Autosegmentation capability, which can improve the segmentation efficiency, is crucial for the ART process. In our clinical situation, the autosegmentation model trained using CT images in UIH TPS was regularly for rectal cancer patients’ target and organ delineation. To evaluate the performance of the autosegmentation model on sCT can indirectly evaluate the similarity between CT and sCT and check whether this autosegmentation model is suitable for sCT to improve the efficiency of ART process. So the segmentation model was used to delineate target and organ at risk on sCT. Then, the contours were reviewed and modified by an experienced physician on the sCT. The autocontours and manually modified contours were compared using dice similarity coefficients (DSC) to evaluate the autosegmentation accuracy on sCT.

In order to evaluate the performance difference of the autosegmentation model on sCT and CT. The segmentation model was used to delineate target and organ at risk on CT. Then these contours were transformed to the corresponding sCT. DSC index was used to evaluate the similarity between the autocontours from sCT and CT. The following formula was used to calculate DSC, in which V1 and V2 represents the volume of the two contours for comparison respectively.

Dose calculation capability is also very important of sCT for the ART accuracy. So the autoplanning function in UIH TPS was used to generate IMRT treatment plan on sCT to check whether clinical acceptable plans can be generated. The manually modified contours on sCT were used for planning. Then the plan and contours were transferred to the corresponding CT. The dose volume histogram was used to evaluate the dose distribution difference between sCT and CT based plans. V95%, V100% (volume of the target receiving at least 95% and 100% of the prescribed dose), D99, D5, D95 (doses to 99%, 5% and 95% of the volume) and Dmean (mean dose of the volume) were investigated for PTV (25). For OARs, volumes receiving different dose levels were evaluated. The dose volume statistics of V30, V40 and V50 for bladder and V30 and V40 for femoral heads were investigated (Vx means the percentage of volume receiving xGy dose). As comparison, the plan and structure were also transferred to CBCT and these dosimetric differences were also compared between CBCT-based and CT-based plans.

Note: The content above has been extracted from a research article, so it may not display correctly.

Please log in to submit your questions online.
Your question will be posted on the Bio-101 website. We will send your questions to the authors of this protocol and Bio-protocol community members who are experienced with this method. you will be informed using the email address associated with your Bio-protocol account.

We use cookies on this site to enhance your user experience. By using our website, you are agreeing to allow the storage of cookies on your computer.