ATLAAS: an automatic decision tree-based learning algorithm for advanced image segmentation in positron emission tomography

Beatrice Berthon,Emiliano Spezi,Mererid Evans,Christopher Marshall

doi:10.1088/0031-9155/61/13/4855

Beatrice Berthon, Emiliano Spezi + Show 2 more

Open Access

https://doi.org/10.1088/0031-9155/61/13/4855

Copy DOI

Abstract

Accurate and reliable tumour delineation on positron emission tomography (PET) is crucial for radiotherapy treatment planning. PET automatic segmentation (PET-AS) eliminates intra- and interobserver variability, but there is currently no consensus on the optimal method to use, as different algorithms appear to perform better for different types of tumours. This work aimed to develop a predictive segmentation model, trained to automatically select and apply the best PET-AS method, according to the tumour characteristics.ATLAAS, the automatic decision tree-based learning algorithm for advanced segmentation is based on supervised machine learning using decision trees. The model includes nine PET-AS methods and was trained on a 100 PET scans with known true contour. A decision tree was built for each PET-AS algorithm to predict its accuracy, quantified using the Dice similarity coefficient (DSC), according to the tumour volume, tumour peak to background SUV ratio and a regional texture metric. The performance of ATLAAS was evaluated for 85 PET scans obtained from fillable and printed subresolution sandwich phantoms.ATLAAS showed excellent accuracy across a wide range of phantom data and predicted the best or near-best segmentation algorithm in 93% of cases. ATLAAS outperformed all single PET-AS methods on fillable phantom data with a DSC of 0.881, while the DSC for H&N phantom data was 0.819. DSCs higher than 0.650 were achieved in all cases.ATLAAS is an advanced automatic image segmentation algorithm based on decision tree predictive modelling, which can be trained on images with known true contour, to predict the best PET-AS method when the true contour is unknown. ATLAAS provides robust and accurate image segmentation with potential applications to radiation oncology.

Highlights

The past decade has seen an increasing interest in the use of positron emission tomography (PET) using 18F-fluorodeoxyglucose (18F-FDG) in radiation oncology
We have developed ATLAAS, an advanced and automatic image segmentation algorithm, based on the decision tree (DT) predictive modelling method
We have shown that ATLAAS can be trained to predict the best PET-based automatic segmentation (PET-AS) method when the ground truth is unknown and demonstrated that ATLAAS provides robust and accurate image segmentation that can potentially have wide applicability in radiation oncology, across multiple tumour types

Summary

Introduction

The past decade has seen an increasing interest in the use of positron emission tomography (PET) using 18F-fluorodeoxyglucose (18F-FDG) in radiation oncology. Supervised machine learning allows a given algorithm to be built and optimised using an existing training dataset for which the true tumour geometry and location is known (the ‘ground truth’), in order for it to make the right decisions to achieve optimal performance for cases in which the outcome is not known This includes methods such as K Nearest Neighbours (Anbeek et al 2005, Lyksborg et al 2012), Support Vector Machine (Iordanescu et al 2012, Jayachandran and Dhanasekaran 2013) and Artificial Neural Networks (Bankman 2000, Reyes-Aldasoro 2000), which have been used in the literature for the segmentation of medical imaging by classifying voxels into different categories. We describe the development and validation of a tool based on DT learning, designed to achieve this goal

Objectives

Methods

Results

Discussion

Conclusion