Background: The rate of level V metastases is significantly low and the necessity of routine level V dissection for papillary thyroid microcarcinoma (PTMC) with clinically lateral lymph node metastasis (LNM) is still controversial. Methods: This study enrolled 114 consecutive PTMC patients with clinically suspected lateral LNM (N1b) who underwent modified radical neck dissection (levels II to V) at Xiangya Hospital of Central South University from September 2016 to July 2019. Univariate and multivariate analyses were performed to investigate the predictive factors of level V metastasis. The area under the receiver operating characteristic (ROC) curve (AUC), accuracy, specificity and sensitivity were used to determine the predictive value. Results: The overall and occult rate of level V metastasis were 29.82% (34/114) and 7.02% (8/114), respectively. Univariate analysis showed that level V metastasis was significantly associated with gross extrathyroidal extension (ETE), level IV metastasis and 2-level simultaneous metastasis (all P<0.05). Gross ETE (OR =11.916, 95% CI, 1.404-102.19; P=0.023) and level IV metastasis (OR =8.497, 95% CI, 2.119-34.065; P =0.03) served as independent predictors of level V metastasis in Nib PTMC patients. The sensitivity, specificity, accuracy, positive predictive value (PPV) and negative predictive value (NPV) of gross ETE and level IV metastasis in predicting the level V metastasis were 25.3% vs. 82.4%, 97.5% vs. 73.8%, 82.69% vs. 76.32%, 80% vs. 57.04% and 75% vs. 90.77%, respectively. The AUC of gross ETE was lower than level IV metastasis (0.605 vs. 0.781, P=0.041). Conclusions: Routine level V dissection is necessary in N1b PTMC patients with level IV metastasis or gross ETE. Compared with gross ETE, level IV metastasis is superior in predicting level V metastasis.

Accurate segmentation of lungs in pathological thoracic computed tomography (CT) scans plays an important role in pulmonary disease diagnosis. However, it is still a challenging task due to the variability of pathological lung appearances and shapes. In this paper, we proposed a novel segmentation algorithm based on random forest (RF), deep convolutional network, and multi-scale superpixels for segmenting pathological lungs from thoracic CT images accurately. A pathological thoracic CT image is first segmented based on multi-scale superpixels, and deep features, texture, and intensity features extracted from superpixels are taken as inputs of a group of RF classifiers. With the fusion of classification results of RFs by a fractional-order gray correlation approach, we capture an initial segmentation of pathological lungs. We finally utilize a divide-and-conquer strategy to deal with segmentation refinement combining contour correction of left lungs and region repairing of right lungs. Our algorithm is tested on a group of thoracic CT images affected with interstitial lung diseases. Experiments show that our algorithm can achieve a high segmentation accuracy with an average DSC of 96.45% and PPV of 95.07%. Compared with several existing lung segmentation methods, our algorithm exhibits a robust performance on pathological lung segmentation. Our algorithm can be employed reliably for lung field segmentation of pathologic thoracic CT images with a high accuracy, which is helpful to assist radiologists to detect the presence of pulmonary diseases and quantify its shape and size in regular clinical practices.