Error mitigation enables PET radiomic cancer characterization on quantum computers.

BACKGROUND: Cancer is a leading cause of death worldwide. While routine diagnosis of cancer is performed mainly with biopsy sampling, it is suboptimal to accurately characterize tumor heterogeneity. Positron emission tomography (PET)-driven radiomic research has demonstrated promising results when predicting clinical endpoints. This study aimed to investigate the added value of quantum machine learning both in simulator and in real quantum computers utilizing error mitigation techniques to predict clinical endpoints in various PET cancer patients. METHODS: Previously published PET radiomics datasets including 11C-MET PET glioma, 68GA-PSMA-11 PET prostate and lung 18F-FDG PET with 3-year survival, low-vs-high Gleason risk and 2-year survival as clinical endpoints respectively were utilized in this study. Redundancy reduction with 0.7, 0.8, and 0.9 Spearman rank thresholds (SRT), followed by selecting 8 and 16 features from all cohorts, was performed, resulting in 18 dataset variants. Quantum advantage was estimated by Geometric Difference (GDQ) score in each dataset variant. Five classic machine learning (CML) and their quantum versions (QML) were trained and tested in simulator environments across the dataset variants. Quantum circuit optimization and error mitigation were performed, followed by training and testing selected QML methods on the 21-qubit IonQ Aria quantum computer. Predictive performances were estimated by test balanced accuracy (BACC) values. RESULTS: On average, QML outperformed CML in simulator environments with 16-features (BACC 70% and 69%, respectively), while with 8-features, CML outperformed QML with + 1%. The highest average QML advantage was + 4%. The GDQ scores were ≤ 1.0 in all the 8-feature cases, while they were > 1.0 when QML outperformed CML in 9 out of 11 cases. The test BACC of selected QML methods and datasets in the IonQ device without error mitigation (EM) were 69.94% BACC, while EM increased test BACC to 75.66% (76.77% in noiseless simulators). CONCLUSIONS: We demonstrated that with error mitigation, quantum advantage can be achieved in real existing quantum computers when predicting clinical endpoints in clinically relevant PET cancer cohorts. Quantum advantage can already be achieved in simulator environments in these cohorts when relying on QML.

Machine learning predictive performance evaluation of conventional and fuzzy radiomics in clinical cancer imaging cohorts.

BACKGROUND: Hybrid imaging became an instrumental part of medical imaging, particularly cancer imaging processes in clinical routine. To date, several radiomic and machine learning studies investigated the feasibility of in vivo tumor characterization with variable outcomes. This study aims to investigate the effect of recently proposed fuzzy radiomics and compare its predictive performance to conventional radiomics in cancer imaging cohorts. In addition, lesion vs. lesion+surrounding fuzzy and conventional radiomic analysis was conducted. METHODS: Previously published 11C Methionine (MET) positron emission tomography (PET) glioma, 18F-FDG PET/computed tomography (CT) lung, and 68GA-PSMA-11 PET/magneto-resonance imaging (MRI) prostate cancer retrospective cohorts were included in the analysis to predict their respective clinical endpoints. Four delineation methods including manually defined reference binary (Ref-B), its smoothed, fuzzified version (Ref-F), as well as extended binary (Ext-B) and its fuzzified version (Ext-F) were incorporated to extract imaging biomarker standardization initiative (IBSI)-conform radiomic features from each cohort. Machine learning for the four delineation approaches was performed utilizing a Monte Carlo cross-validation scheme to estimate the predictive performance of the four delineation methods. RESULTS: Reference fuzzy (Ref-F) delineation outperformed its binary delineation (Ref-B) counterpart in all cohorts within a volume range of 938-354987 mm3 with relative cross-validation area under the receiver operator characteristics curve (AUC) of +4.7-10.4. Compared to Ref-B, the highest AUC performance difference was observed by the Ref-F delineation in the glioma cohort (Ref-F: 0.74 vs. Ref-B: 0.70) and in the prostate cohort by Ref-F and Ext-F (Ref-F: 0.84, Ext-F: 0.86 vs. Ref-B: 0.80). In addition, fuzzy radiomics decreased feature redundancy by approx. 20%. CONCLUSIONS: Fuzzy radiomics has the potential to increase predictive performance particularly in small lesion sizes compared to conventional binary radiomics in PET. We hypothesize that this effect is due to the ability of fuzzy radiomics to model partial volume effects and delineation uncertainties at small lesion boundaries. In addition, we consider that the lower redundancy of fuzzy radiomic features supports the identification of imaging biomarkers in future studies. Future studies shall consider systematically analyzing lesions and their surroundings with fuzzy and binary radiomics.

