Lukas Zimmermann

Lukas Zimmermann

Auto-generated profile, based on publicly available information

Lukas Zimmermann

Medical University of Vienna

Research Focus & Metrics

Academic Career Length

7Years (2020-Present)

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

Representative Publications

View All Publications
21 May 2020 in Physics in Medicine & Biology
Verified
Journal Article
Open access
PDF file
LFLukas Fetty
Lukas Fetty
TLTommy Löfstedt
Tommy Löfstedt
GHGerd Heilemann
Gerd Heilemann
HFHugo Furtado
Hugo Furtado
NNNicole Nesvacil
Nicole Nesvacil
PKPeter Kuess
Peter Kuess
Recent developments in magnetic resonance (MR) to synthetic computed tomography (sCT) conversion have shown that treatment planning is possible without an initial planning CT. Promising conversion results have been demonstrated recently using conditional generative adversarial networks (cGANs). However, the performance is generally only tested on images from one MR scanner, which neglects the potential of neural networks to find general high-level abstract features. In this study, we explored the generalizability of the generator models, trained on a single field strength scanner, to data acquired with higher field strengths. T2-weighted 0.35T MRIs and CTs from 51 patients treated for prostate (40) and cervical cancer (11) were included. 25 of them were used to train four different generators (SE-ResNet, DenseNet, U-Net, and Embedded Net). Further, an ensemble model was created from the four network outputs. The models were validated on 16 patients from a 0.35T MR scanner. Further, the trained models were tested on the Gold Atlas dataset, containing T2-weighted MR scans of different field strengths; 1.5T(7) and 3T(12), and 10 patients from the 0.35T scanner. The sCTs were dosimetrically compared using clinical VMAT plans for all test patients. For the same scanner (0.35T), the results from the different models were comparable on the test set, with only minor differences in the mean absolute error (MAE) (35-51HU body). Similar results were obtained for conversions of 3T GE Signa and the 3T GE Discovery images (40-62HU MAE) for three of the models. However, larger differences were observed for the 1.5T images (48-65HU MAE). The overall best model was found to be the ensemble model. All dose differences were below 1%. This study shows that it is possible to generalize models trained on images of one scanner to other scanners and different field strengths. The best metric results were achieved by the combination of all networks.

12 September 2022 in Physics in Medicine & Biology
Verified
Journal Article
Open access
ABAaron Babier
Aaron Babier
RMRafid Mahmood
Rafid Mahmood
BZBinghao Zhang
Binghao Zhang
VAVictor G L Alves
Victor G L Alves
AMAna Maria Barragan Montero
Ana Maria Barragan Montero
TCTimothy C Y Chan
Timothy C Y Chan
Objective. To establish an open framework for developing plan optimization models for knowledge-based planning (KBP). Approach. Our framework includes radiotherapy treatment data (i.e. reference plans) for 100 patients with head-and-neck cancer who were treated with intensity-modulated radiotherapy. That data also includes high-quality dose predictions from 19 KBP models that were developed by different research groups using out-of-sample data during the OpenKBP Grand Challenge. The dose predictions were input to four fluence-based dose mimicking models to form 76 unique KBP pipelines that generated 7600 plans (76 pipelines × 100 patients). The predictions and KBP-generated plans were compared to the reference plans via: the dose score, which is the average mean absolute voxel-by-voxel difference in dose; the deviation in dose-volume histogram (DVH) points; and the frequency of clinical planning criteria satisfaction. We also performed a theoretical investigation to justify our dose mimicking models. Main results. The range in rank order correlation of the dose score between predictions and their KBP pipelines was 0.50–0.62, which indicates that the quality of the predictions was generally positively correlated with the quality of the plans. Additionally, compared to the input predictions, the KBP-generated plans performed significantly better ( P < 0.05; one-sided Wilcoxon test) on 18 of 23 DVH points. Similarly, each optimization model generated plans that satisfied a higher percentage of criteria than the reference plans, which satisfied 3.5% more criteria than the set of all dose predictions. Lastly, our theoretical investigation demonstrated that the dose mimicking models generated plans that are also optimal for an inverse planning model. Significance. This was the largest international effort to date for evaluating the combination of KBP prediction and optimization models. We found that the best performing models significantly outperformed the reference dose and dose predictions. In the interest of reproducibility, our data and code is freely available.

