New Paper in IEEE Journal of Biomedical and Health Informatics (Lung Tumor Delineation from PET Images)

In collaboration with Dr. Ivica Kopriva and Dr. Xinjian Chen, we have recently published a new paper for lung tumor segmentation using 3D PET images. Previously I and Xinjian published several papers on tumor segmentation from PET, PET/CT, and MRI/PET images. In this article, with Dr. Kopriva’s expertise and new approach, we considered the segmentation problem as a blind-source separation problem and successfully segmented lung tumors with high accuracy and efficiency. Tumor volume and its activity based measurements are highly significant in radiological and nuclear medicine interpretations where diagnosis, tumor staging, prognosis, and therapy planning are based on these measurements. I would like to congratulate our team for this successful developments. Hopefully, we will see this method in the routine clinics and improve healthcare with personalized medicine.


Single-channel Sparse Nonnegative Blind Source Separation Method for Automatic 3D Delineation of Lung Tumor in PET Images


Journal Website


In this paper, we propose a novel method for single-channel blind separation of non-overlapped sources and, to the best of our knowledge, apply it for the first time to automatic segmentation of lung tumors in Positron Emission Tomography (PET) images. Our approach first converts 3D PET image into a pseudo multichannel image. Afterwards, regularization free sparseness constrained nonnegative matrix factorization is used to separate tumor from other tissues. By using complexity based criterion, we select tumor component as the one with minimal complexity. We have compared the proposed method with threshold based on 40% and 50% maximum standardized uptake value (SUV), graph cuts (GC), random walks (RW) and affinity propagation (AP) algorithms on 18 non-small cell lung cancer datasets with respect to ground truth provided by two radiologists. Dice similarity coefficient averaged with respect to two ground truths is: 0.780.12 by the proposed algorithm, 0.780.1 by GC, 0.770.13 by AP, 0.770.07 by RW, and 0.750.13 by 50% maximum SUV threshold. Since the proposed method achieved performance comparable with interactive methods, considering the unique challenges of lung tumor segmentation from PET images, our findings support possibility of using our fully automated method in routine clinics. The source codes will be available at .
Screen Shot 2016-12-10 at 11.28.54 AM.png

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s