English (United Kingdom)

https://curated-unify.zendy.io/wp-json/zendy-region/v1/featured_content/oa?rat=en

https://curated-unify.zendy.io/wp-json/zendy-region/v1/highlighted_journal/

Global: Zendy Plus annual

Presents the access of premium content as premium feature

Premium Content

Presents the keyphrase highlighting as premium feature

Keyphrase Highlighting

Presents the summarisation as premium feature

Summarisation

Insights

Presents the pdf analysis as premium feature

PDF Analysis

Presents the zaia usage as premium feature

ZAIA

Global: Zendy Tools annual

Global: Zendy Free Plan

Global: Zendy Tools monthly

Global: Zendy Plus monthly

Purpose: To compare the segmentation and detection performance of a deeplearning model trained on a database of human-labelled clinicaldiffusion-weighted (DW) stroke lesions to a model trained on the same databaseenhanced with synthetic DW stroke lesions. Methods: In this institutionalreview board approved study, a stroke database of 962 cases (mean age 65+/-17years, 255 males, 449 scans with DW positive stroke lesions) and a normaldatabase of 2,027 patients (mean age 38+/-24 years,1088 females) were obtained.Brain volumes with synthetic DW stroke lesions were produced by warping therelative signal increase of real strokes to normal brain volumes. A generic 3DU-Net was trained on four different databases to generate four differentmodels: (a) 375 neuroradiologist-labeled clinical DW positive strokecases(CDB);(b) 2,000 synthetic cases(S2DB);(c) CDB+2,000 syntheticcases(CS2DB); or (d) CDB+40,000 synthetic cases(CS40DB). The models were testedon 20%(n=192) of the cases of the stroke database, which were excluded from thetraining set. Segmentation accuracy was characterized using Dice score andlesion volume of the stroke segmentation, and statistical significance wastested using a paired, two-tailed, Student's t-test. Detection sensitivity andspecificity was compared to three neuroradiologists. Results: The performanceof the 3D U-Net model trained on the CS40DB(mean Dice 0.72) was better thanmodels trained on the CS2DB (0.70,P <0.001) or the CDB(0.65,P<0.001). The deeplearning model was also more sensitive (91%[89%-93%]) than each of the threehuman readers(84%[81%-87%],78%[75%-81%],and 79%[76%-82%]), but lessspecific(75%[72%-78%] vs for the three human readers (96%[94%-97%],92%[90%-94%]and 89%[86%-91%]). Conclusion: Deep learning training for segmentation anddetection of DW stroke lesions was significantly improved by enhancing thetraining set with synthetic lesions.

Improved Segmentation and Detection Sensitivity of Diffusion-Weighted Brain Infarct Lesions with Synthetically Enhanced Deep Learning