Lenhard Pennig

Dr. Lenhard Pennig


Research Focus & Metrics

Academic Career Length

8Years (2019-Present)

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

Representative Publications

View All Publications
1 April 2024 in Radiology
Verified
Journal Article
Closed access
RGRoman Johannes Gertz
Roman Johannes Gertz
TDThomas Dratsch
Thomas Dratsch
ABAlexander Christian Bunck
Alexander Christian Bunck
SLSimon Lennartz
Simon Lennartz
AIAndra-Iza Iuga
Andra-Iza Iuga
JKJonathan Kottlors
Jonathan Kottlors
GPT-4 (ChatGPT; OpenAI) matched the performance of radiologists in detecting errors in radiology reports and has the potential to reduce processing time and cost.

25 May 2021 in Journal of Magnetic Resonance Imaging
Verified
Journal Article
Open access
PDF file
SJStephanie T. Jünger
Stephanie T. Jünger
UHUlrike Cornelia Isabel Hoyer
Ulrike Cornelia Isabel Hoyer
DSDiana Schaufler
Diana Schaufler
KLKai Roman Laukamp
Kai Roman Laukamp
LGLukas Goertz
Lukas Goertz
LPLenhard Pennig
Lenhard Pennig
Background: Non‐small cell lung cancer (NSCLC) is the most common tumor entity spreading to the brain and up to 50% of patients develop brain metastases (BMs). Detection of BMs on MRI is challenging with an inherent risk of missed diagnosis.Purpose: To train and evaluate a deep learning model (DLM) for fully automated detection and 3D segmentation of BMs in NSCLC on clinical routine MRI.Study Type: Retrospective.Population: Ninety‐eight NSCLC patients with 315 BMs on pretreatment MRI, divided into training (66 patients, 248 BMs) and independent test (17 patients, 67 BMs) and control (15 patients, 0 BMs) cohorts.Field Strength/Sequence: T1‐/T2‐weighted, T1‐weighted contrast‐enhanced (T1CE; gradient‐echo and spin‐echo sequences), and FLAIR at 1.0, 1.5, and 3.0 T from various vendors and study centers.Assessment: A 3D convolutional neural network (DeepMedic) was trained on the training cohort using 5‐fold cross‐validation and evaluated on the independent test and control sets. Three‐dimensional voxel‐wise manual segmentations of BMs by a neurosurgeon and a radiologist on T1CE served as the reference standard.Statistical Tests: Sensitivity (recall) and false positive (FP) findings per scan, dice similarity coefficient (DSC) to compare the spatial overlap between manual and automated segmentations, Pearson's correlation coefficient (r) to evaluate the relationship between quantitative volumetric measurements of segmentations, and Wilcoxon rank‐sum test to compare the volumes of BMs. A P value <0.05 was considered statistically significant.Results: In the test set, the DLM detected 57 of the 67 BMs (mean volume: 0.99 ± 4.24 cm3), resulting in a sensitivity of 85.1%, while FP findings of 1.5 per scan were observed. Missed BMs had a significantly smaller volume (0.05 ± 0.04 cm3) than detected BMs (0.96 ± 2.4 cm3). Compared with the reference standard, automated segmentations achieved a median DSC of 0.72 and a good volumetric correlation (r = 0.95). In the control set, 1.8 FPs/scan were observed.Data Conclusion: Deep learning provided a high detection sensitivity and good segmentation performance for BMs in NSCLC on heterogeneous scanner data while yielding a low number of FP findings.Level of Evidence: 3Technical Efficacy Stage: 2

4 February 2021 in American Journal of Neuroradiology
Verified
Journal Article
Closed access
PDF file
LPLenhard Pennig
Lenhard Pennig
RSR. Shahzad
R. Shahzad
LCLiliana Caldeira
Liliana Caldeira
SLSimon Lennartz
Simon Lennartz
FTFrank Thiele
Frank Thiele
KLK.R. Laukamp
K.R. Laukamp
BACKGROUND AND PURPOSE: Malignant melanoma is an aggressive skin cancer in which brain metastases are common. Our aim was to establish and evaluate a deep learning model for fully automated detection and segmentation of brain metastases in patients with malignant melanoma using clinical routine MR imaging. MATERIALS AND METHODS: Sixty-nine patients with melanoma with a total of 135 brain metastases at initial diagnosis and available multiparametric MR imaging datasets (T1-/T2-weighted, T1-weighted gadolinium contrast-enhanced, FLAIR) were included. A previously established deep learning model architecture (3D convolutional neural network; DeepMedic) simultaneously operating on the aforementioned MR images was trained on a cohort of 55 patients with 103 metastases using 5-fold cross-validation. The efficacy of the deep learning model was evaluated using an independent test set consisting of 14 patients with 32 metastases. Manual segmentations of metastases in a voxelwise manner (T1-weighted gadolinium contrast-enhanced imaging) performed by 2 radiologists in consensus served as the ground truth. RESULTS: After training, the deep learning model detected 28 of 32 brain metastases (mean volume, 1.0 [SD, 2.4] cm3) in the test cohort correctly (sensitivity of 88%), while false-positive findings of 0.71 per scan were observed. Compared with the ground truth, automated segmentations achieved a median Dice similarity coefficient of 0.75. CONCLUSIONS: Deep learning–based automated detection and segmentation of brain metastases in malignant melanoma yields high detection and segmentation accuracy with false-positive findings of <1 per scan.

