Clinical Applications of Radiomics in Nuclear Medicine.

Radiomics is an emerging field of artificial intelligence that focuses on the extraction and analysis of quantitative features such as intensity, shape, texture and spatial relationships from medical images. These features, often imperceptible to the human eye, can reveal complex patterns and biological insights. They can also be combined with clinical data to create predictive models using machine learning to improve disease characterization in nuclear medicine. This review article examines the current state of radiomics in nuclear medicine and shows its potential to improve patient care. Selected clinical applications for diseases such as cancer, neurodegenerative diseases, cardiovascular problems and thyroid diseases are examined. The article concludes with a brief classification in terms of future perspectives and strategies for linking research findings to clinical practice.

Interobserver Agreement Rates on Fibroblast Activation Protein Inhibitor-Directed Molecular Imaging and Therapy.

OBJECTIVES: Fibroblast activation protein (FAP) has emerged as a novel target for FAP inhibitor (FAPI)-directed molecular imaging and endoradiotherapy (ERT). We aimed to assess the interobserver agreement rates for interpretation of 68Ga-FAPI-4 PET/CT and decision for ERT. PATIENTS AND METHODS: A random order of 68Ga-FAPI-4 PET/CTs from 49 oncology patients were independently interpreted by 4 blinded readers. Per scan, visual assessment was performed, including overall scan impression, number of organ/lymph node (LN) metastases, and number of affected organs/LN regions. Moreover, a maximum of 3 target lesions, defined as largest in size and/or most intense, per organ compartment were identified, which allowed for an additional quantitative interobserver assessment of LN and organ lesions. To investigate potential reference tissues, quantification also included unaffected liver parenchyma and blood pool. Readers also had to indicate whether FAPI-directed ERT should be considered (based on intensity of uptake and widespread disease). Interobserver agreement rates were evaluated using intraclass correlation coefficients (ICCs) and interpreted according to Cicchetti (with 0.4-0.59 indicating fair, and 0.6-0.74 good, agreement). RESULTS: On a visual basis, the agreement rate for an overall scan impression was fair (ICC, 0.42; 95% confidence interval [CI], 0.27-0.57). The concordance rate for number of affected LN areas was also fair (ICC, 0.59; 95% CI, 0.45-0.72), whereas the number of LN metastases, number of affected organs, and number of organ metastases achieved good agreement rates (ICC, ≥0.63). In a quantitative analysis, concordance rates for LN were good (ICC, 0.70; 0.48-0.88), but only fair for organ lesions (ICC, 0.43; 0.26-0.60). In regards to background tissues, ICCs were good for unaffected liver parenchyma (0.68; 0.54-0.79) and fair for blood pool (0.43; 0.29-0.58). When readers should decide on ERT, concordance rates were also fair (ICC, 0.59; 95% CI, 0.46-0.73). CONCLUSIONS: For FAPI-directed molecular imaging and therapy, a fair to good interobserver agreement rate was achieved, supporting the adoption of this radiotracer for clinical routine and multicenter trials.

Fibroblast activation protein inhibitor (FAPi) positive tumour fraction on PET/CT correlates with Ki-67 in liver metastases of neuroendocrine tumours.

AIM: Gallium-68-labelled inhibitors of the fibroblast activation protein (FAPi) enable positron emission tomography/computed tomography (PET/CT) imaging of fibroblast activation. We evaluated if [68Ga]Ga-DATA5m.SA.FAPi PET/CT is related to Ki-67 as a marker of tumour aggressiveness in patients with liver metastases of NET. METHODS: Thirteen patients with liver metastases of a histologically confirmed NET who underwent PET/CT with [68Ga]Ga-DATA5m.SA.FAPi, [18F]FDG and [68Ga]Ga-DOTA-TOC were retrospectively analyzed. PET-positive liver tumour volumes were segmented for calculation of volume, SUVmax and PET-positive tumour fraction (TF). PET parameters were correlated with Ki-67. RESULTS: FDGSUVmax correlated positively (rho = 0.543, p < 0.05) and DOTATOCSUVmax correlated negatively (rho = -0.618, p < 0.05) with Ki-67, the correlation coefficients were in the moderate range. There was no significant correlation between FAPiSUVmax and Ki-67 (rho = 0.382, p > 0.05). FAPiTF correlated positively (rho = 0.770, p < 0.01) and DOTATOCTF correlated negatively (rho = -0.828, p < 0.01) with Ki-67, both significantly with high correlation coefficients. FDGTF also correlated significantly with Ki-67, with a moderate correlation coefficient (rho = 0.524, p < 0.05). The ratio FAPiVOL:DOTATOCVOL showed a significant and strong correlation with Ki-67 (rho = 0.808, p < 0.01). CONCLUSION: The ratio FAPiVOL:DOTATOCVOL might serve as a clinical parameter for the assessment of dedifferentiation and aggressiveness of liver metastases in patients with NET. [68Ga]Ga-DATA5m.SA.FAPi might hold potential for identification of high-risk patients. Further studies are warranted to evaluate its prognostic significance in comparison to [18F]FDG in patients with NET.

