Limits...
A new method of detecting pulmonary nodules with PET/CT based on an improved watershed algorithm.

Zhao J, Ji G, Qiang Y, Han X, Pei B, Shi Z - PLoS ONE (2015)

Bottom Line: Then, an improved watershed method was used to mark suspicious areas on the CT image.Next, the support vector machine (SVM) method was used to classify SPNs based on textural features of CT images and metabolic features of PET images to validate the proposed method.Our proposed method was more efficient than traditional methods and methods based on the CT or PET features alone (sensitivity 95.6%; average of 2.9 false positives per scan).

View Article: PubMed Central - PubMed

Affiliation: College of Computer Science and Technology, Taiyuan University of Technology, Taiyuan 030024, China.

ABSTRACT

Background: Integrated 18F-fluorodeoxyglucose positron emission tomography/computed tomography (18F-FDG PET/CT) is widely performed for staging solitary pulmonary nodules (SPNs). However, the diagnostic efficacy of SPNs based on PET/CT is not optimal. Here, we propose a method of detection based on PET/CT that can differentiate malignant and benign SPNs with few false-positives.

Method: Our proposed method combines the features of positron-emission tomography (PET) and computed tomography (CT). A dynamic threshold segmentation method was used to identify lung parenchyma in CT images and suspicious areas in PET images. Then, an improved watershed method was used to mark suspicious areas on the CT image. Next, the support vector machine (SVM) method was used to classify SPNs based on textural features of CT images and metabolic features of PET images to validate the proposed method.

Results: Our proposed method was more efficient than traditional methods and methods based on the CT or PET features alone (sensitivity 95.6%; average of 2.9 false positives per scan).

No MeSH data available.


Related in: MedlinePlus

Comparison of the traditional methods based on SVM.
© Copyright Policy
Related In: Results  -  Collection

License
getmorefigures.php?uid=PMC4390287&req=5

pone.0123694.g009: Comparison of the traditional methods based on SVM.

Mentions: Blood vessels and non-nodules (nodules with inflammation) always had similar CT features compared to nodules. Table 1 shows that traditional method false-positives were relatively high and most were comprised of vessels and inflammation. Traditional methods are mainly based on nodular CT features and thus may misdiagnose vessels or non-nodule areas as SPNs. Moreover, these regions in PET images do not always have pronounced metabolic features. In contrast, our proposed method can differentiate SPNs from vessels and non-nodule areas more effectively. Fig 8 shows that traditional methods reviewed here misidentified these areas as SPNs, but our proposed method identified them as vessels or nodules with inflammation. Table 1 and Fig 9 confirm that our proposed methods are more sensitive than traditional methods, and that they offer better detection. We had the fewest false-positives (2.9/scan) than other methods and our sensitivity was not approached by the traditional methods reviewed here.


A new method of detecting pulmonary nodules with PET/CT based on an improved watershed algorithm.

Zhao J, Ji G, Qiang Y, Han X, Pei B, Shi Z - PLoS ONE (2015)

Comparison of the traditional methods based on SVM.
© Copyright Policy
Related In: Results  -  Collection

License
Show All Figures
getmorefigures.php?uid=PMC4390287&req=5

pone.0123694.g009: Comparison of the traditional methods based on SVM.
Mentions: Blood vessels and non-nodules (nodules with inflammation) always had similar CT features compared to nodules. Table 1 shows that traditional method false-positives were relatively high and most were comprised of vessels and inflammation. Traditional methods are mainly based on nodular CT features and thus may misdiagnose vessels or non-nodule areas as SPNs. Moreover, these regions in PET images do not always have pronounced metabolic features. In contrast, our proposed method can differentiate SPNs from vessels and non-nodule areas more effectively. Fig 8 shows that traditional methods reviewed here misidentified these areas as SPNs, but our proposed method identified them as vessels or nodules with inflammation. Table 1 and Fig 9 confirm that our proposed methods are more sensitive than traditional methods, and that they offer better detection. We had the fewest false-positives (2.9/scan) than other methods and our sensitivity was not approached by the traditional methods reviewed here.

Bottom Line: Then, an improved watershed method was used to mark suspicious areas on the CT image.Next, the support vector machine (SVM) method was used to classify SPNs based on textural features of CT images and metabolic features of PET images to validate the proposed method.Our proposed method was more efficient than traditional methods and methods based on the CT or PET features alone (sensitivity 95.6%; average of 2.9 false positives per scan).

View Article: PubMed Central - PubMed

Affiliation: College of Computer Science and Technology, Taiyuan University of Technology, Taiyuan 030024, China.

ABSTRACT

Background: Integrated 18F-fluorodeoxyglucose positron emission tomography/computed tomography (18F-FDG PET/CT) is widely performed for staging solitary pulmonary nodules (SPNs). However, the diagnostic efficacy of SPNs based on PET/CT is not optimal. Here, we propose a method of detection based on PET/CT that can differentiate malignant and benign SPNs with few false-positives.

Method: Our proposed method combines the features of positron-emission tomography (PET) and computed tomography (CT). A dynamic threshold segmentation method was used to identify lung parenchyma in CT images and suspicious areas in PET images. Then, an improved watershed method was used to mark suspicious areas on the CT image. Next, the support vector machine (SVM) method was used to classify SPNs based on textural features of CT images and metabolic features of PET images to validate the proposed method.

Results: Our proposed method was more efficient than traditional methods and methods based on the CT or PET features alone (sensitivity 95.6%; average of 2.9 false positives per scan).

No MeSH data available.


Related in: MedlinePlus