U-Net-based transfer learning for automated tumour segmentation enabling fully automated [18F]F-DOPA PET analysis in paediatric gliomas.

PurposeTwo commonly used imaging techniques to aid in the diagnosis of neurodegenerative parkinsonian syndromes are dopamine transporter (DAT) imaging with [123I]-FP-CIT single-photon emission computed tomography (DAT-SPECT) and positron emission tomography with [18F]-FDOPA (FDOPA-PET). This paper provides a unique series of parkinsonian patients who received both FDOPA-PET and DAT-SPECT in routine clinical practice and compares the reported results to assess potential differences between these two imaging techniques.MethodsWe present 11 patients with a clinically uncertain parkinsonian syndrome (CUPS), who received both FDOPA-PET and DAT-SPECT. All patients received an FDOPA-PET scan and DAT-SPECT as part of routine clinical care.ResultsThe median time between the F-DOPA-PET scan and DAT-SPECT scan was 6 months (range 0–15 months). There was a discrepancy in the reported results of the FDOPA-PET and DAT-SPECT scans in nine patients, including 7 patients whose FDOPA-PET scan was reportedly normal, whereas their DAT-SPECT scan was abnormal.ConclusionsIn this case series of CUPS patients, DAT-SPECT was more often rated as abnormal than FDOPA-PET. The striatal loss of FDOPA uptake can be less pronounced than that of DAT binding in CUPS patients in early disease stages. Consequently, the interpretation of FDOPA-PET scans in CUPS can sometimes be challenging in routine practice.

PET imaging with [18F]F-DOPA has demonstrated high potential for the evaluation and management of pediatric brain gliomas. Manual extraction of PET parameters is time-consuming, lacks reproducibility, and varies with operator experience. In this study, we tested whether a semi-automated image processing framework could overcome these limitations. Pediatric patients with available static and/or dynamic [18F]F-DOPA PET studies were evaluated retrospectively. We developed a Python software to automate clinical index calculations, including preprocessing to delineate tumor volumes from structural MRI, accounting for lesions with low [18F]F-DOPA uptake. 73 subjects with treatment-naïve low and high-grade gliomas, who underwent brain MRI within two weeks of [18vF]F-DOPA PET, were included and analyzed. Static analysis was conducted on all subjects, while dynamic analysis was performed on 32 patients. For 68 subjects, the Intraclass Correlation Coefficient for T/S between manual and ground truth segmentation was 0.91. Using our tool, ICC improved to 0.94. Our method demonstrated good reproducibility in extracting static tumor-to-striatum ratio (p = 0.357), though significant differences were observed in tumor slope (p < 0.05). No significant differences were found in time-to-peak (p = 0.167) and striatum slope (p = 0.36). Our framework aids in analyzing [18F]F-DOPA PET images of pediatric brain tumors by automating clinical score extraction, simplifying segmentation and Time Activity Curve extraction, reducing user variability, and enhancing reproducibility.

Background/Objectives: PET imaging with [18F]F-DOPA has demonstrated high potential for the evaluation and management of pediatric brain gliomas. Manual extraction of PET parameters is time-consuming, lacks reproducibility, and varies with operator experience. Methods: In this study, we tested whether a semi-automated image processing framework could overcome these limitations. Pediatric patients with available static and/or dynamic [18F]F-DOPA PET studies were evaluated retrospectively. We developed a Python software to automate clinical index calculations, including preprocessing to delineate tumor volumes from structural MRI, accounting for lesions with low [18F]F-DOPA uptake. A total of 73 subjects with treatment-naïve low- and high-grade gliomas, who underwent brain MRI within two weeks of [18F]F-DOPA PET, were included and analyzed. Static analysis was conducted on all subjects, while dynamic analysis was performed on 32 patients. Results: For 68 subjects, the Intraclass Correlation Coefficient for T/S between manual and ground truth segmentation was 0.91. Using our tool, ICC improved to 0.94. Our method demonstrated good reproducibility in extracting static tumor-to-striatum ratio (p = 0.357); however, significant differences were observed in tumor slope (p < 0.05). No significant differences were found in time-to-peak (p = 0.167) and striatum slope (p = 0.36). Conclusions: Our framework aids in analyzing [18F]F-DOPA PET images of pediatric brain tumors by automating clinical score extraction, simplifying segmentation and Time Activity Curve extraction, reducing user variability, and enhancing reproducibility.

