Abstract
PurposeIn this work, we address image segmentation in the scope of dosimetry using deep learning and make three main contributions: (a) to extend and optimize the architecture of an existing convolutional neural network (CNN) in order to obtain a fast, robust and accurate computed tomography (CT)-based organ segmentation method for kidneys and livers; (b) to train the CNN with an inhomogeneous set of CT scans and validate the CNN for daily dosimetry; and (c) to evaluate dosimetry results obtained using automated organ segmentation in comparison with manual segmentation done by two independent experts.MethodsWe adapted a performant deep learning approach using CT-images to delineate organ boundaries with sufficiently high accuracy and adequate processing time. The segmented organs were consequently used as binary masks for further convolution with a point spread function to retrieve the activity values from quantitatively reconstructed SPECT images for “volumetric”/3D dosimetry. The resulting activities were used to perform dosimetry calculations with the kidneys as source organs.ResultsThe computational expense of the algorithm was sufficient for clinical daily routine, required minimum pre-processing and performed with acceptable accuracy a Dice coefficient of 93% for liver segmentation and of 94% for kidney segmentation, respectively. In addition, kidney self-absorbed doses calculated using automated segmentation differed by 7% from dosimetry performed by two medical physicists in 8 patients.ConclusionThe proposed approach may accelerate volumetric dosimetry of kidneys in molecular radiotherapy with 177Lu-labelled radiopharmaceuticals such as 177Lu-DOTATOC. However, even though a fully automated segmentation methodology based on CT images accelerates organ segmentation and performs with high accuracy, it does not remove the need for supervision and corrections by experts, mostly due to misalignments in the co-registration between SPECT and CT images.Trial registration EudraCT, 2016-001897-13. Registered 26.04.2016, www.clinicaltrialsregister.eu/ctr-search/search?query=2016-001897-13.
Highlights
The molecular radiotherapy (MRT) using tumour-targeting peptide pharmacophores, labelled with radioisotopes such as Lu-177 or Y-90, is increasingly used for treatment of targetable cancers such as neuroendocrine tumours (NETs) [1,2,3], or prostate cancer [4]
Datasets The convolutional neural network (CNN) used in this work was trained and evaluated using databases as per the following: dataset 1, 2 and 3 were consisting of computed tomography (CT) data obtained from various sources used individuality to train, evaluate and test the network
Segmentation accuracy expressed as Dice score coefficient for segmented livers and kidneys is shown in Tables 1 and 2 in comparison with other top performing methods reported in the literature
Summary
The molecular radiotherapy (MRT) using tumour-targeting peptide pharmacophores, labelled with radioisotopes such as Lu-177 or Y-90, is increasingly used for treatment of targetable cancers such as neuroendocrine tumours (NETs) [1,2,3], or prostate cancer [4]. I.e. in order to safely administer MRT agents, various dosimetry methodologies have been developed to estimate and calculate the radiation doses delivered to various organs. The phantom-based dose estimators, lack [6] the specific patient and uptake geometry as the organs are standardized and a homogeneous activity distribution within each organ is assumed. To overcome these limitations, different patient-specific dosimetry methods have been adapted where the radiation dose is calculated on a voxel-by-voxel basis taking into consideration the individual organ shape and activity uptake
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
