Empirical Driven Automatic Detection of Lobulation Imaging Signs in Lung CT.

Guanghui Han,Nouman Q Soomro,Yanfeng Zhao,Xinming Zhao,Jia Sun,Xiabi Liu,Chunwu Zhou

doi:10.1155/2017/3842659

Abstract

Computer-aided detection (CAD) of lobulation can help radiologists to diagnose/detect lung diseases easily and accurately. Compared to CAD of nodule and other lung lesions, CAD of lobulation remained an unexplored problem due to very complex and varying nature of lobulation. Thus, many state-of-the-art methods could not detect successfully. Hence, we revisited classical methods with the capability of extracting undulated characteristics and designed a sliding window based framework for lobulation detection in this paper. Under the designed framework, we investigated three categories of lobulation classification algorithms: template matching, feature based classifier, and bending energy. The resultant detection algorithms were evaluated through experiments on LISS database. The experimental results show that the algorithm based on combination of global context feature and BOF encoding has best overall performance, resulting in F1 score of 0.1009. Furthermore, bending energy method is shown to be appropriate for reducing false positives. We performed bending energy method following the LIOP-LBP mixture feature, the average positive detection per image was reduced from 30 to 22, and F1 score increased to 0.0643 from 0.0599. To the best of our knowledge this is the first kind of work for direct lobulation detection and first application of bending energy to any kind of lobulation work.

Highlights

Lung disease is one of the fatal diseases for human
We performed A1 algorithm to test the effectiveness of a template matching method
We found that A6 (PHOW-Shape Context (SC) feature) does not increase the recall rate of lobulation detection, while it has higher computing efficiency than HOG feature

Summary

Introduction

Lung disease is one of the fatal diseases for human. The fatality ratio of lung cancer can be minimized if the cancer sings are detected and treated earlier. The computed tomography (CT) examinations play an important role in early detection and classification of lung lesions. We regard what radiologists see in lung nodules for diagnosing diseases as CT imaging signs, which are often called “CT features,” “CT findings,” “CT patterns,” or “CT manifestation.”. These CT imaging signs are very crucial in disease diagnosis and research works [2] We regard what radiologists see in lung nodules for diagnosing diseases as CT imaging signs, which are often called “CT features,” “CT findings,” “CT patterns,” or “CT manifestation.” These CT imaging signs are very crucial in disease diagnosis and research works [2]

Methods

Results

Conclusion