Gerhard Adam

Gerhard Adam

Auto-generated profile, based on publicly available information

Gerhard Adam

Department of Diagnostic and Interventional Radiology and Nuclear Medicine, University Medical Center Hamburg-Eppendorf, Hamburg, Germany

Research Focus & Metrics

Academic Career Length

4Years (2023-Present)

(Calculated once Scilit publication data is synced: current year − year of first publication + 1)

Representative Publications

View All Publications
1 January 2026 in Radiology: Artificial Intelligence
Verified
Journal Article
Closed access
IMIsabel Molwitz
Isabel Molwitz
IRInka Ristow
Inka Ristow
JEJennifer Erley
Jennifer Erley
TDTugba Akinci D'Antonoli
Tugba Akinci D'Antonoli
ATAli S. Tejani
Ali S. Tejani
LALisa Adams
Lisa Adams
Artificial intelligence has the potential to increase economic value in radiology, but its financial impact is highly context dependent, varying with task complexity, diagnostic performance, payment models, and willingness-to-pay thresholds for quality-adjusted life-year gains.

1 October 2024 in Radiology
Verified
Journal Article
Closed access
CRChristoph Riedel
Christoph Riedel
MHMarko Hoffmann
Marko Hoffmann
MIMuhammad Ismahil
Muhammad Ismahil
ALAlexander Lenz
Alexander Lenz
FPFelix Piecha
Felix Piecha
PBPeter Bannas
Peter Bannas
Four-dimensional flow MRI plus computational fluid dynamics for noninvasive portosystemic pressure gradient assessment in participants with cirrhosis with a transjugular intrahepatic portosystemic shunt (TIPS) was deemed feasible and potentially useful for detecting TIPS stenoses.

1 May 2023 in Radiology: Artificial Intelligence
Verified
Journal Article
Closed access
HRHaissam Ragab
Haissam Ragab
FWFabian Westhaeusser
Fabian Westhaeusser
AEAnne Ernst
Anne Ernst
JYJin Yamamura
Jin Yamamura
PFPatrick Fuhlert
Patrick Fuhlert
CSChristoph Schramm
Christoph Schramm
To develop, train, and validate a multiview deep convolutional neural network (DeePSC) for the automated diagnosis of primary sclerosing cholangitis (PSC) on two-dimensional MR cholangiopancreatography (MRCP) images. This retrospective study included two-dimensional MRCP datasets of 342 patients (45 years ± 14 [SD]; 207 male patients) with confirmed diagnosis of PSC and 264 controls (51 years ± 16; 150 male patients). MRCP images were separated into 3-T (n = 361) and 1.5-T (n = 398) datasets, of which 39 samples each were randomly chosen as unseen test sets. Additionally, 37 MRCP images obtained with a 3-T MRI scanner from a different manufacturer were included for external testing. A multiview convolutional neural network was developed, specialized in simultaneously processing the seven images taken at different rotational angles per MRCP examination. The final model, DeePSC, derived its classification per patient from the instance expressing the highest confidence in an ensemble of 20 individually trained multiview convolutional neural networks. Predictive performance on both test sets was compared with that of four licensed radiologists using the Welch t test. DeePSC achieved an accuracy of 80.5% ± 1.3 (sensitivity, 80.0% ± 1.9; specificity, 81.1% ± 2.7) on the 3-T and 82.6% ± 3.0 (sensitivity, 83.6% ± 1.8; specificity, 80.0% ± 8.9) on the 1.5-T test set and scored even higher on the external test set (accuracy, 92.4% ± 1.1; sensitivity, 100.0% ± 0.0; specificity, 83.5% ± 2.4). DeePSC outperformed radiologists in average prediction accuracy by 5.5 (P = .34, 3 T) and 10.1 (P = .13, 1.5 T) percentage points. Automated classification of PSC-compatible findings based on two-dimensional MRCP was achievable and demonstrated high accuracy on internal and external test sets. Keywords: Neural Networks, Deep Learning, Liver Disease, MRI, Primary Sclerosing Cholangitis, MR Cholangiopancreatography Supplemental material is available for this article. © RSNA, 2023

