system in 26 women and 15 men (total = 41) treated with ablation between December 2011 and August 2016 was performed. Mean age of the cohort was 42.9 years ± 16.0 years. Two subgroups were treated: oligometastatic disease (n = 13) and widely metastatic disease (n = 51). A variety of sarcoma histologies were treated with average tumor volume of 42.5 cm3 (range 0.1-484.7 cm3). Pain scores were recorded before and 4 weeks after therapy for 59% (38/64) of treated lesions. Follow-up imaging was evaluated for local control and to monitor sites of untreated disease as an internal control. Fifty-eight percent (37/64) were lost to imaging follow-up at varying time points over a year. Complication rate was 5% (3/64; one minor and two major events). RESULTS: One-year local tumor control rates were 70% (19/27) in all patients, 67% (12/18) in the setting of progression of untreated metastases, and 100% (10/10) in the setting of oligometastatic disease. Median pain scores decreased from 8 (interquartile range 5.0-9.0) to 3 (interquartile range 0.1-4.0) 1 month after the procedure (P < 0.001). CONCLUSION: Image-guided percutaneous ablation is an effective option for local tumor control and pain palliation of metastatic sarcomas within the musculoskeletal system. Treatment in the setting of oligometastatic disease offers potential for remission. LEVEL OF EVIDENCE: Level 4, Retrospective Review.