Artificial Intelligence-Aided Diagnosis Software to Identify Highly Suspicious Pulmonary Nodules

Prognostic Predicting Model of Pancreatic Body Tail Carcinoma Using Clinical and CT Radiomic Data

Objective: To collect the clinical, pathological, and computed tomography (CT) data of 143 accepted surgical cases of pancreatic body tail cancer (PBTC) and to model and predict its prognosis. Methods: The clinical, pathological, and CT data of 143 PBTC patients who underwent surgical resection or endoscopic ultrasound biopsy and were pathologically diagnosed in Xiangyang No.1 People's Hospital Hospital from December 2012 to December 2022 were retrospectively analyzed. The Kaplan-Meier method was adopted to make survival curves based on the 1 to 5 years' follow-up data, and then the log-rank was employed to analyze the survival. According to the median survival of 6 months, the PBTC patients were divided into a group with a good prognosis (survival time ≥ 6 months) and a group with a poor prognosis (survival time < 6 months), and further the training set and test set were set at a ratio of 7/3. Then logistic regression was conducted to find independent risk factors, establish predictive models, and further the models were validated. Results: The Kaplan-Meier analysis showed that age, diabetes, tumor, node, and metastasis stage, CT enhancement mode, peripancreatic lymph node swelling, nerve invasion, surgery in a top hospital, tumor size, carbohydrate antigen 19-9, carcinoembryonic antigen, Radscore 1/2/3 were the influencing factors of PBTC recurrence. The overall average survival was 7.4 months in this study. The multivariate logistic analysis confirmed that nerve invasion, surgery in top hospital, dilation of the main pancreatic duct, and Radscore 2 were independent factors affecting the mortality of PBTC (P < .05). In the test set, the combined model achieved the best predictive performance [AUC 0.944, 95% CI (0.826-0.991)], significantly superior to the clinicopathological model [AUC 0.770, 95% CI (0.615-0.886), P = .0145], and the CT radiomics model [AUC 0.883, 95% CI (0.746-0.961), P = .1311], with a good clinical net benefit confirmed by decision curve. The same results were subsequently validated on the test set. Conclusion: The diagnosis and treatment of PBTC are challenging, and survival is poor. Nevertheless, the combined model benefits the clinical management and prognosis of PBTC.

Fully automatic cardiac four chamber and great vessel segmentation on CT pulmonary angiography using deep learning

Introduction

Computed tomography pulmonary angiography (CTPA) is an essential test in the work-up of suspected pulmonary vascular disease including pulmonary hypertension and pulmonary embolism. Cardiac and great vessel assessments on CTPA are based on visual assessment and manual measurements which are known to have poor reproducibility. The primary aim of this study was to develop an automated whole heart segmentation (four chamber and great vessels) model for CTPA.

Methods

A nine structure semantic segmentation model of the heart and great vessels was developed using 200 patients (80/20/100 training/validation/internal testing) with testing in 20 external patients. Ground truth segmentations were performed by consultant cardiothoracic radiologists. Failure analysis was conducted in 1,333 patients with mixed pulmonary vascular disease. Segmentation was achieved using deep learning via a convolutional neural network. Volumetric imaging biomarkers were correlated with invasive haemodynamics in the test cohort.

Results

Dice similarity coefficients (DSC) for segmented structures were in the range 0.58-0.93 for both the internal and external test cohorts. The left and right ventricle myocardium segmentations had lower DSC of 0.83 and 0.58 respectively while all other structures had DSC >0.89 in the internal test cohort and >0.87 in the external test cohort. Interobserver comparison found that the left and right ventricle myocardium segmentations showed the most variation between observers: mean DSC (range) of 0.795 (0.785-0.801) and 0.520 (0.482-0.542) respectively. Right ventricle myocardial volume had strong correlation with mean pulmonary artery pressure (Spearman's correlation coefficient = 0.7). The volume of segmented cardiac structures by deep learning had higher or equivalent correlation with invasive haemodynamics than by manual segmentations. The model demonstrated good generalisability to different vendors and hospitals with similar performance in the external test cohort. The failure rates in mixed pulmonary vascular disease were low (<3.9%) indicating good generalisability of the model to different diseases.

Conclusion

Fully automated segmentation of the four cardiac chambers and great vessels has been achieved in CTPA with high accuracy and low rates of failure. DL volumetric biomarkers can potentially improve CTPA cardiac assessment and invasive haemodynamic prediction.