Full kinetic modeling analysis of [18F]fluorocholine Positron Emission Tomography (PET) at initial diagnosis of high-grade glioma.

PURPOSE: The main objective was to characterize the tracer uptake kinetics of [18F]fluoromethylcholine ([18F]F-CHO) in high-grade gliomas (HGG) through a full PET kinetic modeling approach. Secondarily, we aimed to explore the relationship between the PET uptake measures and the HGG molecular features. MATERIALS AND METHODS: Twenty-four patients with a suspected diagnosis of HGG were prospectively included. They underwent a dynamic brain [18F]F-CHO-PET/CT, from which a tumoral time-activity curve was extracted. The plasma input function was obtained through arterial blood sampling with metabolite correction. These data were fitted to 1- and 2-tissue-compartment models, the best of which was selected through the Akaike information criterion. We assessed the correlation between the kinetic parameters and the conventional static PET metrics (SUVmax, SUVmean and tumor-to-background ratio TBR). We explored the association between the [18F]F-CHO-PET quantitative parameters and relevant molecular biomarkers in HGG. RESULTS: Tumoral time-activity curves in all patients showed a rapid rise of [18F]F-CHO uptake followed by a plateau-like shape. Best fits were obtained with near-irreversible 2-tissue-compartment models. The perfusion-transport constant K1 and the net influx rate Ki showed strong correlation with SUVmax (r = 0.808-0.861), SUVmean (r = 0.794-0.851) and TBR (r = 0.643-0.784), p < 0.002. HGG was confirmed in 21 patients, of which those with methylation of the O-6-methylguanine-DNA methyltransferase (MGMT) gene promoter showed higher mean Ki (p = 0.020), K1 (p = 0.025) and TBR (p = 0.001) than the unmethylated ones. CONCLUSION: [18F]F-CHO uptake kinetics in HGG is best explained by a 2-tissue-compartment model. The conventional static [18F]F-CHO-PET measures have been validated against the perfusion-transport constant (K1) and the net influx rate (Ki) derived from kinetic modeling. A relationship between [18F]F-CHO uptake rate and MGMT methylation is suggested but needs further confirmation.

A novel approach for skin lesion symmetry classification with a deep learning model.

Skin cancer has become a public health problem due to its increasing incidence. However, the malignancy risk of the lesions can be reduced if diagnosed at an early stage. To do so, it is essential to identify particular characteristics such as the symmetry of lesions. In this work, we present a novel approach for skin lesion symmetry classification of dermoscopic images based on deep learning techniques. We use a CNN model, which classifies the symmetry of a skin lesion as either "fully asymmetric", "symmetric with respect to one axis", or "symmetric with respect to two axes". Moreover, we introduce a new dataset of labels for 615 skin lesions. During the experimentation framework, we also evaluate whether it is beneficial to rely on transfer learning from pre-trained CNNs or traditional learning-based methods. As a result, we present a new simple, robust and fast classification pipeline that outperforms methods based on traditional approaches or pre-trained networks, with a weighted-average F1-score of 64.5%.

Development of a thin layer chromatography method for plasma correction of [18F]fluorocholine metabolites in positron emission tomography quantification studies in humans.

INTRODUCTION: After its intravenous injection, [18F]fluorocholine is oxidized by choline-oxidase into its main plasma metabolite, [18F]fluorobetaine. If PET kinetic modeling quantification of [18F]fluorocholine uptake is intended, the plasma input time-activity-curve of the parent tracer must be obtained, i.e., the fraction of the total plasma radioactivity corresponding to the nonmetabolized [18F]fluorocholine at each time has to be known. Hence our aim was to develop an easy-routine Thin-Layer-Chromatography (TLC) method to separate and quantify the relative fractions of [18F]fluorocholine and [18F]fluorobetaine as a function of time during PET imaging in humans. METHODS: First, we tested several combinations of solvents systems and layers to select the one showing the best resolution on non-radioactive standards. Thereafter, [18F]fluorobetaine was obtained through chemical oxidation of an [18F]fluorocholine sample at diferent incubation times and we applied the selected TLC-system to aliquots of this oxidation solution, both in a saline and in human deproteinized plasma matrices. The plates were detected by a radio-TLC-scanner. This TLC-system was finally applied to arterial plasma samples from 9 patients with high-grade-glioma undergoing brain PET imaging and a parent fraction curve was obtained in each of them. RESULTS: A TLC-system based on Silica-Gel-60//MeOH-NH3 was selected from the choline/betaine non-radioactive standards assay. Radiochromatograms of [18F]fluorocholine oxidation solution yielded two separated and well-defined peaks, Rf = 0,03 ([18F]fluorocholine) and Rf = 0.78 (18F]fluorobetaine) consistent with those observed on non-radioactive standards. During the oxidation, the [18F]fluorocholine radioactivity peak decreased progressively at several incubation times, while the other peak ([18F]fluorobetaine) increased accordingly. The mean values of the parent fraction of [18F]fluorocholine of the 9 patients studied (mean+/-SD) were 94% ±â€¯6%, 58% ±â€¯15%, 43% ±â€¯10%, 39% ±â€¯6% and 37% ±â€¯6% at 2.8 min, 5.8 min, 8.8 min, 11.7 min and 14.7 min post-injection, respectively. CONCLUSIONS: We have developed a TLC-system, easy to perform in a standard radiopharmacy unit, that enables the metabolite correction of arterial input function of [18F]fluorocholine in patients undergoing PET oncologic quantitative imaging.

Computational texture features of dermoscopic images and their link to the descriptive terminology: A survey.

Computer-extracted texture features are relevant to diagnose cutaneous lesions such as melanomas. Our goal is to set a relationship between a well-established descriptive terminology, which describes the attributes of dermoscopic structures based on their aspect rather than their underlying causes, and the computational methods to extract texture-based features. By tackling this problem, we can ascertain what indicators used by dermatologists are reflected in the extracted texture features. We first review the state-of-the-art models for texture extraction in dermoscopic images. By comparing the methods' performance and goals, we conclude that (I) a single color space does not seem to give performances as good as using several ones, thus the latter is reasonable (II) the optimal number of extracted features seems to vary depending on the method's goal, and extracting a large number of features can lead to a loss of models robustness (III) methods such as GLCM, Sobel or Law energy filters are mainly used to capture local properties to detect specific dermoscopic structures (IV) methods that extract local and global features, like Gabor wavelets or SPT, tend to be used to analyze the presence of certain patterns of dermoscopic structures, e.g. globular, reticular, etc.