Machine Learning Versus Standard Techniques for Updating Searches for Systematic Reviews: A Diagnostic Accuracy Study

2017 ◽  
Vol 167 (3) ◽  
pp. 213 ◽  
Author(s):  
Paul G. Shekelle ◽  
Kanaka Shetty ◽  
Sydne Newberry ◽  
Margaret Maglione ◽  
Aneesa Motala
Keyword(s):  
Machine Learning ◽  
Diagnostic Accuracy ◽  
Systematic Reviews ◽  
Diagnostic Accuracy Study ◽  
Accuracy Study ◽  
Standard Techniques

Machine learning volumetry of ischemic brain lesions on CT after thrombectomy—prospective diagnostic accuracy study in ischemic stroke patients

2020 ◽  
Vol 62 (10) ◽  
pp. 1239-1245
Author(s):  
Jiri Kral ◽  
Martin Cabal ◽  
Linda Kasickova ◽  
Jaroslav Havelka ◽  
Tomas Jonszta ◽  
...  
Keyword(s):  
Machine Learning ◽  
Ischemic Stroke ◽  
Diagnostic Accuracy ◽  
Brain Lesions ◽  
Diagnostic Accuracy Study ◽  
Stroke Patients ◽  
Ischemic Brain ◽  
Accuracy Study

Ultrasonographic 3D Evaluation in the Diagnosis of Bladder Endometriosis: A Prospective Comparative Diagnostic Accuracy Study

2021 ◽  
pp. 1-8
Author(s):  
Fabio Barra ◽  
Franco Alessandri ◽  
Carolina Scala ◽  
Simone Ferrero
Keyword(s):  
Diagnostic Accuracy ◽  
Three Dimensional ◽  
3D Ultrasound ◽  
Diagnostic Accuracy Study ◽  
Extensive Experience ◽  
Deep Endometriosis ◽  
3D Reconstructions ◽  
Analysis Group ◽  
Bladder Endometriosis ◽  
Accuracy Study

<b><i>Objective:</i></b> The use of three-dimensional (3D) transvaginal ultrasonography (TVS) has been investigated for the diagnosis of deep endometriosis (DE). This study aimed to evaluate if 3D reconstructions improve the performance of TVS) in assessing the presence and characteristics of bladder endometriosis (BE). <b><i>Design:</i></b> This was a single-center comparative diagnostic accuracy study. <b><i>Participants/Materials, Setting, Methods:</i></b> Patients referred to our institution (Piazza della Vittoria 14 Srl, Genova, Italy) with clinical suspicion of DE were included. In case of surgery, women underwent systematic preoperative ultrasonographic imaging; an experienced sonographer performed a conventional TVS; another experienced sonographer, blinded to results of the previous exam, performed TVS, with the addition of 3D modality. The presence and characteristics of BE nodules were described in accord with International DE Analysis group consensus. Ultrasound data were compared with surgical and histological results. <b><i>Results:</i></b> Overall, BE was intraoperatively found in 34 out of 194 women who underwent surgery for DE (17.5%; 95% confidence interval: 12.8–23.5%). TVS without and with 3D reconstructions were able to detect endometriotic BE in 82.2% (<i>n</i> = 28/34) and 85.3% (<i>n</i> = 29/34) of the cases (<i>p</i> = 0.125). Both the exams similarly estimated the largest diameter of BE (<i>p</i> = 0.652) and the distance between the endometriotic nodule and the closest ureteral meatus (<i>p</i> = 0.341). However, TVS with 3D reconstructions was more precise in estimating the volume of BE (<i>p</i> = 0.031). In one case (2.9%), TVS without and with 3D reconstructions detected the infiltration of the intramural ureter, which was confirmed at surgery and required laparoscopic ureterovesical reimplantation. <b><i>Limitations:</i></b> The extensive experience of the gynecologists performing the ultrasonographic scans, the lack of prestudy power analysis, and the population selected, which may have been influenced by the position of the institution as a referral center specialized in the treatment of severe endometriosis, are limitations of the current study. <b><i>Conclusion:</i></b> Our results demonstrated the high accuracy of ultrasound for diagnosing BE. The addition of 3D reconstructions does not improve the performance of TVS in diagnosing the presence and characteristics of BE. However, the volume of BE may be more precisely assessed by 3D ultrasound.

Sign in / Sign up

Export Citation Format

Share Document