Automatic Detection of Lung Cancers in Chest CT Images by the Variable N‐Quoit Filter

Abstract

SUMMARYThe authors have developed the quoit filter, which is a kind of mathematical morphological filter, for automatic extraction of candidate pathological areas of lung cancer. The method has problems, however, in processing speed or extraction accuracy. To overcome these problems, this paper proposes variable quoit filtering, in which the filter size is adjusted flexibly according to the pathological shadow, and distance transformation with gray-level weight is applied as preprocessing before the main filtering procedure. First, the performance of the method is analyzed using a model, and the effectiveness of the proposed method is shown. Then, trial applications to images of 82 actual cases (including 21 cancer areas) show that all of the cancer areas were correctly extracted. Compared to the conventional algorithm, the processing time is reduced to less than 1/20.

Image processing algorithm of computer-aided diagnosis in lung cancer screening by CT

“…The details will be found in Ref. 8. As the condition to provide a higher value for the tumor than for the vessel, it is assumed that the radius of the tumor region is larger on a relative scale than the radius of the vessel.…”

“…Consequently, the N-Q (new quoit) filter was developed as a new type of filter which emphasizes detection of location information while sacrificing the ability to restore the candidate shadow [8].…”

Reduction of false positives at vessel bifurcations in computerized detection of lung nodules

“…A number of research groups have reported a variety of CAD systems for detecting lung nodules in chest CT images, including multiple grayscale thresholding [5,6] , local density maximum algorithm [7] , fuzzy clustering [8] , genetic algorithm template matching of Gaussian spheres and discs [9] , filters enhancing spherical structures [10][11][12][13] , curved surface morphology analysis [14] , and volumetric curvature-based thresholding and region growing [15] . Commercial CAD systems for detecting lung nodules in chest CT images have also been developed, including the ImageChecker CT Lung system (R2 Technology Inc., Sunnyvale, CA, USA), Lung VCAR (GE Healthcare Technologies, Waukesha, WI, USA), and Syngo Lung CAD (Siemens Medical Solutions, Erlangen, Germany).…”

CT Image Based Computer-Aided Lung Cancer Diagnosis

“…The radiologists' exhaustion and physical tiredness might cause a wrong diagnosis especially for a group medical examination where most of CT images are healthy and only very few images involve the pathological changes. Therefore, some computer-aided diagnosis (CAD) systems have been developed to help their diagnosis work (Okumura et al, 1998;Lee et al, 1997;Yamamoto et al, 1994;Miwa et al, 1999). Core techniques of CAD systems can be found in feature extraction and pattern recognition.…”

“…Because of the fuzziness of the diagnosis target in the medical images, it often requires different methods from those for artificial targets. Miwa et al have developed a variable N-quoit filter to detect isolated pulmonary nodules (Miwa et al, 1999) and Homma et al have further improved the detection accuracy by discriminating between the isolated nodules and blood vessels those are both in a circle-like shape in CT images (Homma et al, 2008). The discrimination was achieved by developing new feature extraction techniques and combining those features extracted by the techniques.…”