Bone age determination using only the index finger: a novel approach using a convolutional neural network compared with human radiologists.

BACKGROUND: Recently developed convolutional neural network (CNN) models determine bone age more accurately than radiologists. OBJECTIVE: The purpose of this study was to determine whether a CNN and radiologists can accurately predict bone age from radiographs using only the index finger rather than the whole hand. MATERIALS AND METHODS: We used a public anonymized dataset provided by the Radiological Society of North America (RSNA) pediatric bone age challenge. The dataset contains 12,611 hand radiographs for training and 200 radiographs for testing. The index finger was cropped from these images to create a second dataset. Separate CNN models were trained using the whole-hand radiographs and the cropped second-digit dataset using the consensus ground truth provided by the RSNA bone age challenge. Bone age determination using both models was compared with ground truth as provided by the RSNA dataset. Separately, three pediatric radiologists determined bone age from the whole-hand and index-finger radiographs, and the consensus was compared to the ground truth and CNN-model-determined bone ages. RESULTS: The mean absolute difference between the ground truth and CNN bone age for whole-hand and index-finger was similar (4.7 months vs. 5.1 months, P=0.14), and both values were significantly smaller than that for radiologist bone age determination from the single-finger radiographs (8.0 months, P<0.0001). CONCLUSION: CNN-model-determined bone ages from index-finger radiographs are similar to whole-hand bone age interpreted by radiologists in the dataset, as well as a model trained on the whole-hand radiograph. In addition, the index-finger model performed better than the ground truth compared to subspecialty trained pediatric radiologists also using only the index finger to determine bone age. The radiologist interpreting bone age can use the second digit as a reliable starting point in their search pattern.

Ruptured pancreaticoduodenal artery aneurysm presenting with duodenal obstruction.

Visceral aneurysms of the pancreaticoduodenal arcades are rare. Although these aneurysms are often asymptomatic and identified incidentally on cross-sectional imaging, aneurysm rupture presents significant morbidity. Ruptured pancreaticoduodenal arcade aneurysms typically present with abdominal pain, hemorrhagic shock, or gastrointestinal bleeding. A 72-year-old male presented with nausea and vomiting and was found to have imaging evidence of duodenal obstruction. This was due to a duodenal intramural hematoma caused by a ruptured submucosal aneurysm supplied by a branch of the inferior pancreaticoduodenal artery in the presence of median arcuate ligament compression of the celiac artery. This was subsequently treated with endovascular embolization with clinical improvement in duodenal obstruction. This case illustrates an unusual presentation of a ruptured pancreaticoduodenal arcade aneurysm.

Binomial Classification of Pediatric Elbow Fractures Using a Deep Learning Multiview Approach Emulating Radiologist Decision Making.

PURPOSE: To determine the feasibility of using deep learning with a multiview approach, similar to how a human radiologist reviews multiple images, for binomial classification of acute pediatric elbow radiographic abnormalities. MATERIALS AND METHODS: A total of 21 456 radiographic studies containing 58 817 images of the elbow and associated radiology reports over the course of a 4-year period from January 2014 through December 2017 at a dedicated children's hospital were retrospectively retrieved. Mean age was 7.2 years, and 43% were female patients. The studies were binomially classified, based on the reports, as either positive or negative for acute or subacute traumatic abnormality. The studies were randomly divided into a training set containing 20 350 studies and a validation set containing the remaining 1106 studies. A multiview approach was used for the model by combining both a convolutional neural network and recurrent neural network to interpret an entire series of three radiographs together. Sensitivity, specificity, positive predictive value, negative predictive value, area under the receiver operating characteristic curve (AUC), and their 95% confidence intervals were calculated. RESULTS: AUC was 0.95, and accuracy was 88% for the model on the studied dataset. Sensitivity for the model was 91% (536 of 590), while the specificity for the model was 84% (434 of 516). Of 241 supracondylar fractures, one was missed. Of 88 lateral condylar fractures, one was missed. Of 77 elbow effusions without fracture, 15 were missed. Of 184 other abnormalities, 37 were missed. CONCLUSION: Deep learning can effectively classify acute and nonacute pediatric elbow abnormalities on radiographs in the setting of trauma. A recurrent neural network was used to classify an entire radiographic series, arrive at a decision based on all views, and identify fractures in pediatric patients with variable skeletal immaturity.Supplemental material is available for this article.© RSNA, 2019.

Thorn in My Spine: A case of a retained intradural extramedullary foreign body.

Foreign bodies in the spine are most commonly traumatic and managed in an acute setting. A few case reports describe foreign bodies resulting in delayed neurologic dysfunction, most commonly iatrogenic or from penetrating injury. We present a 30-year old man with lower extremity weakness from an intradural extramedullary foreign body granuloma secondary to a thorn, which was initially thought to represent an aggressive malignant process with intra and extramedullary involvement on MR. Postoperatively, the patient endorsed a causative trauma several years prior. We also present a review of the few similar published cases as well as the described prototypical imaging features and pathologic process.

Experimental platform for the study of region specific excitation and inhibition in neural tissue.

Willfully controlling the focus of an extracellular stimulus remains a significant challenge in the development of neural prosthetics and therapeutic devices. In part, this is due to the fact that experimental validation of the evoked response to stimuli is an arduous and time-consuming task. The development of a high-throughput data acquisition and analysis tool would greatly facilitate the design of spatially selective stimulation protocols. We present an automated imaging system that can optically track and identify the action potentials of individual neurons evoked by coordinated stimulus waveforms applied at multiple electrodes. This system can simultaneously provide arbitrary current waveforms to four electrodes, and it is capable of automatically monitoring the cellular responses of every neuron in a cultured network within a 1.6 x 1.6 mm area. The purpose of this platform is to develop stimulus protocols that exploit the benefits of multi-polar field shaping and temporal ion-channel manipulation to localize cellular excitation beyond the vicinity of the electrode. Preliminary single electrode experiments demonstrate that spatially selective stimulus suppression may be achieved with cathodic, depolarizing prepulses that induce a sub-threshold refractory state in neighboring neurons. Coordinated, multi-site stimuli could potentially take advantage of this refractory state to direct the stimulus focus away from the surrounding area of the electrode and into the inter-electrode spaces.