4 November 2024 in Orphanet Journal of Rare Diseases
Verified
Journal Article
Open access
PDF file
MSMarie-Lena Schmalhofer
Marie-Lena Schmalhofer
SFSaid Farschtschi
Said Farschtschi
LKLan Kluwe
Lan Kluwe
VMVictor Felix Mautner
Victor Felix Mautner
GAGerhard Adam
Gerhard Adam
IRInka Ristow
Inka Ristow
Background: Patients with neurofibromatosis type 1 (NF1) can develop plexiform neurofibromas (PN). Large tumor burden is a predictor for the development of malignant peripheral nerve sheath tumors. Whole-body magnetic resonance imaging (WB-MRI) is the recommended imaging method for the evaluation of PN. WB-MRI is recommended for NF1 patients at transition from adolescence to adulthood. In the absence of internal PN further follow-up WB-MRI is not considered necessary. PN are often detected in early childhood, leading to the assumption that they may be congenital lesions. It remains unclear whether this invariably applies to all patients or whether patients who initially displayed no tumors can still develop PN over time. Therefore, we retrospectively reviewed WB-MRI scans of pediatric patients with NF1 without initial tumor burden and compared these with long-term follow-up scans for presence of newly developed PN. Methods: We retrospectively reviewed WB-MRI scans of 17 NF1-children (twelve male; median age at initial scan: 9 [IQR 6.1–11.9] years) who initially displayed no PN. MRI scans with a follow-up interval of at least 6 years (median follow-up interval: 9 [IQR 5.6–12.4] years) were reviewed in consensus by two radiologists regarding the development of new PN over time. Results: New PN were identified in two out of 17 children without initial tumor burden in follow-up examinations. One of these two patients developed two larger distinct PN of 4.5 cm on the right upper arm and of 2.5 cm on the left thoracic wall between the age of ten and twelve. The second child developed multiple smaller PN along the major peripheral nerves between the age of eleven and 16. In addition, 15 of the children without initial tumor burden did not develop any distinct tumors for a period of at least 6 years. Conclusion: Our results indicate that PN can be newly detected in pediatric patients over time, even if no PN were detected on initial MRI scans. Therefore, it seems reasonable to perform at least a second MRI in pediatric NF1 patients at transition to adulthood, even if they did not display any tumor burden on initial MRI, and when the MRI was performed significantly under the age of 18. With this approach, tumors that may have developed between scans can be detected and patients at risk for complications can be identified.

27 December 2024 in EJNMMI Research
Verified
Journal Article
Open access
PDF file
IRInka Ristow
Inka Ristow
IAIvayla Apostolova
Ivayla Apostolova
MKMichael G. Kaul
Michael G. Kaul
MSMaria Stark
Maria Stark
AZAntonia Zapf
Antonia Zapf
LWLennart Well
Lennart Well
Background: To intraindividually compare the diagnostic performance of positron emission computed tomography (F-18-FDG-PET/CT) and diffusion-weighted magnetic resonance imaging (DW-MRI) in a non-inferiority design for the discrimination of peripheral nerve sheath tumours as benign (BPNST), atypical (ANF), or malignant (MPNST) in patients with neurofibromatosis type 1 (NF1). Results: In this prospective single-centre study, thirty-four NF1 patients (18 male; 30 ± 11 years) underwent F-18-FDG-PET/CT and multi-b-value DW-MRI (11 b-values 0 – 800 s/mm²) at 3T. Sixty-six lesions corresponding to 39 BPNST, 11 ANF, and 16 MPNST were evaluated. Two radiologists independently assessed the maximum standardized uptake value (SUVmax) and mean and minimum apparent diffusion coefficient (ADCmean/min) as well as the ADC in areas of lowest signal intensity in each lesion (ADCdark). The AUCs of DW-MRI and F-18-FDG-PET/CT were compared to determine whether the ADC is non-inferior to SUVmax (non-inferiority margin equal to -10%). Follow-up of ≥ 24 months (BPNST) or histopathological evaluation (MPNST + ANF) served as diagnostic reference standard. Both SUVmax and ADC parameters demonstrated good diagnostic accuracy (AUCSUVmax 94.0%; AUCADCmean/min/dark 91.6% / 90.1% / 92.5%). However, non-inferiority could not be demonstrated for any of the three ADC parameters (lower limits of the confidence intervals of the difference between the AUC of ADCmean/min/dark and SUVmax -12.9% / -14.5% / -11.6%). Inter-rater reliability was excellent for both imaging techniques (Krippendorff’s alpha all > 0.94). Conclusions: Both PET/CT-derived SUVmax and MRI-derived ADC allow sensitive and non-invasive differentiation of benign and (pre)-malignant peripheral nerve sheath tumours. Nevertheless, DW-MRI cannot be considered as non-inferior to F-18-FDG-PET/CT in this prospective single-centre study.

