Ossification area localization in pediatric hand radiographs using deep neural networks for object detection

Abstract

BackgroundDetection of ossification areas of hand bones in X-ray images is an important task, e.g. as a preprocessing step in automated bone age estimation. Deep neural networks have emerged recently as de facto standard detection methods, but their drawback is the need of large annotated datasets. Finetuning pre-trained networks is a viable alternative, but it is not clear a priori if training with small annotated datasets will be successful, as it depends on the problem at hand. In this paper, we show that pre-trained networks can be utilized to produce an effective detector of ossification areas in pediatric X-ray images of hands.Methods and findingsA publicly available Faster R-CNN network, pre-trained on the COCO dataset, was utilized and finetuned with 240 manually annotated radiographs from the RSNA Pediatric Bone Age Challenge, which comprises over 14.000 pediatric radiographs. The validation is done on another 89 radiographs from the dataset and the performance is measured by Intersection-over-Union (IoU). To understand the effect of the data size on the pre-trained network, subsampling was applied to the training data and the training was repeated. Additionally, the network was trained from scratch without any pre-trained weights. Finally, to understand whether the trained model could be useful, we compared the inference of the network to an annotation of an expert radiologist. The finetuned network was able to achieve an average precision (mAP@0.5IoU) of 92.92 ± 1.93. Apart from the wrist region, all ossification areas were able to benefit from more data. In contrast, the network trained from scratch was not able to produce any correct results. When compared to the annotations of the expert radiologist, the network was able to localize the regions quite well, as the F1-Score was on average 91.85 ± 1.06.ConclusionsBy finetuning a pre-trained deep neural network, with 240 annotated radiographs, we were able to successfully detect ossification areas in prediatric hand radiographs.

Fine-grained precise-bone age assessment by integrating prior knowledge and recursive feature pyramid network

“…The system was based on Faster R-CNN, which detected 17 regions from the 6 ossification areas (DIP, PIP, MCP, radius, ulna, and wrist). The performance was measured by the average precision of the Intersection over Union (0.5IoU) for the 6 ossification areas [33] and a detected region is considered as a good match if the overlapped area accounts for at least 50% region. It is a lax measure of subregion detection for BAA.…”

Detection of Pediatric Femur Configuration on X-ray Images

“…Recently, individual studies have made attempts to apply CNN to solve regression tasks for children's medical images [18][19][20]. Nevertheless, there have been issues considering the lack of input data, as pediatric medical image datasets are rarely publicly available.…”

Accurate Instance Segmentation in Pediatric Elbow Radiographs

“…The improved network aims to detect tiny fractures in elbow radiographs. Object detection is also successfully applied on the diagnosis of fractures in wrist radiographs [ 27 ], detecting intervertebral discs in lateral lumbar radiographs [ 28 ], localizing ossification areas of hand bones [ 29 ], and detecting distal radius fractures in anteroposterior arm radiographs [ 30 ]. However, object detection can only obtain a rough location of a bone or fracture, which is not enough for further diagnosis.…”