Automatic detection and localization of thighbone fractures in X-ray based on improved deep learning method

Guan, Bin; Yao, Jinkun; Wang, Shaoquan; Zhang, Guoshan; Zhang, Yueming; Wang, Xinbo; Wang, Mengxuan

doi:10.1016/j.cviu.2021.103345

Cited by 29 publications

(25 citation statements)

References 17 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…We adopted the Pytorch_EHR framework established by Zhi Group where Deep learning models based multiple recurrent neural networks 21 were used to analyze and predict clinical outcomes 22 . In conjunction with their algorithm and our previous DeepBiomarker model, we further modi ed the framework as DeepBiomarker2 by (a) integrating individual lab tests, SDoH parameters and medications along with the diagnosis groups as the input, so that we can assess the important clinical and nonclinical factors associated with ASUD risk; and (b) re ning contribution analysis 23 module by re ning the relative contribution analysis for the identi cation of key factors (see below for more details). Contribution analysis is a method used to calculate the relative contribution of key factors which are identi ed by determining the observed change in the model.…”

Section: Deepbiomarker2mentioning

confidence: 99%

DeepBiomarker2: Prediction of alcohol and substance use disorder risk in post-traumatic stress disorder patients using electronic medical records and multiple social determinants of health

Miranda

Fan

et al. 2023

Preprint

View full text Add to dashboard Cite

Introduction: Prediction of high-risk events amongst patients with mental disorders is critical for personalized interventions. In our previous study, we developed a deep learning-based model, DeepBiomarker by utilizing electronic medical records (EMR) to predict the outcomes of patients with suicide-related events in post-traumatic stress disorder (PTSD) patients. Methods We improved our deep learning model to develop DeepBiomarker2 through data integration of multimodal information: lab tests, medication use, diagnosis, and social determinants of health (SDoH) parameters (both individual and neighborhood level) from EMR data for outcome prediction. We further refined our contribution analysis for identifying key factors. We applied DeepBiomarker2 to analyze EMR data of 38,807 patients from University of Pittsburgh Medical Center diagnosed with PTSD to determine their risk of developing alcohol and substance use disorder (ASUD). Results DeepBiomarker2 predicted whether a PTSD patient will have a diagnosis of ASUD within the following 3 months with a c-statistic (receiver operating characteristic AUC) of 0·93. We used contribution analysis technology to identify key lab tests, medication use and diagnosis for ASUD prediction. These identified factors imply that the regulation of the energy metabolism, blood circulation, inflammation, and microbiome is involved in shaping the pathophysiological pathways promoting ASUD risks in PTSD patients. Our study found protective medications such as oxybutynin, magnesium oxide, clindamycin, cetirizine, montelukast and venlafaxine all have a potential to reduce risk of ASUDs. Discussion DeepBiomarker2 can predict ASUD risk with high accuracy and can further identify potential risk factors along with medications with beneficial effects. We believe that our approach will help in personalized interventions of PTSD for a variety of clinical scenarios.

show abstract

Section: Deepbiomarker2mentioning

confidence: 99%

DeepBiomarker2: Prediction of alcohol and substance use disorder risk in post-traumatic stress disorder patients using electronic medical records and multiple social determinants of health

Miranda

Fan

et al. 2023

Preprint

View full text Add to dashboard Cite

show abstract

“…achieved a mean AP (mAP) score of 68.8% using the anchor-based Faster R-CNN model with multi-resolution FPN and ResNet50 backbone network on 2333 femoral fracture X-ray images. Thian et al 18 A deep learning approach for bone fracture detection in radiographs by 24 presents a CNN for the detection of fractures in radiographs.…”

Section: Related Workmentioning

confidence: 99%

HD-NET: Humerus deep-net for humerus fracture and bony callus formation analysis

Abdullah

Shoaib

Arshad

et al. 2023

Preprint

View full text Add to dashboard Cite

When employing x-ray images, fracture identification in orthopaedics is a difficult task. A large percentage of humerus fracture patients are seen in hospitals, particularly in their emergency departments. Similar to this, after a fracture, accurate callus production monitoring is crucial for bone healing. Thus, a fractured patient’s diagnosis and therapy must be accurate and administered promptly. This work investigates the use of deep learning on X-ray images of the humerus for fracture snd bone callus formation analysis to help physicians in the diagnosis of such fractures, especially in emergency settings.This study is named HD-NET, which stands for Humerus Deep Net. The framework includes image enhancement using a Gaussian filter and histogram equalization, two-stage object detection, image super-resolution, U-NET segmentation with feature recalibration. Finally, an LSTM with a sequence length of 2 is used to analyze callus formation at the fracture site. The LSTM takes the segmented area as input and outputs a prediction for the stage of healing and potential complications. The proposed framework was evaluated on a combination of the MURA dataset and a self-collected dataset.Results demonstrated that in terms of specificity, sensitivity, and accuracy, the suggested framework performed better than earlier studies. This research can be expanded to different bone types and is useful for orthopaedic practitioners.

show abstract

“…For example, the DL techniques work on bone fracture detection are using either x-ray or MRI data. 59,61 The average accuracy of DL based x-ray and MRI data falls around 89.56% to 88.98%. At the same time, the novel experiment gets 96.77% of fracture detection accuracy.…”

Section: Mean Precisionmentioning

confidence: 99%

“…Comparing to existing DL based bone fracture detection techniques, the newly developed experiment has better accuracy rate. For example, the DL techniques work on bone fracture detection are using either x‐ray or MRI data 59,61 . The average accuracy of DL based x‐ray and MRI data falls around 89.56% to 88.98%.…”

Section: Techniques For Multi‐fracture Detection Using X‐ray Imagesmentioning

confidence: 99%

A comparative experimental analysis and deep evaluation practices on human bone fracture detection using x‐ray images

Kumar

Prabu

Pandimurugan

et al. 2022

Concurrency and Computation

View full text Add to dashboard Cite

Summary X‐ray images are widely used to identify fractures in human bones. Radiographic diagnosis models take more time and manual procedures for establishing physical analysis of bone fractures. Furthermore, the lack of clinical resources and the medical support systems lead to inaccurate bone data extractions. In this case, the need for detail is required to understand the scientific issues in x‐ray and magnetic resonance imaging (MRI) based bone fracture diagnosis solutions. Particularly, the current bone fracture detection models are emerging with computerized frameworks and health informatics consoles. In this regard, this article lists five phases. First, data preparation and data collection tasks are initiated. Second, bone fracture diagnosis models are comparatively analyzed to find optimal observations. Consequently, the third phase of the article examines the treatments against bone fracture observations with traditional and deep learning (DL) techniques. The fourth phase consists of relative diagnosis solutions and the treatment benefits on various types of technical frameworks. Finally, this work concludes the comparison with recent DL‐based bone fracture identification models.

show abstract

Automatic detection and localization of thighbone fractures in X-ray based on improved deep learning method

Cited by 29 publications

References 17 publications

DeepBiomarker2: Prediction of alcohol and substance use disorder risk in post-traumatic stress disorder patients using electronic medical records and multiple social determinants of health

DeepBiomarker2: Prediction of alcohol and substance use disorder risk in post-traumatic stress disorder patients using electronic medical records and multiple social determinants of health

HD-NET: Humerus deep-net for humerus fracture and bony callus formation analysis

A comparative experimental analysis and deep evaluation practices on human bone fracture detection using x‐ray images

Contact Info

Product

Resources

About