Classification and evaluation strategies of auto-segmentation approaches for PET: Report of AAPM task group No. 211

Abstract

Purpose The purpose of this educational report is to provide an overview of the present state-of-the-art PET auto-segmentation (PET-AS) algorithms and their respective validation, with an emphasis on providing the user with help in understanding the challenges and pitfalls associated with selecting and implementing a PET-AS algorithm for a particular application. Approach A brief description of the different types of PET-AS algorithms is provided using a classification based on method complexity and type. The advantages and the limitations of the current PET-AS algorithms are highlighted based on current publications and existing comparison studies. A review of the available image datasets and contour evaluation metrics in terms of their applicability for establishing a standardized evaluation of PET-AS algorithms is provided. The performance requirements for the algorithms and their dependence on the application, the radiotracer used and the evaluation criteria are described and discussed. Finally, a procedure for algorithm acceptance and implementation, as well as the complementary role of manual and auto-segmentation are addressed. Findings A large number of PET-AS algorithms have been developed within the last 20 years. Many of the proposed algorithms are based on either fixed or adaptively selected thresholds. More recently, numerous papers have proposed the use of more advanced image analysis paradigms to perform semi-automated delineation of the PET images. However, the level of algorithm validation is variable and for most published algorithms is either insufficient or inconsistent which prevents recommending a single algorithm. This is compounded by the fact that realistic image configurations with low signal-to-noise ratios (SNR) and heterogeneous tracer distributions have rarely been used. Large variations in the evaluation methods used in the literature point to the need for a standardized evaluation protocol. Conclusions Available comparison studies suggest that PET-AS algorithms relying on advanced image analysis paradigms provide generally more accurate segmentation than approaches based on PET activity thresholds, particularly for realistic configurations. However, this may not be the case for simple shape lesions in situations with a narrower range of parameters, where simpler methods may also perform well. Recent algorithms which employ some type of consensus or automatic selection between several PET-AS methods have potential to overcome the limitations of the individual methods when appropriately trained. In either case, accuracy evaluation is required for each different PET scanner and scanning and image reconstruction protocol. For the simpler, less robust approaches, adaptation to scanning conditions, tumor type, and tumor location by optimization of parameters is necessary. The results from the method evaluation stage can be used to estimate the contouring uncertainty. All PET-AS contours should be critically verified by a physician. A standard test, i.e., a benchmark dedicated to ...

Does simplified quantitative analysis of 18F-FDG PET in cardiac inflammatory disease work?

“…Yet, for diseases presenting with low incidence rates, such as cardiac inflammatory processes making the proof of an added clinical value of the more advanced quantification tools using dynamic imaging will take a tremendous amount of time. In this context, comparison of historical, contemporary, and future 18 F-FDG cardiac inflammatory PET studies using refined second-order metrics as emphasized by Manabe et al 23 in this issue of the Journal of Nuclear Cardiology might provide a common reporting ground and enable pooling of data and experience over time.…”

“…[3][4][5] The detection of sarcoidosis, a systemic inflammatory disease characterized by non-necrotizing granuloma and with cardiac involvement in 20%-50% of cases that affects survival, 6,7 has benefited from these metabolic suppression interventions. This is attested by the growing evidence supporting possible diagnosis of this condition with cardiac 18 F-FDG PET 8 giving at the same time sizable information on prognosis 9 and on monitoring medical treatment. 10 Consensus is expected to promote standardized clinical scenarios, imaging preparation, acquisition, and reporting procedures so as to deliver the best level of evidence in favor of an added value of 18 F-FDG PET in the diagnosis and management of such conditions.…”

“…This is attested by the growing evidence supporting possible diagnosis of this condition with cardiac 18 F-FDG PET 8 giving at the same time sizable information on prognosis 9 and on monitoring medical treatment. 10 Consensus is expected to promote standardized clinical scenarios, imaging preparation, acquisition, and reporting procedures so as to deliver the best level of evidence in favor of an added value of 18 F-FDG PET in the diagnosis and management of such conditions. [11][12][13] A key component lies in quantifying the extent of the inflammatory process in a manner that enables pooling of data among institutions and low variability for repeated measurements.…”

“…The versatile substrate utilization of healthy myocardium is a persistent challenge when contemplating the use of 18 F-labeled fluorodeoxyglucose ( 18 F-FDG) for the depiction of cardiac inflammatory disease. 1,2 Strategies aiming at refraining from glucose utilization in the normal myocardium in a way to contrast with abnormal inflammatory processes (that exclusively rely on carbohydrates as metabolic substrate) have been advocated.…”

“…Along with visual qualitative interpretation of the static images, a number of image-derived PET metrics have been proposed in clinical oncology to achieve the common goal of objective and reproducible quantification of tracer uptake across vendors and over time that can also be applied to cardiac inflammatory PET studies: 14,15 -Basic first-order metrics provide a semiquantitative value reflecting tracer uptake in the hottest voxel (SUV max ) 16 or mean over a 1 mL spherical volume of interest (VOI) centered at the hottest voxel (SUV--Second-order metrics (SUV mean , MTV, TLG) add a refinement through definition of tumor volumes using either manual delineation (prone to large intra-and interobserver variability) or semi-or fully automated PET image segmentation techniques (usually provide better reproducibility of results). 10,17,18 -Third-order metrics (textural features and radiomics) remain an active field of research aiming at capitalizing on the heterogeneity of tracer uptake and its textural features within the selected VOI to discriminate patterns of disease. 19,20 -The most advanced quantification techniques rely on the use of parametric imaging derived from dynamic imaging and simplified or more complex compartmental modeling techniques in a way similar to coronary flow reserve quantification.…”