Association of norepinephrine transporter methylation with in vivo NET expression and hyperactivity-impulsivity symptoms in ADHD measured with PET.

Attention deficit hyperactivity disorder (ADHD) is a common neurodevelopmental disorder with a robust genetic influence. The norepinephrine transporter (NET) is of particular interest as it is one of the main targets in treatment of the disorder. As ADHD is a complex and polygenetic condition, the possible regulation by epigenetic processes has received increased attention. We sought to determine possible differences in NET promoter DNA methylation between patients with ADHD and healthy controls. DNA methylation levels in the promoter region of the NET were determined in 23 adult patients with ADHD and 23 healthy controls. A subgroup of 18 patients with ADHD and 18 healthy controls underwent positron emission tomography (PET) with the radioligand (S,S)-[18F]FMeNER-D2 to quantify the NET in several brain areas in vivo. Analyses revealed significant differences in NET methylation levels at several cytosine-phosphate-guanine (CpG) sites between groups. A defined segment of the NET promoter ("region 1") was hypermethylated in patients in comparison with controls. In ADHD patients, a negative correlation between methylation of a CpG site in this region and NET distribution in the thalamus, locus coeruleus, and the raphe nuclei was detected. Furthermore, methylation of several sites in region 1 was negatively associated with the severity of hyperactivity-impulsivity symptoms. Our results point to an epigenetic dysregulation in ADHD, possibly due to a compensatory mechanisms or additional factors involved in transcriptional processing.

Assessing Corticospinal Tract Asymmetry in Unilateral Polymicrogyria.

BACKGROUND AND PURPOSE: Asymmetry of the corticospinal tract in congenital lesions is a good prognostic marker for preserved motor function after hemispherectomy. This study aimed to assess this marker and provide a clinically feasible approach in selected cases of unilateral polymicrogyria. MATERIALS AND METHODS: Corticospinal tract asymmetry of 9 patients with unilateral polymicrogyria substantially affecting the central region was retrospectively assessed on axial T1WI and DTI. Volumes of the brain stem and thalamus and DTI parameters of the internal capsule were measured. Two neuroradiologists independently rated the right-left asymmetry at 4 levels along the corticospinal tract. DTI tractography was used to determine the motor cortex within polymicrogyria, with task-based functional MR imaging available in 3/9 cases. RESULTS: Visual assessment of the brain stem asymmetry showed excellent correlation with quantitative measures on both T1WI and color-coded DTI maps (P = .007 and P = .023). Interrater reliability regarding structural and DTI-based corticospinal tract asymmetry was best at the midbrain (Cohen κ = 0.77, P = .018). Three patients underwent functional hemispherectomy with postsurgical stable motor function, all showing marked corticospinal tract asymmetry preoperatively. Following the DTI-based corticospinal tract trajectories allowed identifying the presumed primary motor region within the dysplastic cortex in 9/9 patients, confirmed by functional MR imaging in 3/3 cases. CONCLUSIONS: Visual assessment of corticospinal tract asymmetry in unilateral polymicrogyria involving the motor cortex is most reliable with T1WI and color-coded DTI maps at the level of the midbrain. Pronounced asymmetry predicts preserved motor function after hemispherectomy. DTI-based tractography can be used as a guidance tool to the motor cortex within polymicrogyria.

