Machine learning techniques, also known as artificial intelligence (AI), is about to dramatically change workflow and diagnostic capabilities in diagnostic radiology. The interest in AI in Interventional Radiology is rapidly gathering pace. With this early interest in AI in procedural medicine, IR could lead the way to AI research and clinical applications for all interventional medical fields. This review will address an overview of machine learning, radiomics and AI in the field of interventional radiology, enumerating the possible applications of such techniques, while also describing techniques to overcome the challenge of limited data when applying these techniques in interventional radiology. Lastly, this review will address common errors in research in this field and suggest pathways for those interested in learning and becoming involved about AI.

Background: As an adjunct to non-operative management, splenic artery embolization (SAE) has been increasingly utilized throughout the world and is now the standard of care for hemodynamically stable patients. This study aimed to retrospectively assess the rate of splenic salvage and complications after SAE for blunt trauma at a level 1 trauma center using the 2018 update to the AAST criteria, and further sub-stratify the role of angiography in AAST grade III injuries with significant hemoperitoneum. All patients between 1 January 2009 and 1 January 2019 who underwent blunt trauma and proceeded to embolization were included. Data was collected concerning initial injury grade, location of embolization, type of embolic material used, complications, and need for subsequent splenectomy. Technical success was defined as successful angiographic occlusion of the target artery at the conclusion of embolization. Clinical success was defined as splenic salvage at discharge. Vascular lesions were characterized including those with active bleeding, pseudoaneurysm, and arterio-venous fistula. Results: Two hundred thirty-two patients were included in the study. Treatments were performed at a median of 0 days (range 0–28 days) and the median AAST grade was IV (range III-V). Technical success was achieved in all patients. There were 13 complications (5.6%) consisting of re-bleed (9, 3.9%), infarction (3, 1.3%), and access site haematoma (1, 0.43%). Clinical success was achieved in 97% of patients with 7 patients requiring splenectomy after SAE (3.0%) at a median time of 4 days (range 0–17 days). Angiography in patients with grade III injuries identified 18 occult vascular injuries not identified at initial CT (p < 0.0001). Conclusions: The SPLEEN-IN study shows that treatment of intermediate-high grade blunt force traumatic splenic injuries using SAE resulted in a low rate of complication and splenic salvage in 97% of patients, providing a safe and effective treatment in stable patients. In addition, angiography of grade III injuries identified occult vascular lesions and may warrant treatment of select patients in this cohort. Level of evidence: Level 3.

Introduction: Incorporating artificial intelligence (AI) in diagnostic medical imaging reports has the potential to improve efficiency. Although perception of radiologists, radiographers, medical students and patients on AI use in image reporting has been explored, there is limited literature on non‐radiologist clinicians' opinion on this topic.Method: Single‐centre online survey targeting non‐radiologist medical staff conducted from May to August 2021 at a tertiary referral hospital in Melbourne, Australia. Survey questions revolved around clinicians' level of comfort acting on AI‐generated reports with varying levels of radiologist involvement and scan complexity, opinion on medicolegal responsibility for erroneous AI‐issued reports and perception of data privacy and security.Results: Eighty‐eight responses were collected, including 47.9% of consultants. Non‐radiologist clinicians across all seniorities and specialties felt significantly less comfortable acting on AI‐issued reports compared with radiologist‐issued reports (mean comfort radiologist 6.44/7, mean comfort AI 3.35/7, P < 0.001) but felt equally comfortable with an AI‐hybrid model of care (mean comfort hybrid 6.38/7, P = 0.676). Non‐radiologist clinicians believed that medicolegal responsibility with errors in AI‐issued reports mostly lay with hospitals or health service providers (65.9%) and radiologists (54.5%). Regarding data privacy and security, non‐radiologist clinicians felt significantly less comfortable with AI issuing image reports instead of radiologists (P < 0.001).Conclusion: A hybrid AI‐generated radiologist‐confirmed method of image reporting may be the ideal way of integrating AI into clinical practice based on the perception of our referring non‐radiologist medical colleagues. Formal guidelines on medicolegal responsibility and data privacy should be established prior to utilising AI in the clinical setting.

Purpose: The AngioJet Hemolytic Thrombectomy Device “AngioJet” (Boston Scientific, Malborough, MA, USA) is a rheolytic thrombectomy device that can be used in the treatment of hemodialysis arteriovenous fistula and graft thrombosis. We aim to appraise the current evidence on AngioJet thrombectomy in hemodialysis vascular access thrombosis, and to determine its safety, including complications and efficacy. Materials and methods: An electronic literature search was performed on the MEDLINE and Embase databases. References of retrieved articles were also reviewed. Articles using AngioJet for the treatment of arteriovenous fistula or graft were included. Patients less than 18 years old, non-English articles, single case reports, small case series (<5 cases), conference publications, and nonhuman models were excluded. Results: A total of 688 articles were initially evaluated and 10 articles fulfilled the inclusion criteria, which included 836 patients with 431 grafts and 234 fistulas. Results were heterogeneously reported. The adjusted mean primary patency rates for 1, 3, 6, and 12 months were 64.6%, 43.8%, 42.5%, and 30.5%, respectively; adjusted mean secondary patency rates for 3, 6, and 12 months were 76.5%, 75.1%, and 74.5%, and assisted primary patency rates were 61.9%, 47.3%, and 35.3%, respectively. There were 126 complications, most were minor. There were seven deaths, but were unable to be attributable to the AngioJet. Conclusion: The AngioJet is an effective and safe treatment option for thrombosed fistulas and grafts. However, the quality of the current evidence is poor. Further research with prospective randomized controlled trials using standardized data reporting methods is needed to prove its safety and efficacy.