Tumor heterogeneity for differentiation between liver tumors and normal liver tissue in 18F-FDG PET/CT.

AIM: Malignancies show higher spatial heterogeneity than normal tissue. We investigated, if textural parameters from FDG PET describing the heterogeneity function as tool to differentiate between tumor and normal liver tissue. METHODS: FDG PET/CT scans of 80 patients with liver metastases and 80 patients with results negative upper abdominal organs were analyzed. Metastases and normal liver tissue were analyzed drawing up to three VOIs with a diameter of 25 mm in healthy liver tissue of the tumoral affected and results negative liver, whilst up to 3 metastases per patient were delineated. Within these VOIs 30 different textural parameters were calculated as well as SUV. The parameters were compared in terms of intra-patient and inter-patient variability (2-sided t test). ROC analysis was performed to analyze predictive power and cut-off values. RESULTS: 28 textural parameters differentiated healthy and pathological tissue (p < 0.05) with high sensitivity and specificity. SUV showed ability to differentiate but with a lower significance. 15 textural parameters as well as SUV showed a significant variation between healthy tissues out of tumour infested and negative livers. Mean intra- and inter-patient variability of metastases were found comparable or lower for 6 of the textural features than the ones of SUV. They also showed good values of mean intra- and inter-patient variability of VOIs drawn in liver tissue of patients with metastases and of results negative ones. CONCLUSION: Heterogeneity parameters assessed in FDG PET are promising to classify tissue and differentiate malignant lesions usable for more personalized treatment planning, therapy response evaluation and precise delineation of tumors for target volume determination as part of radiation therapy planning.

Tumour delineation in oesophageal cancer - A prospective study of delineation in PET and CT with and without endoscopically placed clip markers.

PURPOSE: The objective was to analyse the value of F-18-fluorodesoxyglucose (FDG)-positron emission tomography/computed tomography (PET/CT) for delineation of the Gross Tumour Volumes (GTVs) in primary radiotherapy of oesophageal cancer. METHOD: 20 consecutive and prospective patients (13 men, 7 women) underwent FDG-PET/CT for initial staging and radiation treatment planning. After endoscopy-guided clipping of the tumour another CT study was acquired. The CT and the FDG-PET/CT were registered with a rigid and a non-rigid registration algorithm to compare the overlap between GTV contours defined with the following methods: manual GTV definition in (1) the CT image of the FDG-PET/CT, (2) the PET image of the FDG-PET/CT, (3) the CT study based on endoscopic clips (CT clip), and (4) in the PET-data using different semi-automatic PET segmentation algorithms including a gradient-based algorithm. The absolute tumour volumes, tumour length in cranio-caudal direction, as well as the overlap with the reference volume (CT-clip) were compared for all lesions and separately for proximal/distal tumours. RESULTS: In 6 of the patients, FDG-PET/CT discovered previously unknown tumour locations, which resulted in either altered target volumes (n=3) or altered intent of treatment from curative to palliative (n=3) by upstaging to stage IV. For tumour segmentation a large variability between all algorithms was found. For the absolute tumour volumes with CT-clip as reference, no single PET-based segmentation algorithm performed better compared to using the manual CT delineation alone. The best correlation was found between the CT-clip and the gradient based segmentation algorithm (PET-edge, R(2)=0.84) as well as the manual CT-delineation (CT-manual R(2)=0.89). Non-rigid registration between CT and image FDG-PET/CT did not decrease variability between segmentation methods compared to rigid registration statistically significant. For the analysis of tumour length no homogeneous correlation was found. CONCLUSION: Whereas FDG-PET was highly relevant for staging purposes, CT imaging with clipping of the tumour extension remains the gold standard for GTV delineation.

Focal uptake of 68Ga-DOTATOC in the pancreas: pathological or physiological correlate in patients with neuroendocrine tumours?

