Role of CT angiography and therapeutic anticoagulation in patients presenting to the emergency department with acute gastrointestinal bleeding.

INTRODUCTION: Acute gastrointestinal bleeding (GIB) is associated with morbidity and mortality. There can be a low threshold for practitioners to assess for active GIB and computed tomography angiography (CTA) examinations are performed frequently, even for stable patients and those who are therapeutically anticoagulated. We aimed to assess the predictive value of CTA for acute GIB and the influence of CTA on treatment. METHODS: Retrospective single-centre study over a 2-year period. RESULTS: A total of 227 patients with mean age 67.7 years (SD 17.86), 58.6% male. 84.4% were for lower GIB. 49 patients were on therapeutic anticoagulation (21.6%). 45 CTAs were positive (19.8%). 22 patients received embolisation, and 15 received acute endoscopic treatment. CTA sensitivity was 68.6% and specificity 89.1%. The PPV was 53.3% and NPV 93.9%. The odds ratio of a positive CTA requiring treatment for patients on therapeutic anticoagulation was 1.1 (P = 0.932) compared with the odds of patients not taking therapeutic anticoagulation 21.5 (P < 0.001). The risk ratio for requiring treatment if not taking anticoagulation was 6.2. A total of 19 patients (9.1%) met the definition of CI-AKI as a result of the CTA. A pre-existing eGFR of less than 20 was associated with significantly increased odds of developing CI-AKI (OR 3.95, P = 0.031, 95%CI 1.135-13.782). CONCLUSIONS: The presence of anticoagulation has a significant impact on the decision not to perform interventional treatments on patients with acute GIB when CTA is positive. Anticoagulant reversal and volume resuscitation are important front-line measures, and CTA may have a role for those anticoagulated who are haemodynamically unstable after resuscitation.

Identification of the lateral femoral safe zone for drilling during LISS plate fixation of distal femur fractures.

BACKGROUND: The incidence of arterial injury associated with femoral fractures is approximately 1%. Lateral sub-muscular plate fixation is gaining popularity for the management of distal femoral fractures. The objective of this study was to assess the iatrogenic risk to the superficial femoral artery (SFA) during Less Invasive Stabilisation System (LISS) plate fixation of distal femoral fractures by analysing the range of distances and angles between LISS plate screws and the drilling line to the SFA. MATERIALS AND METHODS: We identified all patients who underwent LISS plate fixation of distal femoral fractures between 2008 and 2018 in our level-1 trauma centre. Patients who underwent postoperative computed tomography for any reason were eligible for inclusion in the study. Twenty-five patients met the inclusion criteria. The sample comprised 10 male and 15 female patients with a mean age of 55 years. The most common fracture type was a supracondylar femur fracture (56%), followed by an intercondylar fracture (36%). A 13-hole LISS plate was the most common plate length used (44%). A consultant radiologist reviewed all scans to verify the visibility and marking of the SFA. RESULTS: The median distance between the screw tip and the SFA was 21 mm (range, 8-65 mm). There was a negative correlation between the LISS plate hole number and the trajectory of drilling (Pearson coefficient: -0.87, p < 0.001). Using a linear regression model, the SFA was more likely to be in the line of drilling when the 6th to 10th holes in the LISS plate were used. CONCLUSION: Extra care is needed when drilling into the LISS plate holes from the lateral to the medial direction in order to reduce the risk of iatrogenic injury to the SFA, especially in the high-risk plate zone where the artery can be close to the drilling line.

Computer vs human: Deep learning versus perceptual training for the detection of neck of femur fractures.

INTRODUCTION: To evaluate the accuracy of deep convolutional neural networks (DCNNs) for detecting neck of femur (NoF) fractures on radiographs, in comparison with perceptual training in medically-naïve individuals. METHODS: This study extends a previous study that conducted perceptual training in medically-naïve individuals for the detection of NoF fractures on a variety of dataset sizes. The same anteroposterior hip radiograph dataset was used to train two DCNNs (AlexNet and GoogLeNet) to detect NoF fractures. For direct comparison with perceptual training results, deep learning was completed across a variety of dataset sizes (200, 320 and 640 images) with images split into training (80%) and validation (20%). An additional 160 images were used as the final test set. Multiple pre-processing and augmentation techniques were utilised. RESULTS: AlexNet and GoogLeNet DCNNs NoF fracture detection accuracy increased with larger training dataset sizes and mildly with augmentation. Accuracy increased from 81.9% and 88.1% to 89.4% and 94.4% for AlexNet and GoogLeNet respectively. Similarly, the test accuracy for the perceptual training in top-performing medically-naïve individuals increased from 87.6% to 90.5% when trained on 640 images compared with 200 images. CONCLUSIONS: Single detection tasks in radiology are commonly used in DCNN research with their results often used to make broader claims about machine learning being able to perform as well as subspecialty radiologists. This study suggests that as impressive as recognising fractures is for a DCNN, similar learning can be achieved by top-performing medically-naïve humans with less than 1 hour of perceptual training.

Perceptual training to improve hip fracture identification in conventional radiographs.

Diagnosing certain fractures in conventional radiographs can be a difficult task, usually taking years to master. Typically, students are trained ad-hoc, in a primarily-rule based fashion. Our study investigated whether students can more rapidly learn to diagnose proximal neck of femur fractures via perceptual training, without having to learn an explicit set of rules. One hundred and thirty-nine students with no prior medical or radiology training were shown a sequence of plain film X-ray images of the right hip and for each image were asked to indicate whether a fracture was present. Students were told if they were correct and the location of any fracture, if present. No other feedback was given. The more able students achieved the same level of accuracy as board certified radiologists at identifying hip fractures in less than an hour of training. Surprisingly, perceptual learning was reduced when the training set was constructed to over-represent the types of images participants found more difficult to categorise. Conversely, repeating training images did not reduce post-training performance relative to showing an equivalent number of unique images. Perceptual training is an effective way of helping novices learn to identify hip fractures in X-ray images and should supplement the current education programme for students.

Internet use by orthopaedic outpatients - current trends and practices.

BACKGROUND: The e-patient revolution increasingly enables patients to self diagnose and self educate, influencing decisions affecting their health. This poses a challenge for both patients and health care professionals due to the highly variable and often poor quality information available on the internet. AIMS: This study aims to measure the current internet usage in patients attending outpatient clinics, in both a public and private setting. All patients were recruited whilst consulting orthopaedic surgeons. METHOD: We developed a 29 question survey which asked questions related to patient demographics, general internet usage and internet usage related to the patient's orthopaedic condition. Patients were recruited for the public cohort during Western Health outpatient clinics and for the private cohort during private surgical consults in the waiting rooms of eight surgeons' clinics. RESULTS: A total of 400 surveys were completed; 200 in both the private and public cohorts of the study. Of all surveyed participants, 79% (n = 316) had access to the internet. Of people who had access to the internet 65.2% (n = 206) used the internet to investigate their orthopaedic condition. 29.6% (n = 61) of participants asked their surgeon questions related to information they had read on the internet. Of patients that had access to the internet 36.1% (n = 114) used the internet to research their surgeon. CONCLUSION: Patients are commonly using the internet as an information resource, in spite of the highly variable quality of this information. This highlights the need for patient information websites which reflect the current standards of clinical practice.