24 March 2025 in European Radiology
Verified
Journal Article
Open access
PDF file
IRInka Ristow
Inka Ristow
SZShuo ZHANG
Shuo ZHANG
CRChristoph Riedel
Christoph Riedel
ALAlexander Lenz
Alexander Lenz
RARalph Akoto
Ralph Akoto
LWLennart Well
Lennart Well
Objectives: To evaluate the feasibility and diagnostic performance of a 3D FRACTURE (fast field echo resembling a CT using restricted echo-spacing) MRI sequence for the detection and classification of proximal tibial fractures compared with CT. Methods: We retrospectively included 126 patients (85 male; 39.6 ± 14.5 years) from two centers following acute knee injury. Patients underwent knee MRI at 3 T including FRACTURE-MRI. Additional CT was performed in patients with tibial fractures (32.5%; n = 41) as the reference standard for fracture classification. Two radiologists independently evaluated FRACTURE-MRI for the presence of fractures and classified them according to AO/OTA, Schatzker, and the 10-segment classification. Diagnostic performance of FRACTURE-MRI was assessed using crosstabulations. Inter-reader agreement was estimated using Krippendorff’s alpha. Image quality was graded on a five-point scale (5 = excellent; 1 = inadequate definition of fracture lines and fracture displacement) and assessed using estimated marginal means. Results: Fractures were detected by FRACTURE-MRI with a sensitivity of 91.5% (83.2–96.5%) and a specificity of 97.1% (93.3–99.0%). Regarding fracture classification, diagnostic performances were slightly lower, with the 10-segment classification yielding the best sensitivity of 85.7% (81.4–89.3%) and specificity of 97.4% (96.6–98.0%), and the Schatzker classification yielding the lowest sensitivity of 78.2% (67.4–86.8%) and specificity of 97.7% (94.1–99.4%). Inter-reader agreement across the whole cohort was excellent (Krippendorff’s alpha 0.89–0.96) and when considering only patients with fractures, good to acceptable (0.48–0.91). Image quality was rated good (estimated marginal mean 4.3 (4.1–4.4)). Conclusion: FRACTURE-MRI is feasible at 3 T enabling accurate delineation of fracture lines for precise diagnosis and classification of proximal tibial fractures. Key Points: Question CT-like MRI is increasingly being evaluated for its advantages in bone imaging but is not yet established in routine practice. Findings The 3D FRACTURE (fast field echo resembling a CT using restricted echo-spacing) MRI sequence is feasible at 3 T, allowing for diagnosis and classification of proximal tibial fractures. Clinical relevance FRACTURE-MRI might be a helpful alternative to computed tomography in an acute trauma setting by reducing costs and radiation exposure in patients requiring a preoperative MRI anyway. Graphical