Dataset related to article "Deep learning and atlas-based models to streamline the segmentation workflow of total marrow and lymphoid irradiation"

This record contains raw data related to article “Deep learning and atlas-based models to streamline the segmentation workflow of total marrow and lymphoid irradiation”

Purpose To improve the workfow of total marrow and lymphoid irradiation (TMLI) by enhancing the delineation of organsat risk (OARs) and clinical target volume (CTV) using deep learning (DL) and atlas-based (AB) segmentation models.Materials and methods Ninety-fve TMLI plans optimized in our institute were analyzed. Two commercial DL softwarewere tested for segmenting 18 OARs. An AB model for lymph node CTV (CTV_LN) delineation was built using 20 TMLIpatients. The AB model was evaluated on 20 independent patients, and a semiautomatic approach was tested by correctingthe automatic contours. The generated OARs and CTV_LN contours were compared to manual contours in terms of topological agreement, dose statistics, and time workload. A clinical decision tree was developed to defne a specifc contouringstrategy for each OAR.Results The two DL models achieved a median [interquartile range] dice similarity coefcient (DSC) of 0.84 [0.71;0.93]and 0.85 [0.70;0.93] across the OARs. The absolute median Dmean diference between manual and the two DL models was2.0 [0.7;6.6]% and 2.4 [0.9;7.1]%. The AB model achieved a median DSC of 0.70 [0.66;0.74] for CTV_LN delineation,increasing to 0.94 [0.94;0.95] after manual revision, with minimal Dmean diferences. Since September 2022, our institution has implemented DL and AB models for all TMLI patients, reducing from 5 to 2 h the time required to complete theentire segmentation process.Conclusion DL models can streamline the TMLI contouring process of OARs. Manual revision is still necessary for lymphnode delineation using AB models.