Simple and rapid quantification of serotonin transporter binding using [11C]DASB bolus plus constant infusion.

INTRODUCTION: In-vivo quantification of serotonin transporters (SERT) in human brain has been a mainstay of molecular imaging in the field of neuropsychiatric disorders and helped to explore the underpinnings of several medical conditions, therapeutic and environmental influences. The emergence of PET/MR hybrid systems and the heterogeneity of SERT binding call for the development of efficient methods making the investigation of larger or vulnerable populations with limited scanner time and simultaneous changes in molecular and functional measures possible. We propose [11C]DASB bolus plus constant infusion for these applications and validate it against standard analyses of dynamic PET data. METHODS: [11C]DASB bolus/infusion optimization was performed on data acquired after [11C]DASB bolus in 8 healthy subjects. Subsequently, 16 subjects underwent one scan using [11C]DASB bolus plus constant infusion with Kbol 160-179min and one scan after [11C]DASB bolus for inter-method reliability analysis. Arterial blood sampling and metabolite analysis were performed for all scans. Distribution volumes (VT) were obtained using Logan plots for bolus scans and ratios between tissue and plasma parent activity for bolus plus infusion scans for different time spans of the scan (VT-70 for 60-70min after start of tracer infusion, VT-90 for 75-90min, VT-120 for 100-120min) in 9 subjects. Omitting blood data, binding potentials (BPND) obtained using multilinear reference tissue modeling (MRTM2) and cerebellar gray matter as reference region were compared in 11 subjects. RESULTS: A Kbol of 160min was observed to be optimal for rapid equilibration in thalamus and striatum. VT-70 showed good intraclass correlation coefficients (ICCs) of 0.61-0.70 for thalamus, striatal regions and olfactory cortex with bias ≤5.1% compared to bolus scans. ICCs increased to 0.72-0.78 for VT-90 and 0.77-0.93 for VT-120 in these regions. BPND-90 had negligible bias ≤2.5%, low variability ≤7.9% and ICCs of 0.74-0.87; BPND-120 had ICCs of 0.73-0.90. Low-binding cortical regions and cerebellar gray matter showed a positive bias of ~8% and ICCs 0.57-0.68 at VT-90. Cortical BPND suffered from high variability and bias, best results were obtained for olfactory cortex and anterior cingulate cortex with ICC=0.74-0.75 for BPND-90. High-density regions amygdala and midbrain had a negative bias of -5.5% and -22.5% at VT-90 with ICC 0.70 and 0.63, respectively. CONCLUSIONS: We have optimized the equilibrium method with [11C]DASB bolus plus constant infusion and demonstrated good inter-method reliability with accepted standard methods and for SERT quantification using both VT and BPND in a range of different brain regions. With as little as 10-15min of scanning valid estimates of SERT VT and BPND in thalamus, amygdala, striatal and high-binding cortical regions could be obtained. Blood sampling seems vital for valid quantification of SERT in low-binding cortical regions. These methods allow the investigation of up to three subjects with a single radiosynthesis.

Association of Protein Distribution and Gene Expression Revealed by PET and Post-Mortem Quantification in the Serotonergic System of the Human Brain.

Regional differences in posttranscriptional mechanisms may influence in vivo protein densities. The association of positron emission tomography (PET) imaging data from 112 healthy controls and gene expression values from the Allen Human Brain Atlas, based on post-mortem brains, was investigated for key serotonergic proteins. PET binding values and gene expression intensities were correlated for the main inhibitory (5-HT1A) and excitatory (5-HT2A) serotonin receptor, the serotonin transporter (SERT) as well as monoamine oxidase-A (MAO-A), using Spearman's correlation coefficients (rs) in a voxel-wise and region-wise analysis. Correlations indicated a strong linear relationship between gene and protein expression for both the 5-HT1A (voxel-wise rs = 0.71; region-wise rs = 0.93) and the 5-HT2A receptor (rs = 0.66; 0.75), but only a weak association for MAO-A (rs = 0.26; 0.66) and no clear correlation for SERT (rs = 0.17; 0.29). Additionally, region-wise correlations were performed using mRNA expression from the HBT, yielding comparable results (5-HT1Ars = 0.82; 5-HT2Ars = 0.88; MAO-A rs = 0.50; SERT rs = -0.01). The SERT and MAO-A appear to be regulated in a region-specific manner across the whole brain. In contrast, the serotonin-1A and -2A receptors are presumably targeted by common posttranscriptional processes similar in all brain areas suggesting the applicability of mRNA expression as surrogate parameter for density of these proteins.

