oblique, transverse) using three regions of interest (ROI) sizes. Three repeated measures were acquired per location, orientation and ROI size combination. The Supersonic Aixplorer system was used to compare LE9 performance. Within-subject coefficient of variance (WSCV), intraclass correlation coefficients (ICC) and limits of agreement (LOA) were calculated. Results: The small ROI produced the highest WSCV of 6.8%, while medium and large were both 4.3%. Using medium ROI, mean SWE velocity was highest in transverse (2.43m/s) and lowest in longitudinal (1.96m/s) orientation. Reliability increased gradually from transverse to oblique to longitudinal. LE9 intra-system reliability ranged from fair (.41) to good (.77). In contrast, Aixplorer reliability ranged from .81 to .97. The most reliable combination was longitudinal/lateral [ICC (95% CI) LE9¼.77 (.58, .90); Aixplorer¼.97 (.92, 1.00)]. The best inter-system agreement was again found for the longitudinal/lateral combination, with an ICC of .71 (.48, 1.00) and a mean bias of 0.06m/s (LOA Æ0.36m/s). Conclusion: SWE using LE9 produces its most reliable results in a longi- tudinal plane from the lateral location using the medium ROI. Elastography estimations agree between the two SWE systems but with different de- grees of internal measurement reliability. Does unit of reporting, depth and probe load influence the reliability of muscle shear wave elastography? Authors: Abdulrahman M. Alfuraih, Phil O’Connor, Ai Lyn Tan, Elizabeth Hensor, Paul Emery, Richard Wakefield Purpose: To investigate the effect of using different units, probe loads and acquisition depths on the reliability of shear wave elastography (SWE) in healthy skeletal muscles. Methods and materials: Using the new General Electric LOGIQ-E9 SWE system, the vastus lateralis, biceps femoris, biceps brachii and abductor digiti minimi muscles were investigated in twenty healthy subjects. Readings were acquired in shear wave velocity (m/s) and Young’s modulus (kPa) from various depths. The effect of using a layer of standoff gel was compared against minimal probe load. Intraclass correlations (ICC) of three repeated measurements were used to report reliability. Results: Velocity differed between the muscles and was lowest for biceps femoris (1.54 m/s) and highest for abductor digiti minimi (2.55 m/s) with a difference (95% CI) of 0.79 m/s (0.64, 0.95)]. In the vastus lateralis, using unit of m/s resulted in higher ICC of .83 (.65, 93) in comparison to .77 (.52, .90) for kPa which overestimates elasticity. The standoff gel method decreased reliability to .62 (.20, .84) despite the mean velocity being similar to minimal probe load. The variance of the repeated readings increased significantly with depth from 0.07 at 4 cm to 0.17 at 6 cm (p<0.001). Conclusion: Acquisition factors have an influence on the reliability of muscle SWE. Elasticity units are not synonymous, and using kPa may produce less reliable results. We recommend acquiring readings using minimal probe load from depths less than 4 cm. Radiology volunteering in Tanzania Author: Paul Michael Kelly Purpose: To illustrate the worth of clinical radiologists volunteering for charity work in Africa. I spent ten days aboard a medical ship on Lake Victoria, Tanzania in 2017. My primary purpose was to undertake and guide local staff in the use of ultrasound using the Sonosite 180 machine that they had on board. Methods and materials: The Jubilee Hope is a medical boat operated by the Vine Trust charity. It was the subject of a recent BBC Scotland docu- mentary. This former Royal Navy tender vessel sailed from the Clyde, round the Horn of Africa to Mombasa and was transported over land to Mwanza on the Tanzanian coast of Lake Victoria. Since arriving in 2014, the boat has undertaken several visits to islands in the south west of the lake where there is no formal local health care. I volunteered to spend ten days on the boat performing ultrasound and training local healthcare workers in the use of the Sonosite 180 machine. Results: In cramped conditions, in the bowels of the boat, I undertook 60 ultrasound examinations e approximately seven per day for nine working days. These comprised 34 abdomen, ten pelvis, ten obstetric, three neck, two small parts and one breast. Pregnancies were dated and two unex- pected pregnancies and twins diagnosed. Other diagnoses made were: one ascites with probable cirrhosis; two goiters; two complex ovarian cysts and one nasal dermoid. Unfortunately, teaching opportunities were limited due to the work commitments of others. I also undertook 150 of the 350 HIV tests performed. 12% were positive and patients were referred on to mainland for treatment. Cost to volunteer was £1,100 plus cost of flights. Conclusion: I found it extremely rewarding to take my ultrasound skills to a remote corner of Africa. I have suggested a local staff member go on a practical ultrasound course and I hope to return thereafter to offer further training. Success and complication rates of ultrasound-guided core biopsies of abdominal and thoracic lesions Authors: John Reicher, Madhangi Parameswaran, Alexandru Calciu, Naail Al-Zuhir, Tim Hoare, Andrew McNeill Purpose: To determine success and complication rates for ultrasound (US)-guided core biopsies of focal thoracic and abdominal lesions at Newcastle Upon Tyne Hospitals from 2014e2016. Methods and materials: We reviewed pathology reports for relevant bi- opsies between 01/01/2014 and 31/12/2016. Biopsies were deemed suc- cessful if the sample was diagnostic. Unsuccessful biopsies were divided into N1 (inadequate sample) and N2 (geographical miss). We reviewed all cases for radiologically confirmed complications. Results: Liver - 278 biopsies; successful e 246 (88.1%), N1 e 7(2.5%), N2 e 26(9.4%). - Complications e 4(1.4%), including one bile leak requiring drainage. Peritoneum - Seven biopsies; successful e 45(95.8%), N1 e 1(2.1%), N2 e 1(2.1%). Thorax (including lung, pleura, mediastinum) - 20 biopsies; successful e 19(95%), N1 e 1(5%). Kidney - 123 biopsies; successful e 110(89.4%), N1 e 5(4.1%), N2 e 8(6.5%). - Complications e 3(2.4%), including one peri-nephric haematoma requiring thrombin injection. Other (for example, adrenals, adnexae, lymph nodes) - 99 biopsies; successful e 92(92.9%), N1 e 2(2.0%), N2 e 5(5.1%). Overall - 567 biopsies; successful e 511(90.1%), N1 e 16(2.8%), N2 e 40(7.1%). - Complications e 7(1.2%), only two requiring invasive intervention. Conclusion: US-guided lesional biopsies in our institution have a low complication rate, but the number of N2 (missed) biopsies of hepatic and renal lesions is relatively high. We are currently reviewing procedural factors (for example, needle gauge, number of passes, lesion size/depth) and their correlation with success rates, to determine how these can be improved. A closed loop audit on FAST scan documentation in the emergency department Authors: Ravindran Karthigan, Kashif Burney, Simon Lambracos Purpose: Focused assessment with sonography for trauma (FAST) scans are a useful bedside tool used in trauma to detect the presence of free fluid Abstract / Clinical Radiology 72 (2017) S1eS13 S12