PURPOSE: Neuroendocrine tumours are frequently located in the upper abdomen and especially in the pancreas. Imaging of the abdomen with somatostatin analogs such as (68)Ga-DOTA-Phe(1)-Tyr(3)-octreotide (DOTATOC) is a standard approach for imaging neuroendocrine cancer, but is still challenging due to physiological and technical considerations in this area. Therefore, the aim of this study was to further investigate the origin of (68)Ga-DOTATOC findings in the pancreas. METHODS: Forty-three consecutive patients with neuroendocrine tumours were examined by (68)Ga-DOTATOC positron emission tomography (PET)/CT for staging or restaging. As imaging of the upper abdomen is frequently affected by breathing artefacts, PET and CT data were analysed for misalignment and rearranged if necessary. Any noticeable uptake in the pancreas was described. Tracer uptake in the head of the pancreas and the liver was measured by means of maximum and average standard uptake value (SUV(max), SUV(av)). The reference standards (malignant versus benign) for correlation with PET findings were clinical and radiological follow-up (mean follow-up time 14 months) (n = 37) or histological confirmation (n = 6). RESULTS: In 23 of 43 studies (54%) misalignment between PET and CT data was found with a mean value of 1.4 cm. Visual assessment demonstrated that 20 of 43 scans (46.6%) showed no uptake in the head of the pancreas. Of 43 scans, 23 (53.4%) showed noticeable uptake with focal pattern in the head of the pancreas in 10 scans and irregular pattern in 13 scans. Follow-up indicated malignant pancreatic lesions in three patients. The pancreatic head to liver SUV(av) ratios in these patients ranged from 1.62 to 6.85, whereas in cases of uptake without known malignancy ratios ranged from 0.56 to 1.19. Considering SUV(max), the ratio ranged from 3.24 to 9.1 and from 0.84 to 1.47, respectively. No statistically significant difference was noted between uptake in the head of the pancreas and the liver in patients without malignant pancreatic tumours (p > 0.05). CONCLUSION: (68)Ga-DOTATOC uptake in the head of the pancreas is a common finding in patients undergoing (68)Ga-DOTATOC PET/CT. However, this finding most likely represents a physiological condition, especially if the uptake in the pancreatic head is similar to the uptake in the liver (uptake ratio head to liver SUV(av) < 1.4). Therefore, quantification is recommended to avoid false-positive diagnosis. Misalignment due to respiratory motion must always be taken into account.

The sensitivity of [11C]choline PET/CT to localize prostate cancer depends on the tumor configuration.

PURPOSE: To evaluate the dependency of the sensitivity of [(11)C]choline positron emission tomography/computed tomography (PET/CT) for detecting and localizing primary prostate cancer (PCa) on tumor configuration in the histologic specimen. EXPERIMENTAL DESIGN: Forty-three patients with biopsy-proven PCa were included. They underwent radical prostatectomy within 31 days after [(11)C]choline PET/CT. The transaxial image slices and the histologic specimens were analyzed by comparing the respective slices. Maximum standardized uptake values (SUV(max)) were calculated in each segment and correlated with histopathology. The tumor configuration in the histologic specimen was grouped as: I, unifocal; II, multifocal; III, rind-like shaped; IV, size <5 mm. Data analysis included the investigation of detection of PCa by SUV(max), the assessment of the influence of potential contributing factors on tumor prediction, and the evaluation of whether SUV could discriminate cancer tissue from benign prostate hyperplasia (BPH), prostatitis, HGPIN (high-grade prostate intraepithelial neoplasm), or normal prostate tissue. General estimation equation models were used for statistical analysis. RESULTS: Tumor configuration in histology was classified as I in 21 patients, as II in 9, as III in 5, and as IV in 8. The prostate segment involved by cancer is identified in 79% of the patients. SUV(max) was located in the same side of the prostate in 95% of patients. Tumor configuration was the only factor significantly negatively influencing tumor prediction (P < 0.001). PCa-SUV(max) (median SUV(max) = 4.9) was not significantly different from BPH-SUV (median SUV(max) = 4.5) and prostatitis-SUV (median SUV(max) = 3.9), P = 0.102 and P = 0.054, respectively. CONCLUSIONS: The detection and localization of PCa in the prostate with [(11)C]choline PET/CT is impaired by tumor configuration. Additionally, in our patient population, PCa tissue could not be distinguished from benign pathologies in the prostate.