14 June 2023 in Medical Physics
Verified
Journal Article
Open access
Gerd HeilemannGerd Heilemann
Gerd Heilemann
LZLukas Zimmermann
Lukas Zimmermann
RSRaphael Schotola
Raphael Schotola
WLWolfgang Lechner
Wolfgang Lechner
MPMarco Peer
Marco Peer
PKPeter Kuess
Peter Kuess
Background: Deep learning‐based auto‐planning is an active research field; however, for some tasks a treatment planning system (TPS) is still required.Purpose: To introduce a deep learning‐based model generating deliverable DICOM RT treatment plans that can be directly irradiated by a linear accelerator (LINAC). The model was based on an encoder‐decoder network and can predict multileaf collimator (MLC) motion sequences for prostate VMAT radiotherapy.Methods: A total of 619 treatment plans from 460 patients treated for prostate cancer with single‐arc VMAT were included in this study. An encoder‐decoder network was trained using 465 clinical treatment plans and validated on 77 plans. The performance was analyzed on a separate test set of 77 treatment plans. Separate L1 losses were computed for the leaf and jaw positions as well as the monitor units, with the leaf loss being weighted by a factor of 100 before being added to the other losses. The generated treatment plans were recalculated in a treatment planning system and the dose‐volume metrics and gamma passing rates were compared to the original dose.Results: All generated treatment plans showed good agreement with the original data, with an average gamma passing rate (3%/3 mm) of 91.9 ± 7.1%. However, the coverage of the PTVs. was slightly lower for the generated plans (D98% = 92.9 ± 2.6%) in comparison to the original plans (D98% = 95.7 ± 2.2%). There was no significant difference in mean dose to the bladder between the predicted and original plan (Dmean of 28.0 ± 13.5 vs. 28.1 ± 13.3% of prescribed dose) or rectum (Dmean of 42.3 ± 7.4 vs. 42.6 ± 7.5%). The maximum dose to bladder was only slightly higher in the predicted plans (D2% of 100.7 ± 5.3 vs. 99.8 ± 4.0%) and for the rectum it was even lower (D2% of 100.5 ± 3.7 vs. 100.1 ± 4.3).Conclusions: The deep learning‐based model could predict MLC motion sequences in prostate VMAT plans, eliminating the need for sequencing inside a TPS, thus revolutionizing autonomous treatment planning workflows. This research completes the loop in deep learning‐based treatment planning processes, enabling more efficient workflows for real‐time or online adaptive radiotherapy.

1 May 2022 in Zeitschrift für Medizinische Physik
Verified
Journal Article
Open access
LZLukas Zimmermann
Lukas Zimmermann
BKBarbara Knäusl
Barbara Knäusl
MSMarkus Stock
Markus Stock
CLCarola Lütgendorf-Caucig
Carola Lütgendorf-Caucig
Dietmar GeorgDietmar Georg
Dietmar Georg
PKPeter Kuess
Peter Kuess
A magnetic resonance imaging (MRI) sequence independent deep learning technique was developed and validated to generate synthetic computed tomography (sCT) scans for MR guided proton therapy. 47 meningioma patients previously undergoing proton therapy based on pencil beam scanning were divided into training (33), validation (6), and test (8) cohorts. T1, T2, and contrast enhanced T1 (T1CM) MRI sequences were used in combination with the planning CT (pCT) data to train a 3D U-Net architecture with ResNet-Blocks. A hyperparameter search was performed including two loss functions, two group sizes of normalisation, and depth of the network. Training outcome was compared between models trained for each individual MRI sequence and for all sequences combined. The performance was evaluated based on a metric and dosimetric analysis as well as spot difference maps. Furthermore, the influence of immobilisation masks that are not visible on MRIs was investigated. Based on the hyperparameter search, the final model was trained with fixed features per group for the group normalisation, six down-convolution steps, an input size of 128 × 192 × 192, and feature loss. For the test dataset for body/bone the mean absolute error (MAE) values were on average 79.8/216.3 Houndsfield unit (HU) when trained using T1 images, 71.1/186.1 HU for T2, and 82.9/236.4 HU for T1CM. The structural similarity metric (SSIM) ranged from 0.95 to 0.98 for all sequences. The investigated dose parameters of the target structures agreed within 1% between original proton treatment plans and plans recalculated on sCTs. The spot difference maps had peaks at ±0.2 cm and for 98% of all spots the difference was less than 1 cm. A novel MRI sequence independent sCT generator was developed, which suggests that the training phase of neural networks can be disengaged from specific MRI acquisition protocols. In contrast to previous studies, the patient cohort consisted exclusively of actual proton therapy patients (i.e. “real-world data”).

