Formation and progression of atherosclerotic plaque is a dynamic and complex process involving various pathophysiologic steps including inflammation and calcification. The purpose of this study was to compare macrophage activity as determined by (18)F-FDG PET and ongoing mineral deposition as measured by (18)F-sodium fluoride PET in atherosclerotic plaque and to correlate these findings with calcified plaque burden as assessed by CT. Forty-five patients were examined by whole-body (18)F-FDG PET, (18)F-sodium fluoride PET, and CT. Tracer uptake in various arterial segments was analyzed both qualitatively and semiquantitatively by measuring the blood-pool-corrected standardized uptake value (target-to-background ratio [TBR]). The pattern of tracer uptake in atherosclerotic lesions was compared after color-coded multistudy image fusion of PET and CT studies. The Fisher exact test and the Spearman correlation coefficient r(s) were used for statistical analysis of image-based results and cardiovascular risk factors. Intra- and interrater reproducibility were evaluated using the Cohen κ. (18)F-sodium fluoride uptake was observed at 105 sites in 27 (60%) of the 45 study patients, and mean TBR was 2.3 ± 0.7. (18)F-FDG uptake was seen at 124 sites in 34 (75.6%) patients, and mean TBR was 1.5 ± 0.3. Calcified atherosclerotic lesions were observed at 503 sites in 34 (75.6%) patients. Eighty-one (77.1%) of the 105 lesions with marked (18)F-sodium fluoride uptake and only 18 (14.5%) of the 124 lesions with (18)F-FDG accumulation were colocalized with arterial calcification. Coincident uptake of both (18)F-sodium fluoride and (18)F-FDG was observed in only 14 (6.5%) of the 215 arterial lesions with radiotracer accumulation. PET/CT with (18)F-FDG and (18)F-sodium fluoride may allow evaluation of distinct pathophysiologic processes in atherosclerotic lesions and might provide information on the complex interactions involved in formation and progression of atherosclerotic plaque.

¹⁸F-FDG PET/CT is effective in the assessment of therapy response. Changes in glucose uptake or tumor size are used as a measure. Tumor heterogeneity was found to be a promising predictive and prognostic factor. We investigated textural parameters for their predictive and prognostic capability in patients with rectal cancer using histopathology as the gold standard. In addition, a comparison to clinical outcome was performed. Methods: Twenty-seven patients with rectal cancer underwent ¹⁸F-FDG PET/CT before, 2 wk after the start, and 4 wk after the completion of neoadjuvant chemoradiotherapy. In all PET/CT scans, conventional parameters (tumor volume, diameter, maximum and mean standardized uptake values, and total lesion glycolysis [TLG]) and textural parameters (coefficient of variation [COV], skewness, and kurtosis) were determined to assess tumor heterogeneity. Values on pretherapeutic PET/CT as well as changes early in the course of therapy and after therapy were compared with histopathologic response. In addition, the prognostic value was assessed by correlation with time to progression and survival time. Results: The COV showed a statistically significant capability to assess histopathologic response early in therapy (sensitivity, 68%; specificity, 88%) and after therapy (79% and 88%, respectively). Thereby, the COV had a higher area under the curve in receiver-operating-characteristic analysis than did any analyzed conventional parameter for early and late response assessment. The COV showed a statistically significant capability to evaluate disease progression and to predict survival, although the latter was not statistically significant. Conclusion: Tumor heterogeneity assessed by the COV, being superior to the investigated conventional parameters, is an important predictive factor in patients with rectal cancer. Furthermore, it can provide prognostic information. Therefore, its application is an important step for personalized treatment of rectal cancer.

