ABSTRACT
Objective:To explore the feasibility of predicting axillary lymph node metastasis of breast cancer using radiomics analysis based on dynamic contrast-enhanced (DCE) MRI.Methods:The retrospective study enrolled 163 patients (163 lesions) with breast cancer diagnosed by core needle biopsy from January 2013 to December 2013 in Peking University First Hospital. The status of axillary lymph nodes in all patients was pathologically confirmed, and they had complete preoperative breast MRI images. Among the 163 patients, 94 patients were confirmed with axillary lymph node metastasis, and 69 patients without axillary lymph node metastasis. They were randomly divided into the training dataset ( n=115) and testing dataset ( n=48) in a 7∶3 ratio. The radiomics analysis was performed in the training dataset, including image preprocessing and labeling, radiomics feature extraction, radiomics model establishment and model predictive performance inspection. Model performance was tested in the testing dataset. Receiver operating characteristic curve and area under curve (AUC) was used to analyze the model prediction performance. Results:Of the 1 075 features extracted from the training dataset, principal component analyses (PCA) features 8, 41 and 67 were selected by random forest classifier. The radiomics model including 3 PCA features reached an AUC of 0.956 (95%CI 0.907-0.988), with sensitivity of 91.2%, specificity of 100% and accuracy of 94.8%. In the testing dataset, the radiomics model including 3 PCA features reached an AUC of 0.767 (95%CI 0.652-0.890), with sensitivity of 80.8%, specificity of 72.7% and accuracy of 77.1%.Conclusion:It is feasible to predict axillary lymph node metastasis using radiomics features based on DCE-MRI of breast cancer.
ABSTRACT
Objective:To develop and validate a cascaded deep learning algorithm for kidney stone detection on plain CT images.Methods:Plain CT images of the patients with kidney stones were retrospectively archived from January 2018 to July 2018 in Peking University First Hospital. The cases were divided into two datasets according to the date of the CT scanning: training dataset （ n=30) and held-out test dataset （ n=29). The development of the kidney stone detection method consisted of three steps. First, a U-Net model was trained on the training dataset for kidney segmentation, and the model′s performance was estimated with the dice coefficient. Second, the thresholding and region growing methods were used to detect the stones in the renal region predicted by the trained U-Net model. Third, the stones′ lengths （maximal, middle and minimal length) and CT attenuation were calculated and integrated into a structured report automatically. Using the radiologist′s labels and measurements (maximal, middle, minimal length and CT attenuation) as ground truth, the stone detection algorithm performance was evaluated with sensitivity, specificity and precision, and the stone measurement algorithm performance was evaluated with Bland-Altman plots. Results:The held-out test dataset consisted of 29 cases, containing 58 kidneys and 11 358 CT slices. The 38 kidneys containing 56 stones and 20 kidneys did not contain stones. The U-Net model showed good performance, with a mean dice coefficient of 0.96. And 10 945 of 11 358 CT slices had a dice coefficient no less than 0.90. The sensitivity, precision, and specificity of stone detection were 100% （38/38), 100% （38/38) and 100% （20/20) in the organ-level. The sensitivity and precision of stone detection were 100% （56/56) and 96.6% （56/58) in the lesion-level.Conclusion:A cascaded algorithm is constructed and can be used to detect kidney stones in plain CT images. The algorithm can improve efficiency with results automatically integrated into the structured report in clinical practice.