1 August 2022 in Zeitschrift für Medizinische Physik
Verified
Journal Article
Open access
Gerd HeilemannGerd Heilemann
Gerd Heilemann
MMMark Matthewman
Mark Matthewman
PKPeter Kuess
Peter Kuess
GGGregor Goldner
Gregor Goldner
JWJoachim Widder
Joachim Widder
LZLukas Zimmermann
Lukas Zimmermann
For image translational tasks, the application of deep learning methods showed that Generative Adversarial Network (GAN) architectures outperform the traditional U-Net networks, when using the same training data size. This study investigates whether this performance boost can also be expected for segmentation tasks with small training dataset size. Two models were trained on varying training dataset sizes ranging from 1—100 patients: a) U-Net and b) U-Net with patch discriminator (conditional GAN). The performance of both models to segment the male pelvis on CT-data was evaluated (Dice similarity coefficient, Hausdorff) with respect to training data size. No significant differences were observed between the U-Net and cGAN when the models were trained with the same training sizes up to 100 patients. The training dataset size had a significant impact on the models’ performances, with vast improvements when increasing dataset sizes from 1 to 20 patients. When introducing GANs for the segmentation task no significant performance boost was observed in our experiments, even in segmentation models developed on small datasets.

25 November 2025 in Radiotherapy and Oncology
Verified
Journal Article
Open access
LZLukas Zimmermann
Lukas Zimmermann
BKBarbara Knäusl
Barbara Knäusl
JKJohannes Knoth
Johannes Knoth
ASAlina E. Sturdza
Alina E. Sturdza
VDVincent Dick
Vincent Dick
MSMaximilian Schmid
Maximilian Schmid
Accurate gross tumor volume (GTV) delineation is critical for successful radiochemotherapy and image-guided adaptive brachytherapy (BT) in cervical cancer. This study investigated whether diffusion-weighted imaging (DWI) improves GTV delineation accuracy compared to T2-weighted (T2w) MRI alone, across different physician experience levels. Twenty-seven patients with locally advanced cervical carcinoma undergoing primary radiochemotherapy were analyzed. Six physicians (three experts, three residents) delineated GTVs at three time points: diagnosis (init), pre-brachytherapy (preBT), and pre-brachytherapy with applicator in situ (BT). Segmentations were performed using T2w images alone and T2w plus DWI (b=800smm-2) guidance. Expert consensus served as reference standard using STAPLE algorithm. Inter-observer agreement was assessed using conformity index, Dice-Sørensen coefficient, and Hausdorff distance. DWI guidance significantly improved inter-observer agreement among experts at init (conformity index: 0.62→ 0.70, p<0.05) and BT (0.33→ 0.39, p<0.05) time points. For residents, DWI guidance enhanced agreement with expert consensus, particularly during BT, with significant improvements in Dice coefficient (median increase 9%, p<0.05) and reduced Hausdorff distance (median decrease 1.3 mm, p<0.05). Tumor volume correlation between preBT and BT time points improved with DWI guidance for both groups. Incorporating DWI into the segmentation workflow reduces inter-observer variability for both expert and resident radiation oncologists. DWI guidance particularly benefits less experienced physicians, enabling them to achieve contours closer to expert consensus standards through additional functional information.