23 January 2020 in Journal of Cardiovascular Magnetic Resonance
Verified
Journal Article
Open access
PDF file
LPLenhard Pennig
Lenhard Pennig
AWAnton Wagner
Anton Wagner
KWKilian Weiss
Kilian Weiss
SLSimon Lennartz
Simon Lennartz
JGJan-Peter Grunz
Jan-Peter Grunz
JDJonas Doerner
Jonas Doerner
Patients with Congenital heart disease (CHD) require repetitive imaging of the pulmonary vasculature throughout their life. In this study, we compared a novel Compressed SENSE accelerated (factor 9) electrocardiogram (ECG)- and respiratory-triggered 3D modified Relaxation-Enhanced Angiography without Contrast and Triggering (modified REACT-non-contrast-enhanced magnetic resonance angiography (modified REACT-non-CE-MRA)) with standard non-ECG-triggered time-resolved 4D CE-MRA for imaging of the pulmonary arteries and veins in patients with CHD. This retrospective analysis of 25 patients (June 2018–April 2019) with known or suspected CHD was independently conducted by two radiologists executing measurements on modified REACT-non-CE-MRA and 4D CE-MRA on seven dedicated points (inner edge): Main pulmonary artery (MPA), right and left pulmonary artery, right superior and inferior pulmonary vein, left superior (LSPV) and inferior pulmonary vein. Image quality for arteries and veins was evaluated on a four-point scale in consensus. Twenty-three of the 25 included patients presented a CHD. There was a high interobserver agreement for both methods of imaging at the pulmonary arteries (ICC ≥ 0.96); at the pulmonary veins, modified REACT-non-CE-MRA showed a slightly higher agreement, pronounced at LSPV (ICC 0.946 vs. 0.895). Measurements in 4D CE-MRA showed higher diameter values compared to modified REACT-non-CE-MRA, at the pulmonary arteries reaching significant difference (e.g. MPA: mean 0.408 mm, p = 0.002). Modified REACT-non-CE-MRA (average acquisition time 07:01 ± 02:44 min) showed significant better image quality than 4D CE-MRA at the pulmonary arteries (3.84 vs. 3.32, p < 0.001) and veins (3.32 vs. 2.72, p = 0.015). Compressed SENSE accelerated (factor 9) ECG- and respiratory-triggered 3D modified REACT-non-CE-MRA allows for reliable and fast imaging of the pulmonary arteries and veins with higher image quality and slightly higher interobserver agreement than 4D CE-MRA without contrast agent and associated disadvantages. Therefore, it represents a clinically suitable technique for patients requiring repetitive imaging of the pulmonary vasculature, e.g. patients with CHD.

