Ernst J. Rummeny

Online supplemental material is available for this article.

There is no generally accepted screening strategy for prostate cancer (PCa). From February 2014 to December 2019 a randomized trial (PROBASE) recruited 46 642 men at age 45 to determine the efficacy of risk‐adapted prostate‐specific antigen‐based (PSA) screening, starting at either 45 or 50 years. PSA tests are used to classify participants into a low (<1.5 ng/mL), intermediate (1.5‐2.99 ng/mL) or high (≥3 ng/mL) risk group. In cases of confirmed PSA values ≥3 ng/mL participants are recommended a prostate biopsy with multiparametric magnetic resonance imaging (mpMRI). Half of the participants (N = 23 341) were offered PSA screening immediately at age 45; the other half (N = 23 301) were offered digital rectal examination (DRE) with delayed PSA screening at age 50. Of 23 301 participants who accepted baseline PSA testing in the immediate screening arm, 89.2% fell into the low, 9.3% into intermediate, and 1.5% (N = 344) into the high risk group. Repeat PSA measurement confirmed high‐risk status for 186 men (0.8%), of whom 120 (64.5%) underwent a biopsy. A total of 48 PCas was detected (overall prevalence 0.2%), of which 15 had International Society of Uropathology (ISUP) grade 1, 29 had ISUP 2 and only 4 had ISUP ≥3 cancers. In the delayed screening arm, 23 194 participants were enrolled and 6537 underwent a DRE with 57 suspicious findings, two of which showed PCa (both ISUP 1; detection rate 0.03%). In conclusion, the prevalence of screen‐detected aggressive (ISUP ≥3) PCa in 45‐year‐old men is very low. DRE did not turn out effective for early detection of PCa.

Purpose: Brown adipose tissue (BAT) plays a potential role in energy and glucose metabolism in humans. Thyroid hormones (TH) are main regulators of BAT development and function. However, it remains unknown how the magnetic resonance (MR)-based proton density fat fraction (PDFF) of supraclavicular adipose tissue used as a surrogate marker for BAT presence relates to TH. Therefore, the purpose of this analysis was to investigate the relationship between supraclavicular PDFF and serum levels of TH. Methods: In total, 96 adult volunteers from a large cross-sectional study who underwent additional MR examination of the neck and pelvis were included in this analysis. Segmented PDFF maps of the supraclavicular and gluteal subcutaneous adipose tissue were generated. Delta PDFF was calculated as the difference between gluteal and supraclavicular PDFF and grouped as high (≥12%) or low (2. Statistical analyses included the use of paired samples ttest, simple linear regression analysis and a multivariable linear regression analysis. Results: The median age of the subjects (77% female) was 33 years, BMI ranged from 17.2 to 43.1 kg/m². Supraclavicular and gluteal PDFF differed significantly (76.5 ± 4.8 vs. 89.4 ± 3.5 %, p < 0.01). Supraclavicular PDFF was associated with FT3 in subjects with high delta PDFF (R² = 0.17, p < 0.01), with higher FT3 being associated with lower supraclavicular PDFF (y = 85.2 + –3.6 x). In a multivariable linear regression analysis considering further potential prognostic factors, the interaction between the delta PDFF group and FT3 remained a predictor for supraclavicular PDFF (B = –4.65, p < 0.01). Discussion/Conclusions: Supraclavicular PDFF corresponds to the presence of BAT. In the present analysis, supraclavicular PDFF is correlated with FT3 in subjects with high delta PDFF. Therefore, the present findings suggest that biologically active T3 may be involved in the development of supraclavicular BAT.

Imaging techniques such as computed tomographies (CT) play a major role in clinical imaging and diagnosis of malignant lesions. In recent years, metal nanoparticle platforms enabled effective payload delivery for several imaging techniques. Due to the possibility of surface modification, metal nanoparticles are predestined to facilitate molecular tumor targeting. In this work, we demonstrate the feasibility of anti-plasma membrane Heat shock protein 70 (Hsp70) antibody functionalized gold nanoparticles (cmHsp70.1-AuNPs) for tumor-specific multimodal imaging. Membrane-associated Hsp70 is exclusively presented on the plasma membrane of malignant cells of multiple tumor entities but not on corresponding normal cells, predestining this target for a tumor-selective in vivo imaging. In vitro microscopic analysis revealed the presence of cmHsp70.1-AuNPs in the cytosol of tumor cell lines after internalization via the endo-lysosomal pathway. In preclinical models, the biodistribution as well as the intratumoral enrichment of AuNPs were examined 24 h after i.v. injection in tumor-bearing mice. In parallel to spectral CT analysis, histological analysis confirmed the presence of AuNPs within tumor cells. In contrast to control AuNPs, a significant enrichment of cmHsp70.1-AuNPs has been detected selectively inside tumor cells in different tumor mouse models. Furthermore, a machine-learning approach was developed to analyze AuNP accumulations in tumor tissues and organs. In summary, utilizing mHsp70 on tumor cells as a target for the guidance of cmHsp70.1-AuNPs facilitates an enrichment and uniform distribution of nanoparticles in mHsp70-expressing tumor cells that enables various microscopic imaging techniques and spectral-CT-based tumor delineation in vivo.

