Quantitative Intensity Harmonization of Dopamine Transporter SPECT Images Using Gamma Mixture Models

Abstract

Purpose: Differences in site, device, and/or settings may cause large variations in the intensity profile of dopamine transporter (DAT) single-photon emission computed tomography (SPECT) images. However, the current standard to evaluate these images, the striatal binding ratio (SBR), does not efficiently account for this heterogeneity and the assessment can be unequivalent across distinct acquisition pipelines. In this work, we present a voxel-based automated approach to intensity normalize such type of data that improves on cross-session interpretation. Procedures: The normalization method consists of a reparametrization of the voxel values based on the cumulative density function (CDF) of a Gamma distribution modeling the specific region intensity. The harmonization ability was tested in 1342 SPECT images from the PPMI repository, acquired with 7 distinct gamma camera models and at 24 different sites. We compared the striatal quantification across distinct cameras for raw intensities, SBR values, and after applying the Gamma CDF (GDCF) harmonization. As a proof-of-concept, we evaluated the impact of GCDF normalization in a classification task between controls and Parkinson disease patients. Results: Raw striatal intensities and SBR values presented significant differences across distinct camera models. We demonstrate that GCDF normalization efficiently alleviated these differences in striatal quantification and with values constrained to a fixed interval [0,1]. Also, our method allowed a fully automated image assessment that provided maximal classification ability, given by an area under the curve (AUC) of AUC = 0.94 when used mean regional variables and AUC = 0.98 when used voxel-based variables. Conclusion: The GCDF normalization method is useful to standardize the intensity of DAT SPECT images in an automated fashion and enables the development of unbiased algorithms using multicenter datasets. This method may constitute a key pre-processing step in the analysis of this type of images.Alberto Llera and Ismael Huertas shared first authorship. Electronic supplementary material The online version of this article (https://

Mapping dopaminergic projections in the human brain with resting-state fMRI

“…However, here we were interested in the detailed spatial profile of DaT availability across the striatum. To obtain this spatial profile, we intensity-normalized all raw DaT SPECT images ( Llera et al, 2019 ) so as to optimize contrast in the DaT SPECT image and take into account variability in the DaT SPECT scans across the PPMI dataset as a result of different cameras and different scan sites. Finally, we averaged across all subjects and masked the striatum to obtain the average DaT SPECT image of the striatum.…”

Data harmonisation for information fusion in digital healthcare: A state-of-the-art systematic review, meta-analysis and future research directions

“…Gamma cumulative distribution functions alignment (GCDF): The voxel intensities were re-parameterised using a mixture model of two Gamma distributions that fitted a reference histogram [92] . This reparameterisation was based on the CDF of the Gamma component, which modelled the particular uptake, and constrained the new feature space to [0, 1].…”

“…The assessment is usually done by the Dice coefficient (Dice) or the intersection over union (IoU) [ 62 , 100 , 106 , 108 , 126 , 129 , 144 ]. On the contrary, the classification tasks are variously evaluated, e.g., using area under the receiver operating characteristics curve (AUC) [ 50 , 54 , 86 , 92 , 106 , 111 , 115 , 138 , 143 , 146 , 154 , 159 ], accuracy [ 38 , 48 , 50 , 59 , 60 , 62 , 63 , 69 , 77 , 89 , 99 , 106 , 131 133 , 140 , 143 , 145 , 159 ], true positive rate [ 67 , 135 ], sensitivity [ 48 , 143 ], specificity [48] and Matthews correlation coefficient (MCC) [69] . Note that MCC is a balanced measurement for the binary classification tasks, with comprehensive evaluations of TP, TN, FP, and FN, therefore it is divided into the main task based performance evaluation.…”

“…Radiomics refers to the extraction and analysis of a large number of quantitative image features from medical images obtained by CT, PET, MRI, and other imaging modalities [164] . In addition to the studies on phantoms [ 61 , 65 , 94 ], research refer to MRI mainly focus on the brain [ 17 , 47 49 , 51 53 , 55 , 57 , 58 , 63 , 66 , 86 , 87 , 90 92 , 95 , 97 , 99 , 100 , 115 , 117 - 119 , 125 , 126 , 129 131 , 143 , 144 , 157 ], breast [ 101 , 111 ], while those refer to CT focus on ear [60] , liver [ 50 , 64 ] and lung [ 34 , 120 ]. In addition to CT, PET and MRI, optical coherence tomography also suffers severe scanner variability, which can be eased by computational data harmonisation approaches [129] .…”

Mapping dopaminergic projections in the human brain with resting-state fMRI

“…However, here we were interested in the detailed spatial profile of DaT availability across the striatum. To obtain this spatial profile, we intensity-normalized all raw DaT SPECT images 34 so as to optimize contrast in the DaT SPECT image and take into account variability in the DaT SPECT scans across the PPMI dataset as a result of different cameras and different scan sites. Finally, we averaged across all subjects and masked the striatum to obtain the average DaT SPECT image of the striatum.…”