Textural features in FDG-PET have been shown to provide prognostic information in a variety of tumor entities. Here we evaluate their predictive value for recurrence and prognosis in NSCLC patients receiving primary stereotactic radiation therapy (SBRT). 45 patients with early stage NSCLC (T1 or T2 tumor, no lymph node or distant metastases) were included in this retrospective study and followed over a median of 21.4 months (range 3.1–71.1). All patients were considered non-operable due to concomitant disease and referred to SBRT as the primary treatment modality. Pre-treatment FDG-PET/CT scans were obtained from all patients. SUV and volume-based analysis as well as extraction of textural features based on neighborhood gray-tone difference matrices (NGTDM) and gray-level co-occurence matrices (GLCM) were performed using InterView Fusion™ (Mediso Inc., Budapest). The ability to predict local recurrence (LR), lymph node (LN) and distant metastases (DM) was measured using the receiver operating characteristic (ROC). Univariate and multivariate analysis of overall and disease-specific survival were executed. 7 out of 45 patients (16%) experienced LR, 11 (24%) LN and 11 (24%) DM. ROC revealed a significant correlation of several textural parameters with LR with an AUC value for entropy of 0.872. While there was also a significant correlation of LR with tumor size in the overall cohort, only texture was predictive when examining T1 (tumor diameter < = 3 cm) and T2 (>3 cm) subgroups. No correlation of the examined PET parameters with LN or DM was shown. In univariate survival analysis, both heterogeneity and tumor size were predictive for disease-specific survival, but only texture determined by entropy was determined as an independent factor in multivariate analysis (hazard ratio 7.48, p = .016). Overall survival was not significantly correlated to any examined parameter, most likely due to the high comorbidity in our cohort. Our study adds to the growing evidence that tumor heterogeneity as described by FDG-PET texture is associated with response to radiation therapy in NSCLC. The results may be helpful into identifying patients who might profit from an intensified treatment regime, but need to be verified in a prospective patient cohort before being incorporated into routine clinical practice.

Gliomas are the most common types of tumors in the brain. While the definite diagnosis is routinely made ex vivo by histopathologic and molecular examination, diagnostic work-up of patients with suspected glioma is mainly done by using magnetic resonance imaging (MRI). Nevertheless, L-S-methyl-¹¹C-methionine (¹¹C-MET) Positron Emission Tomography (PET) holds a great potential in characterization of gliomas. The aim of this study was to establish machine learning (ML) driven survival models for glioma built on ¹¹C-MET-PET, ex vivo and patient characteristics. Methods: 70 patients with a treatment naïve glioma, who had a positive ¹¹C-MET-PET and histopathology-derived ex vivo feature extraction, such as World Health Organization (WHO) 2007 tumor grade, histology and isocitrate dehydrogenase (IDH1-R132H) mutation status were included. The ¹¹C-MET-positive primary tumors were delineated semi-automatically on PET images followed by the feature extraction of tumor-to-background ratio based general and higher-order textural features by applying five different binning approaches. In vivo and ex vivo features, as well as patient characteristics (age, weight, height, body-mass-index, Karnofsky-score) were merged to characterize the tumors. Machine learning approaches were utilized to identify relevant in vivo, ex vivo and patient features and their relative weights for 36 months survival prediction. The resulting feature weights were used to establish three predictive models per binning configuration based on a combination of: in vivo/ex vivo and clinical patient information (M36IEP), in vivo and patient-only information (M36IP), and in vivo only (M36I). In addition a binning-independent ex vivo and patient-only (M36EP) model was created. The established models were validated in a Monte Carlo (MC) cross-validation scheme. Results: Most prominent ML-selected and -weighted features were patient and ex vivo based followed by in vivo features. The highest area under the curve (AUC) values of our models as revealed by the MC cross-validation were: 0.9 (M36IEP), 0.87 (M36EP), 0.77 (M36IP) and 0.72 (M36I). Conclusion: Survival prediction of glioma patients based on amino acid PET using computer-supported predictive models based on in vivo, ex vivo and patient features is highly accurate.

Quantum machine learning has experienced significant progress in both software and hardware development in the recent years and has emerged as an applicable area of near-term quantum computers. In this work, we investigate the feasibility of utilizing quantum machine learning (QML) on real clinical datasets. We propose two QML algorithms for data classification on IBM quantum hardware: a quantum distance classifier (qDS) and a simplified quantum-kernel support vector machine (sqKSVM). We utilize these different methods using the linear time quantum data encoding technique ( $${\mathrm{log}}_{2}N$$log 2 N ) for embedding classical data into quantum states and estimating the inner product on the 15-qubit IBMQ Melbourne quantum computer. We match the predictive performance of our QML approaches with prior QML methods and with their classical counterpart algorithms for three open-access clinical datasets. Our results imply that the qDS in small sample and feature count datasets outperforms kernel-based methods. In contrast, quantum kernel approaches outperform qDS in high sample and feature count datasets. We demonstrate that the $${\mathrm{log}}_{2}N$$log 2 N encoding increases predictive performance with up to + 2% area under the receiver operator characteristics curve across all quantum machine learning approaches, thus, making it ideal for machine learning tasks executed in Noisy Intermediate Scale Quantum computers.

Survival prediction in patients undergoing radionuclide therapy based on intratumoral somatostatin-receptor heterogeneity