Improved quality of life in patients treated with Peptide radionuclides.

Peptide receptor radionuclide therapy (PRRT) has recently been established as an important treatment modality for somatostatin receptor (SSTR)-positive tumors. The purpose of this study was to evaluate the clinical response, side-effects as well as the quality of life following (90)Y-DOTA-lanreotide (DOTALAN) and/or (90)Y-DOTA-Tyr (3)-DPhe(1)-octreotide (DOTATOC) therapy in patients with progressive metastatic disease during a 6-year follow-up period. Following dosimetric evaluation with (111)In-DOTALAN and (111)In-DOTATOC, 13 patients with estimated absorbed tumor doses of >5 Gy/GBq (carcinoid, n = 5; radioiodine-negative thyroid cancer, n = 4; gastrinoma, n = 1; insulinoma, n = 1; glucagonoma, n = 1; glomus jugularis tumor, n = 1) were assigned for PRRT. A dose of 925 MBq of (90)Y-DOTALAN (four patients) or 1.85-3.7 GBq of (90)Y-DOTATOC (10 patients) was administered intravenously and repeated every 4-8 weeks. Tumor dosimetry was performed prior to and under therapy, re-staging every 2-3 months. Pain intensity, Karnofsky score and general symptoms were evaluated in order to determine quality of life. Patients were followed until death. Altogether, 53 infusions of PRRT (1.85-14.1 GBq) were administered. After the first follow-up of 3 months of (90)Y-DOTALAN therapy, stable disease (SD) was observed in one patient and progressive disease (PD) in three patients. With (90)Y-DOTATOC therapy, SD was found in all 10 patients. During the re-evaluation period (4-27 months), one patient had to be shifted from (90)Y-DOTALAN to (90)Y-DOTATOC therapy due to reduced (111)In-DOTALAN uptake after 5.5 GBq. In the first 6 months after PRRT with DOTATOC, SD was found in nine of 10 patients and PD in one patient. Thereafter, SD was observed in two patients and PD in eight patients. Nine of 13 patients after PRRT with either DOTALAN or DOTATOC died. None of the patients had experienced severe acute hematological side-effects. Transient thrombocytopenia or lymphocytopenia was seen in 10 patients after 3.7 GBq, and a skin reaction in one patient. Total accumulated kidney dose ranged between 4 and 64 Gy, with reduced creatinine clearance in two patients. Pain relief was achieved in three of three patients after ~3.7 GBq ERT within 4-6 months. Appetite, weight, Karnofsky score and general well-being had improved in patients with SD during and after therapy. Based on the results of this study conducted on a small group of patients, we conclude that PRRT may offer an alternative treatment option for SSTR-positive tumors, with only mild transient side-effects and a marked improvement in the quality of life.

Preliminary experience with (68)Ga-DOTA-lanreotide positron emission tomography.

