Developing an artificial intelligence diagnostic tool for paediatric distal radius fractures, a proof of concept study.

INTRODUCTION: In the UK, 1 in 50 children sustain a fractured bone yearly, yet studies have shown that 34% of children sustaining an injury do not have a visible fracture on initial radiographs. Wrist fractures are particularly difficult to identify because the growth plate poses diagnostic challenges when interpreting radiographs. METHODS: We developed Convolutional Neural Network (CNN) image recognition software to detect fractures in radiographs from children. A consecutive data set of 5,000 radiographs of the distal radius in children aged less than 19 years from 2014 to 2019 was used to train the CNN. In addition, transfer learning from a VGG16 CNN pretrained on non-radiological images was applied to improve generalisation of the network and the classification of radiographs. Hyperparameter tuning techniques were used to compare the model with the radiology reports that accompanied the original images to determine diagnostic test accuracy. RESULTS: The training set consisted of 2,881 radiographs with a fracture and 1,571 without; 548 radiographs were outliers. With additional augmentation, the final data set consisted of 15,498 images. The data set was randomly split into three subsets: training (70%), validation (10%) and test (20%). After training for 20 epochs, the diagnostic test accuracy was 85%. DISCUSSION: A CNN model is feasible in diagnosing paediatric wrist fractures. We demonstrated that this application could be utilised as a tool for improving diagnostic accuracy. Future work would involve developing automated treatment pathways for diagnosis, reducing unnecessary hospital visits and allowing staff redeployment to other areas.

Outcome of one-stage correction of deformities of the ankle and hindfoot and fusion in Charcot neuroarthropathy using a retrograde intramedullary hindfoot arthrodesis nail.

We report the outcomes of 20 patients (12 men, 8 women, 21 feet) with Charcot neuro-arthropathy who underwent correction of deformities of the ankle and hindfoot using retrograde intramedullary nail arthrodesis. The mean age of the patients was 62.6 years (46 to 83); their mean BMI was 32.7 (15 to 47) and their median American Society of Anaesthetists score was 3 (2 to 4). All presented with severe deformities and 15 had chronic ulceration. All were treated with reconstructive surgery and seven underwent simultaneous midfoot fusion using a bolt, locking plate or a combination of both. At a mean follow-up of 26 months (8 to 54), limb salvage was achieved in all patients and 12 patients (80%) with ulceration achieved healing and all but one patient regained independent mobilisation. There was failure of fixation with a broken nail requiring revision surgery in one patient. Migration of distal locking screws occurred only when standard screws had been used but not with hydroxyapatite-coated screws. The mean American Academy of Orthopaedic Surgeons Foot and Ankle (AAOS-FAO) score improved from 50.7 (17 to 88) to 65.2 (22 to 88), (p = 0.015). The mean Short Form (SF)-36 Health Survey Physical Component Score improved from 25.2 (16.4 to 42.8) to 29.8 (17.7 to 44.2), (p = 0.003) and the mean Euroqol EQ­5D­5L score improved from 0.63 (0.51 to 0.78) to 0.67 (0.57 to 0.84), (p = 0.012). Single-stage correction of deformity using an intramedullary hindfoot arthrodesis nail is a good form of treatment for patients with severe Charcot hindfoot deformity, ulceration and instability provided a multidisciplinary care plan is delivered.