Diagnosing Parkinson's disease (PD) on clinical grounds may be difficult, especially in the early stages of the disease. F-DOPA PET and FP-CIT SPECT scans are able to determine presynaptic dopaminergic activity in different ways. The aim of this study was to determine and compare the sensitivity and specificity of the two methods in the detection of striatal dopaminergic deficits in the same cohort of PD patients and healthy controls. Movement disorder specialists recruited 11 patients with early-stage PD and 17 patients with advanced PD. The patients underwent both an FP-CIT SPECT scan and an F-DOPA PET scan. In addition, 10 FP-CIT SPECT scans or 10 F-DOPA PET scans were performed in 20 healthy controls. A template with regions of interest was used to sample tracer activity of the caudate, putamen and a reference region in the brain. The outcome parameter was the striatooccipital ratio (SOR). Normal SOR values were determined in the controls. The sensitivity and specificity of both scanning methods were calculated. FP-CIT SPECT and F-DOPA PET scans were both able to discriminate PD patients from healthy controls. For the early phases of the disease, sensitivity and specificity of the contralateral striatal and putaminal uptake of FP-CIT and F-DOPA was 100%. When only caudate uptake was considered, the specificities were 100% and 90% for FP-CIT and F-DOPA, respectively, while the sensitivity was 91% for both scanning techniques. FP-CIT SPECT and F-DOPA PET scans are both able to diagnose presynaptic dopaminergic deficits in early phases of PD with excellent sensitivity and specificity.

Automatic segmentation of gliomas on amino acid PET is essential for quantitative tumor assessment, a pillar in monitoring gliomas under treatment. This study aimed to develop a deep learning model for the automated extraction of PET RANO criteria from [18F]FDOPA PET, with external validation. A total of 635 static [18F]FDOPA PET scans from three European centers were retrospectively included for glioma diagnosis, recurrence assessment, or treatment monitoring. The training cohort comprised 530 scans from Nancy Hospital, with external validation and test sets from Pitié-Salpêtrière Hospital (n = 66) and Turin Hospital (n = 39). Ground truth segmentations followed international guidelines. A 3D U-Net was trained to segment tumor and healthy brain volumes. Performance was evaluated using the Dice coefficient using the whole tumor volume. Quantitative agreement for PET RANO criteria 1.0 parameters, tumor-to-background ratios (TBRmean, TBRmax) and metabolic tumor volume (MTV), was assessed at the lesion-level. Tumor segmentation achieved Dice of 0.925 [0.841; 0.970] in training, 0.885 [0.829; 0.925] in validation, and 0.851 [0.733; 0.911] in the test set. At lesion level, agreement with expert quantification was high, with low bias and strong reliability for MTV (2.293 [-4.734; 9.321] mL), TBRmax (0.056 [-0.189; 0.301]), and TBRmean (-0.139 [-0.424; 0.146]) and intra-class correlation coefficients superior to 0.93. Measurable lesions were correctly identified in more than 97% of cases. Our [18F]FDOPA PET deep learning model (available at https://github.com/IADI-Nancy/FDOPA-PET-GliomaSeg) demonstrates robust multicenter performance and enables fully automated, reproducible quantification, supporting broader clinical adoption of amino-acid PET in neuro-oncology.

A simple micromethod has been devised for estimating the fat and energy content of human milk based on the centrifugation of milk in a haematocrit centrifuge. The percentage of cream, or "creamatocrit," is read from the haematocrit capillary tube and is linearly related to the fat and energy content. The technique, which is rapid and cheap, may be used in clinical practice, in research, and in epidemiological studies.

Do static and dynamic insulin resistance indices perform similarly in predicting pre-diabetes and type 2 diabetes?