AIM: Positron emission tomography (PET) of (68)Ga-radiolabelled (SST) somatostatin receptor (R) binding peptides has recently been evaluated in SSTR positive tumor patients. First promising results in lung and thyroid tumor patients with (111)In-DOTA-Lanreotide (DOTA-LAN) scintigraphy have been described. We report our first experience with (68)Ga-labeled DOTA-LAN. METHODS: Eleven patients (3 non small cell lung cancer [NSCLC], 3 small cell lung cancer [SCLC], 3 radioiodine negative thyroid cancer, 2 medullary thyroid cancer [MTC]) were investigated. After intravenous injection of 75-150 MBq (68)Ga-DOTA-LAN dynamic studies were acquired over the tumor site for the first 40 min with a dedicated PET scanner in 3 patients, and 2 partial body scans were acquired at 20 and 50 min p.i. in 2 patients. Whole body acquisitions at 90 min after injection were acquired in all 11 patients. Image reconstruction was performed by iterative reconstruction utilizing additional transmission scans for attenuation correction. Vital parameters were recorded during the PET study and up to 24 h p.i. Blood and urinary sampling was done up to 4 hr after tracer injection in 8 patients. PET results were compared to conventional imaging techniques (CIT), i.e. computed tomography (CT) and/or magnetic resonance imaging (MRI). In 5 patients, (68)Ga-DOTA-LAN was compared with 2-[(18)F]fluoro-2-deoxy-D-glucose ((18)F-FDG). RESULTS: After intravenous (i.v.) injection of (68)Ga-DOTA-LAN the radioactivity in the blood rapidly decreased to less then 20% of the injected dose (ID) within the first 20 min and further decreased to less than 9% ID after 4 h. A cumulative urinary excretion of (68)Ga-DOTA-LAN up to 29.2 + or - 13.2% ID at 4 h was found. No acute side effects were observed. Tumor sites were visualized already during the first min after injection. Comparison of positron emission tomography (PET) and CIT showed concordant results in 3/8 patients and partial concordant results in 5/8 patients with matched results for the primary/recurrent tumor, mediastinal lymph nodes, or adrenal gland metastases. Partial concordant results were seen for the lung, bone, liver and cervical lymph node metastases. Micronodular metastases of the lung and the cerebrum were not visualized by (68)Ga-DOTA-LAN PET. The maximal standardized uptake values of the lung and bone tumor lesions ranged from 6 to 8 g/ml at 90 min p.i.. CONCLUSIONS: (68)Ga-DOTA-LAN visualized the majority of tumor lesions. Further studies are required to assess the clinical value, and to obtain the best imaging protocol of this new PET SSTR tracer.

Iodine-123-vascular endothelial growth factor-165 (123I-VEGF165). Biodistribution, safety and radiation dosimetry in patients with pancreatic carcinoma.

AIM: Imaging with radiolabelled vascular endothelial growth factor (VEGF) has been developed for the localisation and diagnosis of a variety of human solid tumors including gastrointestinal tumors. METHODS: In this study we investigated the biodistribution, safety and absorbed dose of iodine-123 radiolabelled VEGF(165) ((123)I-VEGF(165)) in 9 patients with pancreatic carcinoma. Following intravenous administration of (123)I-VEGF(165) (189+/-17 MBq; <130 pmole (<5 microg) VEGF(165) per patient), sequential images were recorded during the initial 30 min PI. Serial whole-body images were acquired in anterior and posterior views at various time points. All patients underwent single-photon emission tomography (SPET) imaging. Dosimetry calculations were performed on the basis of gamma camera data. Estimates of radiation absorbed dose were calculated using the MIRDOSE 3 program. RESULTS: The highest absorbed organ doses were found to be thyroid (0.058+/-0.004 mGy/MBq), spleen (0.046+/- 0.017 mGy/MBq), urinary bladder (0.04+/-0.02 mGy/MBq), lungs (0.034+/-0.009 mGy/MBq) and kidneys (0.033+/-0.005 mGy/MBq). The effective dose was estimated to be 0.017+/-0.002 mSv/MBq. A majority of primary pancreatic tumors and their metastases were visualized by (123)I-VEGF(165) scan. CONCLUSION: In vitro binding results confirmed specific binding of (123)I-VEGF(165) to pancreatic tumor cells and tissues. (123)I-VEGF(165) shows favorable dosimetry and is a safe radiopharmaceutical that may be of potential value for the imaging of VEGF receptor status in vivo.