Developing an understanding of artificial intelligence lung nodule risk prediction using insights from the Brock model

Abstract

Objectives To determine if predictions of the Lung Cancer Prediction convolutional neural network (LCP-CNN) artificial intelligence (AI) model are analogous to the Brock model. Methods In total, 10,485 lung nodules in 4660 participants from the National Lung Screening Trial (NLST) were analysed. Both manual and automated nodule measurements were inputted into the Brock model, and this was compared to LCP-CNN. The performance of an experimental AI model was tested after ablating imaging features in a manner analogous to removing predictors from the Brock model. First, the nodule was ablated leaving lung parenchyma only. Second, a sphere of the same size as the nodule was implanted in the parenchyma. Third, internal texture of both nodule and parenchyma was ablated. Results Automated axial diameter (AUC 0.883) and automated equivalent spherical diameter (AUC 0.896) significantly improved the accuracy of Brock when compared to manual measurement (AUC 0.873), although not to the level of the LCP-CNN (AUC 0.936). Ablating nodule and parenchyma texture (AUC 0.915) led to a small drop in AI predictive accuracy, as did implanting a sphere of the same size as the nodule (AUC 0.889). Ablating the nodule leaving parenchyma only led to a large drop in AI performance (AUC 0.717). Conclusions Feature ablation is a feasible technique for understanding AI model predictions. Nodule size and morphology play the largest role in AI prediction, with nodule internal texture and background parenchyma playing a limited role. This is broadly analogous to the relative importance of morphological factors over clinical factors within the Brock model. Key Points • Brock lung cancer risk prediction accuracy was significantly improved using automated axial or equivalent spherical measurements of lung nodule diameter, when compared to manual measurements. • Predictive accuracy was further improved by using the Lung Cancer Prediction convolutional neural network, an artificial intelligence-based model which obviates the requirement for nodule measurement. • Nodule size and morphology are important factors in artificial intelligence lung cancer risk prediction, with nodule texture and background parenchyma contributing a small, but measurable, role.

3D-Morphomics, Morphological Features on CT scans for lung nodule malignancy diagnosis

“…Within the top 5 of features with highest predictive power we can observe 2 shape feature (2D size and volume) 2 texture feature (grey level non uniformity of GLSZM and GLDM) and a 3D-morphomics feature of high positive curvature. We also tested the Brock NLST model [24], which combines some clinical and annotators GT features into a logistic model, and obtained an AUC of 0.826, lower than the originally reported but partly confirming recent results [25].…”

Exploring the direction of the English translation of environmental protection articles based on the robot cognitive- emotional interaction model

“…the process to node 1, is the reward for node 6, node 5, node 4, node 3, and node 2 to obtain the tendency to node 1 action step by step, i.e., the periodic affective reward. The environment bonus is set according to (8). The energy replenishment point is the node where the robot replenishes internal energy during the article translation process; the dead-end node is the node where only the "return" action can be selected when there is a single word that cannot be searched; the trap point is the node where the robot loses additional internal energy at this node; and the normal node is the node with internal state gain and is not a dead-end node.…”

“…Emotional interaction cannot be achieved without the technical means of artificial intelligence [8]. For nearly two decades, AI researchers have been trying to empower machines with cognitive abilities to recognize, interpret, and express emotions [9].…”

Assessment of interobserver concordance in radiomic tools for lung nodule classification, with a focus on BRODERS and SILA

“…While only a limited number of radiomic tools have been clinically validated, including, Lung Cancer Prediction (LCP) tool (Optellum), BRODERS (Benign vs aggressive nODule Evaluation using Radiomic Stratification), and CANARY (Computer Aided Nodule Assessment and Risk Yield), many other models are being developed. While LCP automates this process 7 , mitigating inter-segmenter variability, though it lacks a proofreading mechanism for comprehensive nodule inclusion most nodule characterization tools, including BRODERS and CANARY, critically depend on nodule segmentation to determine to input for the model.…”