Objective: To measure T2 values for magnetic resonance neurography (MRN) of the healthy distal sciatic nerve and compare those to T2 changes in patients with nerve compression.Materials and methods: Twenty-one healthy subjects and five patients with sciatica due to disc herniation underwent MRN using a T2-prepared turbo spin echo (TSE) sequence of the distal sciatic nerve bilaterally. Six and one of those healthy subjects further underwent a commonly used multi-echo spin-echo (MESE) sequence and magnetic resonance spectroscopy (MRS), respectively.Results: T2 values derived from the T2-prepared TSE sequence were 44.6 ± 3.0 ms (left) and 44.5 ± 2.6 ms (right) in healthy subjects and showed good inter-reader reliability. In patients, T2 values of 61.5 ± 6.2 ms (affected side) versus 43.3 ± 2.4 ms (unaffected side) were obtained. T2 values of MRS were in good agreement with measurements from the T2-prepared TSE, but not with those of the MESE sequence.Discussion: A T2-prepared TSE sequence enables precise determination of T2 values of the distal sciatic nerve in agreement with MRS. A MESE sequence tends to overestimate nerve T2 compared to T2 from MRS due to the influence of residual fat on T2 quantification. Our approach may enable to quantitatively assess direct nerve affection related to nerve compression.

Zielsetzung Im Rahmen dieser Studie sollten simulierte Sparse-Sampling Computertomografie (SpSCT) Bilder verschiedener Dosisstufen hinsichtlich subjektiver und objektiver Kriterien bezüglich der Diagnostik von Lungenarterienembolien (LAE) evaluiert werden. Material und Methoden 20 klinische CT-Pulmonalangiografien (50% mit der Diagnose LAE) wurden verwendet, um Datensätze mit verschiedenen Dosisleveln (DL) zu simulieren; hierbei wurden 6 Stufen mit 100%-50%-25%-12,5%-6,3%-3,1% der ursprünglichen Dosis generiert. Bei diesen voll-Sampling (full sampling, FS) Daten wurde jede Projektion für die Bilderstellung verwendet. Von diesen FS-Daten wurde nun jede zweite (2-SpSCT) oder vierte (4-SpSCT) Projektion verwendet, um Sparse-Sampling Datensätze zu simulieren. Jeder dieser Datensätze wurde von vier verblindeten Radiologen hinsichtlich subjektiver Kriterien evaluiert und die Richtigkeit der Diagnose wurde ausgewertet. Als objektiver Parameter wurde zusätzlich das Kontrast-zu-Rausch Verhältnis (KRV) bestimmt. Ergebnisse Die Sensitivität betrug 100% mit 2-SpSCT und 4-SpSCT beim 25% und 12,5% DL, unabhängig von der Lokalisation der LAE (eine Subgruppe 98,5%). Mit 2-SpSCT und 4-SpSCT konnten bei allen Dosisstufen höhere Sensitivitäten und Spezifitäten erreicht werden als mit FS. Die subjektive Bildqualität war bei allen DL mit 4-SpSCT signifikant höher als mit FS (p<0.01). Lediglich mit 4-SpSCT wurden allen Datensätzen beim 12,5% DL eine diagnostische Bildqualität bestätigt. SpSCT führte bei niedrigen Dosisstufen zu einem signifikant höheren KRV als FS. Schlußfolgerungen Mit SpSCT wäre in der aktuellen Studie eine Dosisreduktion von 88% bei gleichzeitig beibehaltener hoher Bildqualität und voller diagnostischer Konfidenz möglich, entsprechend einer effektiven Dosis von 0.38 mSv. Zukünftig könnte somit durch SpSCT die Dosis für den Patienten reduziert werden oder die Dosis pro Projektion bei insgesamt konstanter Dosis erhöht werden, wodurch insbesondere bei Niedrigdosisuntersuchungen eine Bildverbesserung erreicht werden könnte.

