Methodology for automatic detection of lung nodules in computerized tomography images

Abstract

Lung cancer is a disease with significant prevalence in several countries around the world. Its difficult treatment and rapid progression make the mortality rates among people affected by this illness to be very high. Aiming to offer a computational alternative for helping in detection of nodules, serving as a second opinion to the specialists, this work proposes a totally automatic methodology based on successive detection refining stages. The automated lung nodules detection scheme consists of six stages: thorax extraction, lung extraction, lung reconstruction, structures extraction, tubular structures elimination, and false positive reduction. In the thorax extraction stage all the artifacts external to the patient's body are discarded. Lung extraction stage is responsible for the identification of the lung parenchyma. The objective of the lung reconstruction stage is to prevent incorrect elimination of portions belonging to the parenchyma. Structures extraction stage comprises the selection of dense structures from inside the lung parenchyma. The next stage, tubular structures elimination eliminates a great part of the pulmonary trees. Finally, the false positive stage selects only structures with great probability to be nodule. Each of the several stages has very specific objectives in detection of particular cases of lung nodules, ensuring good matching rates even in difficult detection situations. We use 33 exams with diversified diagnosis and slices numbers for validating the methodology. We obtained a false positive per exam rate of 0.42 and false negative rate of 0.15. The total classification sensitivity obtained, measured out of the nodule candidates, was 84.84%. The specificity achieved was 96.15% and the total accuracy of the method was 95.21%.

Relationship between the Phase Angle and Volume of Tumours in Patients with Lung Cancer

“…It was based on an anticipated comparison of the mean values of tumour volume obtained in preliminary studies using the Bebúi software [14,15,16], and an assumption of type I and type II errors of 1%. The minimal sample size required was 30 subjects.…”

“…Lesion segmentation was performed using a semi-automatic process with the use of region growth and voxel aggregation algorithm, computing the Moran index and the Geary coefficient, which provide a global measurement of spatial autocorrelation [14,15]. The implemented algorithm calculates the value of the two spatial correlation functions for all voxel pairs that are within a certain spatial arrangement, taking into account the distance between the pairs and the direction and depth angles formed between them.…”

“…The developments in computer graphics and image processing algorithms, together with the rapid growth of computational performance, have made possible the development of highly sophisticated systems for the volumetric assessment of nodules and masses that are detected during tomography [14,15,16]. …”

Automatic Detection of Masses in Mammograms Using Quality Threshold Clustering, Correlogram Function, and SVM

“…5, not matching the expected shape of a mass, which is closer to "circular" [26]. Such clusters were eliminated after the analysis of the eccentricity, circularity, compactness [22], and circular density shape descriptors [29] and were classified by the SVM algorithm. Thus, several clusters with these characteristics were eliminated, which resulted in a significant reduction in false positives for the subsequent processing stages.…”

Computer-aided detection of lung nodules using outer surface features

“…CAD systems use extracted features from ROIs and VOIs to determine nodule or non-nodule candidate. Statistical, histogram, morphological and textural features can be used in the classification phase [13][14][15][16], as well as rule-based approaches [17,18], artificial neural networks [19,20], genetic algorithms [21], clustering algorithms [22], and support vector machines (SVM) [23][24][25].…”