Deep learning (DL) models are gaining popularity in forest variable prediction using Earth observation (EO) images. However, in practical forest inventories, reference datasets are often represented by plot- or stand-level measurements, while high-quality representative wall-to-wall reference data for end-to-end training of DL models are rarely available. Transfer learning facilitates expansion of the use of deep learning models into areas with sub-optimal training data by allowing pretraining of the model in areas where high-quality teaching data are available. In this study, we perform a “model transfer” (or domain adaptation) of a pretrained DL model into a target area using plot-level measurements and compare performance versus other machine learning models. We use an earlier developed UNet based model (SeUNet) to demonstrate the approach on two distinct taiga sites with varying forest structure and composition. The examined SeUNet model uses multi-source EO data to predict forest height. Here, EO data are represented by a combination of Copernicus Sentinel-1 C-band SAR and Sentinel-2 multispectral images, ALOS-2 PALSAR-2 SAR mosaics and TanDEM-X bistatic interferometric radar data. The training study site is located in Finnish Lapland, while the target site is located in Southern Finland. By leveraging transfer learning, the SeUNet prediction achieved root mean squared error (RMSE) of 2.70 m and R2 of 0.882, considerably more accurate than traditional benchmark methods. We expect such forest-specific DL model transfer can be suitable also for other forest variables and other EO data sources that are sensitive to forest structure.

Tackling small sample survival analysis via transfer learning: A study of colorectal cancer prognosis.

PurposeSchizophrenia (SCZ) is a severe psychiatric disorder marked by abnormal dopamine synthesis, measurable through [18F]FDOPA PET imaging. This imaging technique has been proposed as a biomarker for treatment stratification in SCZ, where one-third of patients respond poorly to standard antipsychotics. This study explores the use of radiomics on [18F]FDOPA PET data to examine dopamine synthesis in SCZ and predict antipsychotic response.MethodsWe analysed 273 [18F]FDOPA PET scans from healthy controls (n = 138) and SCZ patients (n = 135) from multiple cohorts, including first-episode psychosis cases. Radiomic features from striatal regions were extracted using the MIRP Python package. Reproducibility was assessed with test–retest scans, selecting features with an intraclass correlation coefficient (ICC) > 0.80. These features were grouped via hierarchical clustering based on Spearman correlation. Regression analysis evaluated sex and age effects on radiomic features. Predictive power for treatment response was tested and compared to standard imaging analysis obtained from the Standardised Uptake Value ratio (SUVr) of striatal over cerebellar tracer activity.ResultsOut of 177 features, 15 met the ICC criteria (ICC: 0.81–0.99). Age and sex influenced features in patients but not in controls. The best performance were was by the GLCM joint maximum feature, which effectively differentiated responders from non-responders (AUC: 0.66–0.87), but did not reach statistical significance in classification over SUVr.ConclusionRadiomic analysis of [18F]FDOPA PET supports its use as a biomarker for assessing antipsychotic efficacy in schizophrenia, highlighting differential striatal tracer uptake based on patient response. While it provides modest classification improvements over standard imaging, further validation in larger datasets and integration with multivariate classification algorithms are needed.

Striatal dopamine synthesis capacity and neuromelanin in the substantia nigra: A multimodal imaging study in schizophrenia and healthy controls

Deep learning is a promising approach to increase reproducibility and time-efficiency of gross tumor volume (GTV) delineation in head and neck cancer, but model evaluation primarily relies on manual GTV delineations as reference annotation, which are subjective and tend to overestimate tumor volume. This study aimed to validate a deep learning model for laryngeal and hypopharyngeal GTV segmentation with pathology and to compare its performance with clinicians' manual delineations. A retrospective data set of 193 patients with laryngeal and hypopharyngeal cancer was used to train a deep learning model with clinical GTV delineations as reference. For validation, a data set comprising 18 patients who underwent imaging before total laryngectomy was used, with histopathology-based (n = 16) tumor delineations as ground truth. The performance of the automatic segmentations was compared with that of clinicians' manual delineations, both quantitatively and qualitatively. Median sensitivity (0.90 and 0.91) and largest required clinical target volume margin (6.4 and 6.6 mm) were comparable between automatic and manual GTV delineations. The positive predictive value yielded the only significant difference between automatic and manual GTV delineations, with medians of 0.52 and 0.61, respectively (P = .03). Clinical target volumes derived from automatic and manual GTVs exhibited similar sizes (median of 44.5 and 40.1 mL) and achieved a sensitivity of 1.00 in 13/16 and 14/16 tumors, respectively. Automatic segmentations were considered clinically acceptable in 67% of cases, compared with 63% of manual delineations. The proposed deep learning model for laryngeal and hypopharyngeal GTV segmentation achieved comparable results with clinicians' manual delineations, showing the potential for more consistency and efficiency in the radiation therapy workflow.

