GLORIA

GEOMAR Library Ocean Research Information Access

X

Your email was sent successfully. Check your inbox.

An error occurred while sending the email. Please try again.

Proceed reservation?

Export
  • 1
    Online Resource
    Online Resource
    Development and external validation of prediction models to predict implantable cardioverter-defibrillator efficacy in primary prevention of sudden cardiac death
    Oxford University Press (OUP) ; 2021
    In:  EP Europace Vol. 23, No. 6 ( 2021-06-07), p. 887-897
    Details
    In: EP Europace, Oxford University Press (OUP), Vol. 23, No. 6 ( 2021-06-07), p. 887-897
    Abstract: This study was performed to develop and externally validate prediction models for appropriate implantable cardioverter-defibrillator (ICD) shock and mortality to identify subgroups with insufficient benefit from ICD implantation. Methods and results We recruited patients scheduled for primary prevention ICD implantation and reduced left ventricular function. Bootstrapping-based Cox proportional hazards and Fine and Gray competing risk models with likely candidate predictors were developed for all-cause mortality and appropriate ICD shock, respectively. Between 2014 and 2018, we included 1441 consecutive patients in the development and 1450 patients in the validation cohort. During a median follow-up of 2.4 (IQR 2.1–2.8) years, 109 (7.6%) patients received appropriate ICD shock and 193 (13.4%) died in the development cohort. During a median follow-up of 2.7 (IQR 2.0–3.4) years, 105 (7.2%) received appropriate ICD shock and 223 (15.4%) died in the validation cohort. Selected predictors of appropriate ICD shock were gender, NSVT, ACE/ARB use, atrial fibrillation history, Aldosterone-antagonist use, Digoxin use, eGFR, (N)OAC use, and peripheral vascular disease. Selected predictors of all-cause mortality were age, diuretic use, sodium, NT-pro-BNP, and ACE/ARB use. C-statistic was 0.61 and 0.60 at respectively internal and external validation for appropriate ICD shock and 0.74 at both internal and external validation for mortality. Conclusion Although this cohort study was specifically designed to develop prediction models, risk stratification still remains challenging and no large group with insufficient benefit of ICD implantation was found. However, the prediction models have some clinical utility as we present several scenarios where ICD implantation might be postponed.
    Type of Medium: Online Resource
    ISSN: 1099-5129 , 1532-2092
    URL: Article
    DOI: 10.1093/europace/euab012
    Language: English
    Publisher: Oxford University Press (OUP)
    Publication Date: 2021
    detail.hit.zdb_id: 2002579-8
    Location Call Number Limitation Availability
    BibTip Others were also interested in ...
    Online Resource
  • 2
    Online Resource
    Online Resource
    Development and external validation of automated detection, classification, and localization of ankle fractures: inside the black box of a convolutional neural network (CNN)
    Springer Science and Business Media LLC ; 2023
    In:  European Journal of Trauma and Emergency Surgery Vol. 49, No. 2 ( 2023-04), p. 1057-1069
    Details
    In: European Journal of Trauma and Emergency Surgery, Springer Science and Business Media LLC, Vol. 49, No. 2 ( 2023-04), p. 1057-1069
    Abstract: Convolutional neural networks (CNNs) are increasingly being developed for automated fracture detection in orthopaedic trauma surgery. Studies to date, however, are limited to providing classification based on the entire image—and only produce heatmaps for approximate fracture localization instead of delineating exact fracture morphology. Therefore, we aimed to answer (1) what is the performance of a CNN that detects, classifies, localizes, and segments an ankle fracture, and (2) would this be externally valid? Methods The training set included 326 isolated fibula fractures and 423 non-fracture radiographs. The Detectron2 implementation of the Mask R-CNN was trained with labelled and annotated radiographs. The internal validation (or ‘test set’) and external validation sets consisted of 300 and 334 radiographs, respectively. Consensus agreement between three experienced fellowship-trained trauma surgeons was defined as the ground truth label. Diagnostic accuracy and area under the receiver operator characteristic curve (AUC) were used to assess classification performance. The Intersection over Union (IoU) was used to quantify accuracy of the segmentation predictions by the CNN, where a value of 0.5 is generally considered an adequate segmentation. Results The final CNN was able to classify fibula fractures according to four classes (Danis-Weber A, B, C and No Fracture) with AUC values ranging from 0.93 to 0.99. Diagnostic accuracy was 89% on the test set with average sensitivity of 89% and specificity of 96%. External validity was 89–90% accurate on a set of radiographs from a different hospital. Accuracies/AUCs observed were 100/0.99 for the ‘No Fracture’ class, 92/0.99 for ‘Weber B’, 88/0.93 for ‘Weber C’, and 76/0.97 for ‘Weber A’. For the fracture bounding box prediction by the CNN, a mean IoU of 0.65 (SD ± 0.16) was observed. The fracture segmentation predictions by the CNN resulted in a mean IoU of 0.47 (SD ± 0.17). Conclusions This study presents a look into the ‘black box’ of CNNs and represents the first automated delineation (segmentation) of fracture lines on (ankle) radiographs. The AUC values presented in this paper indicate good discriminatory capability of the CNN and substantiate further study of CNNs in detecting and classifying ankle fractures. Level of evidence II, Diagnostic imaging study.
    Type of Medium: Online Resource
    ISSN: 1863-9933 , 1863-9941
    URL: Article
    DOI: 10.1007/s00068-022-02136-1
    Language: English
    Publisher: Springer Science and Business Media LLC
    Publication Date: 2023
    detail.hit.zdb_id: 2276432-X
    Location Call Number Limitation Availability
    BibTip Others were also interested in ...
    Online Resource
Close ⊗
This website uses cookies and the analysis tool Matomo. More information can be found here...