13 July 2020 in Journal of Magnetic Resonance Imaging
Verified
Journal Article
Open access
PDF file
LPLenhard Pennig
Lenhard Pennig
UHUlrike Cornelia Isabel Hoyer
Ulrike Cornelia Isabel Hoyer
LGLukas Goertz
Lukas Goertz
RSRahil Shahzad
Rahil Shahzad
TPThorsten Persigehl
Thorsten Persigehl
KLKai Roman Laukamp
Kai Roman Laukamp
Background: Precise volumetric assessment of brain tumors is relevant for treatment planning and monitoring. However, manual segmentations are time‐consuming and impeded by intra‐ and interrater variabilities.Purpose: To investigate the performance of a deep‐learning model (DLM) to automatically detect and segment primary central nervous system lymphoma (PCNSL) on clinical MRI.Study Type: Retrospective.Population: Sixty‐nine scans (at initial and/or follow‐up imaging) from 43 patients with PCNSL referred for clinical MRI tumor assessment.Field Strength/Sequence: T1−/T2‐weighted, T1‐weighted contrast‐enhanced (T1 CE), and FLAIR at 1.0, 1.5, and 3.0T from different vendors and study centers.Assessment: Fully automated voxelwise segmentation of tumor components was performed using a 3D convolutional neural network (DeepMedic) trained on gliomas (n = 220). DLM segmentations were compared to manual segmentations performed in a 3D voxelwise manner by two readers (radiologist and neurosurgeon; consensus reading) from T1 CE and FLAIR, which served as the reference standard.Statistical Tests: Dice similarity coefficient (DSC) for comparison of spatial overlap with the reference standard, Pearson's correlation coefficient (r) to assess the relationship between volumetric measurements of segmentations, and Wilcoxon rank‐sum test for comparison of DSCs obtained in initial and follow‐up imaging.Results: The DLM detected 66 of 69 PCNSL, representing a sensitivity of 95.7%. Compared to the reference standard, DLM achieved good spatial overlap for total tumor volume (TTV, union of tumor volume in T1 CE and FLAIR; average size 77.16 ± 62.4 cm3, median DSC: 0.76) and tumor core (contrast enhancing tumor in T1 CE; average size: 11.67 ± 13.88 cm3, median DSC: 0.73). High volumetric correlation between automated and manual segmentations was observed (TTV: r = 0.88, P < 0.0001; core: r = 0.86, P < 0.0001). Performance of automated segmentations was comparable between pretreatment and follow‐up scans without significant differences (TTV: P = 0.242, core: P = 0.177).Data Conclusion: In clinical MRI scans, a DLM initially trained on gliomas provides segmentation of PCNSL comparable to manual segmentation, despite its complex and multifaceted appearance. Segmentation performance was high in both initial and follow‐up scans, suggesting its potential for application in longitudinal tumor imaging.Level of Evidence: 3Technical Efficacy Stage: 2

27 August 2020 in The International Journal of Cardiovascular Imaging
Verified
Journal Article
Open access
PDF file
LPLenhard Pennig
Lenhard Pennig
AWAnton Wagner
Anton Wagner
KWKilian Weiss
Kilian Weiss
SLSimon Lennartz
Simon Lennartz
MHMichael Huntgeburth
Michael Huntgeburth
JDJonas Doerner
Jonas Doerner
To compare a novel Compressed SENSE accelerated ECG- and respiratory-triggered flow-independent 3D isotropic Relaxation-Enhanced Angiography without Contrast and Triggering (modified REACT) with standard non-ECG-triggered 3D contrast-enhanced magnetic resonance angiography (CE-MRA) for imaging of the thoracic aorta in patients with connective tissue diseases (CTD) or other aortic diseases using manual and semiautomatic measurement approaches. This retrospective, single-center analysis of 30 patients (June–December 2018) was conducted by two radiologists, who independently measured aortic diameters on modified REACT and CE-MRA using manual (Multiplanar-Reconstruction) and semiautomatic (Advanced Vessel Analysis) measurement tools on seven levels (inner edge): Aortic annulus and sinus, sinotubular junction, mid- and high-ascending aorta, aortic isthmus, and descending aorta. Bland–Altman analysis was conducted to evaluate differences between the mean values of aortic width and ICCs were calculated to assess interobserver agreement. For each level, image quality was evaluated on a four-point scale in consensus with Wilcoxon matched-pair test used to evaluate for differences between both MRA techniques. Additionally, evaluation time for each measurement technique was noted, which was compared applying one-way ANOVA. When comparing both imaging and measurement methods, CE-MRA (mean difference 0.24 ± 0.27 mm) and the AVA-tool (− 0.21 ± 0.15 mm) yielded higher differences compared to modified REACT (− 0.11 ± 0.11 mm) and the MPR-tool (0.07 ± 0.21 mm) for all measurement levels combined without yielding clinical significance. There was an excellent interobserver agreement between modified REACT and CE-MRA using both tools of measurement (ICC > 0.9). Modified REACT (average acquisition time 06:34 ± 01:36 min) provided better image quality from aortic annulus to mid-ascending aorta (p < 0.05), whereas at distal measurement levels, no significant differences were noted. Regarding time requirement, no statistical significance was found between both measurement techniques (p = 0.08). As a novel non-CE-MRA technique, modified REACT allows for fast imaging of the thoracic aorta with higher image quality in the proximal aorta than CE-MRA enabling a reliable measurement of vessel dimensions without the need for contrast agent. Thus, it represents a clinically suitable alternative for patients requiring repetitive imaging. Manual and semiautomatic measurement approaches provided comparable results without significant difference in time need.