This study presents a novel approach to combine Digital Twins (DTs) with advanced deep learning (DL) models to deliver accurate, timely and actionable insights. Very high-resolution multi-temporal satellite images from various platforms were employed, covering pre-disaster, immediate post-disaster, and recovery phases. The dataset used consists of 1,303 labeled samples, distributed across three classes: ‘not damaged’, ‘recovered’, and ‘not recovered’. By leveraging transfer learning and spatial and multi-head attention mechanisms with state-of-the-art DL models, we significantly enhance model performance and reduce training time to address the (near) real time processing needs of DTs. Multi-head Attention ShallowNet (MAS), specifically developed for this task, achieved an accuracy of 83% with a training time of just 26 min. RegNetX002 demonstrated the highest accuracy (88%) but required significantly more time for training (519 min). Additionally, this study integrates Explainable AI methods, including Saliency Maps and Grad-CAM, to provide transparency, reliability checks, and detailed insights into the models’ decision-making processes. The results indicate that MAS demonstrates reliable performance with a balanced focus across pre-disaster, event, and post-disaster timeframes, effectively identifying critical regions for damage and recovery assessment. This comprehensive approach lays the foundation for integrated solutions in real-time response and recovery monitoring with DTs.

Background:Current treatments for schizophrenia act directly on dopamine (DA) receptors but are ineffective for many patients, highlighting the need to develop new treatment approaches. Striatal DA dysfunction, indexed using [18F]-FDOPA imaging, is linked to the pathoetiology of schizophrenia. We evaluated the effect of a novel drug, AUT00206, a Kv3.1/3.2 potassium channel modulator, on dopaminergic function in schizophrenia and its relationship with symptom change. Additionally, we investigated the test–retest reliability of [18F]-FDOPA PET in schizophrenia to determine its potential as a biomarker for drug discovery.Methods:Twenty patients with schizophrenia received symptom measures and [18F]-FDOPA PET scans, before and after being randomised to AUT00206 or placebo groups for up to 28 days treatment.Results:AUT00206 had no significant effect on DA synthesis capacity. However, there was a correlation between reduction in striatal dopamine synthesis capacity (indexed as Kicer) and reduction in symptoms, in the AUT00206 group (r = 0.58, p = 0.03). This was not observed in the placebo group (r = −0.15, p = 0.75), although the placebo group may have been underpowered to detect an effect. The intraclass correlation coefficients of [18F]-FDOPA indices in the placebo group ranged from 0.83 to 0.93 across striatal regions.Conclusions:The relationship between reduction in DA synthesis capacity and improvement in symptoms in the AUT00206 group provides evidence for a pharmacodynamic effect of the Kv3 channel modulator. The lack of a significant overall reduction in DA synthesis capacity in the AUT00206 group could be due to variability and the low number of subjects in this study. These findings support further investigation of Kv3 channel modulators for schizophrenia treatment. [18F]-FDOPA PET imaging showed very good test–retest reliability in patients with schizophrenia.

To study the relative rates of progression of early Parkinson's disease (PD) in patients started on a dopamine agonist, ropinirole, or L-dopa. A double-blind study of 45 early PD patients [mean age 61 +/- 9.8 SD and mean symptom duration, 26 +/- 16 SD months] randomized 2 : 1 (ropinirole : L-dopa). Supplementary L-dopa was allowed if, during the trial, there was lack of a therapeutic effect. (18)F-dopa PET scans were performed at baseline (n = 45) and 2 years (n = 37). At two years, the mean percentage reduction in putamen (18)F-dopa uptake (Ki(o)) was not significantly different between the two groups (13% ropinirole, n = 28 versus 18% L-dopa, n = 9). We found no significant overall difference in underlying PD progression, after two years treatment, between patients groups. In summary, (18)F-dopa PET can be employed to objectively evaluate the effect of potential neuroprotective agents on dopaminergic function.