Zielsetzung Mithilfe von Convolutional Neural Networks (CNN) soll die Erkennung des hyperdensen Mediazeichens in der klinischen Diagnostik verbessert werden. Das hyperdense Mediazeichen ist ein Frühzeichen des ischämischen Schlaganfalles und somit von hoher klinischer Bedeutung. Da dieses Zeichen von Radiologen leicht übersehen werden kann, wird in der hier vorgestellten Arbeit ein CNN-basiertes computerassistiertes Diagnosesystem (CAD) entwickelt. Material und Methoden Native CT Scans von 216 Patienten wurden aus dem klinischen Bildarchiv anonymisiert und exportiert. Von den 101 Patienten mit sichtbarem hyperdensen Mediazeichen wurde dieses von einer Radiologin segmentiert (7 Jahre neuroradiologische Erfahrung). Mit den gewonnenen Daten wurden mehrere CNNs trainiert und in einer Kaskade kombiniert (3D CNN, U-Net). Schlussendlich wurde eine quantitative Evaluierung auf Patientenebene durchgeführt, indem die Fläche unter der ROC-Kurve (AUC) bestimmt wurde. Ergebnisse Die vorgestellte Methode erreicht einen AUC-Wert von 0.77 (95% Konfidenzintervall von 0.62 – 0.90). Die Größe des Testdatensatzes beträgt 68 Patienten (34 gesund, 34 mit hyperdensem Mediazeichen). Im Vergleich zu Black-Box CNNs können durch die Verwendung von mehreren CNNs Zwischenergebnisse wie Segmentierungen und Scores für die einzelnen Läsionen erhalten werden. Somit kann die Entscheidung von einem Radiologen nachvollzogen werden, was sich als essentiell für die Einbindung in eine klinische Umgebung erweist. Schlußfolgerungen Wir haben ein System entwickelt, welches langfristig das Potential bietet, Radiologen bei der Schlaganfalldiagnostik zu unterstützen. Die Methode kann ein CT-Volumen automatisch klassifizieren. Somit bietet sie das Potential, als “drittes Auge” fälschlicherweise negative Diagnosen zu vermeiden.

To evaluate image quality and confidence for planning of periradicular infiltrations using virtually lowered tube currents and in-house developed iterative reconstruction (IR) for multidetector computed tomography (MDCT).

Background: Rheumatoid arthritis (RA) is one of the most common rheumatic diseases. Joint inflammation and pathological growth of joint cartilage cause swollen and painful joints, which severely diminishes the patients’ life quality. There is no causal treatment. Symptomatic therapies should start as early as possible to take maximal effect. Hence, diagnostic procedures capable of detecting affected joints before the onset of clinical symptoms are highly desirable. We explored the value of PET imaging of integrin subtypes αvβ3 and α5β1 for early detection of RA foci in collagen-induced arthritis (CIA) mouse models. Results: Development of RA in CIA mice was monitored by paw scoring, and αvβ3- and α5β1-integrin expression was quantified by μPET using ⁶⁸Ga-Avebetrin and ⁶⁸Ga-Aquibeprin. For consecutive sections of selected decalcified joints (knee, ankle), arthritic degeneration and integrin expression were assessed by MOVAT staining and β3/α5 immunohistochemistry (IHC), respectively. β3- and α5-IHC revealed elevated levels of both αvβ3- and α5β1-integrin in arthritic joints. Unlike αvβ3, α5β1 is strongly expressed in the proliferating synovial lining layer, which suggests that its presence is directly related to RA development. For mice with advanced RA (6 weeks after CIA), PET signals for α5β1-integrin were substantially stronger (> 300% of baseline) than that of αvβ3-integrin (< 200%). A longitudinal PET follow-up revealed that the manifestation of clinical symptoms of RA is preceded by upregulation of α5β1- but not of αvβ3-integrin. Conclusion: α5β1-integrin PET could add a new functional imaging aspect to the portfolio of RA diagnostics because it appears to be a sensitive biomarker for early RA development. We suggest α5β1-integrin PET as a valuable tool to achieve a higher precision for early diagnosis of RA, including initial staging, monitoring of the disease course, and drug treatment, and for planning of radiosynoviorthesis (RSO).