Page Not Found
Page not found. Your pixels are in another canvas.
Sitemap
A list of all the posts and pages found on the site. For you robots out there is an XML version available for digesting as well.
Pages
About me
About me
Page not in menu
This is a page not in th emain menu
<a href="https://mibaumgartner.github.io{"from"=>"/", "to"=>"https://mibaumgartner.github.io/highlights/"}" rel="permalink"></a>
<a href="https://mibaumgartner.github.io{"from"=>"/about/", "to"=>"https://mibaumgartner.github.io/highlights/"}" rel="permalink"></a>
<a href="https://mibaumgartner.github.io{"from"=>"/about.html", "to"=>"https://mibaumgartner.github.io/highlights/"}" rel="permalink"></a>
<a href="https://mibaumgartner.github.io{"from"=>"/md/", "to"=>"https://mibaumgartner.github.io/markdown/"}" rel="permalink"></a>
<a href="https://mibaumgartner.github.io{"from"=>"/markdown.html", "to"=>"https://mibaumgartner.github.io/markdown/"}" rel="permalink"></a>
<a href="https://mibaumgartner.github.io{"from"=>"/nmp/", "to"=>"https://mibaumgartner.github.io/non-menu-page/"}" rel="permalink"></a>
<a href="https://mibaumgartner.github.io{"from"=>"/nmp.html", "to"=>"https://mibaumgartner.github.io/non-menu-page/"}" rel="permalink"></a>
<a href="https://mibaumgartner.github.io{"from"=>"/wordpress/blog-posts/", "to"=>"https://mibaumgartner.github.io/year-archive/"}" rel="permalink"></a>
Posts
Future Blog Post
Published:
This post will show up by default. To disable scheduling of future posts, edit config.yml and set future: false.
Blog Post number 4
Published:
This is a sample blog post. Lorem ipsum I can’t remember the rest of lorem ipsum and don’t have an internet connection right now. Testing testing testing this blog post. Blog posts are cool.
Blog Post number 3
Published:
This is a sample blog post. Lorem ipsum I can’t remember the rest of lorem ipsum and don’t have an internet connection right now. Testing testing testing this blog post. Blog posts are cool.
Blog Post number 2
Published:
This is a sample blog post. Lorem ipsum I can’t remember the rest of lorem ipsum and don’t have an internet connection right now. Testing testing testing this blog post. Blog posts are cool.
Blog Post number 1
Published:
This is a sample blog post. Lorem ipsum I can’t remember the rest of lorem ipsum and don’t have an internet connection right now. Testing testing testing this blog post. Blog posts are cool.
highlights
projects
Delira - A Backend Agnostic High Level Deep Learning Library
Lightweight framework for fast prototyping and training of deep neural networks with PyTorch and TensorFlow. delira is designed to work as a backend agnostic high level deep learning library which allwos the user to compare various models written in different backends without rewriting them.
Rising: High-Performance Data Loading and Augmentation for 2D & 3D Images
Rising is a high-performance data loading and augmentation library for 2D and 3D data completely written in PyTorch. Our goal is to provide a seamless integration into the PyTorch Ecosystem without sacrificing usability or features. All transformations are directly implemented in PyTorch which allows gradient propagation and direct execution on GPUs. The provided dataloading module allows for various types of transformations (CPU vs GPU, per sample vs batched) to maximise the resuability of all implemented components.
nnDetection: A Self-configuring Method for Medical Object Detection
Simultaneous localisation and categorization of objects in medical images, also referred to as medical object detection, is of high clinical relevance because diagnostic decisions depend on rating of objects rather than e.g. pixels. For this task, the cumbersome and iterative process of method configuration constitutes a major research bottleneck. Recently, nnU-Net has tackled this challenge for the task of image segmentation with great success. Following nnU-Net’s agenda, in this work we systematize and automate the configuration process for medical object detection. The resulting self-configuring method, nnDetection, adapts itself without any manual intervention to arbitrary medical detection problems while achieving results en par with or superior to the state-of-the-art. We demonstrate the effectiveness of nnDetection on two public benchmarks, ADAM and LUNA16, and propose 10 further public data sets for a comprehensive evaluation of medical object detection methods.
publications
Multi Scale Curriculum CNN for Context-Aware Breast MRI Malignancy Classification
Published in Medical Image Computing and Computer Assisted Intervention–MICCAI 2019, 2019
Citation: Haarburger, Christoph, et al. "Multi scale curriculum CNN for context-aware breast MRI malignancy classification." Medical Image Computing and Computer Assisted Intervention–MICCAI 2019: 22nd International Conference, Shenzhen, China, October 13–17, 2019, Proceedings, Part IV 22. Springer International Publishing, 2019.
Classification of malignancy for breast cancer and other cancer types is usually tackled as an object detection problem: Individual lesions are first localized and then classified with respect to malignancy. However, the drawback of this approach is that abstract features incorporating several lesions and areas that are not labelled as a lesion but contain global medically relevant information are thus disregarded: especially for dynamic contrast-enhanced breast MRI, criteria such as background parenchymal enhancement and location within the breast are important for diagnosis and cannot be captured by object detection approaches properly.
nnDetection: A Self-configuring Method for Medical Object Detection
Published in Medical Image Computing and Computer Assisted Intervention–MICCAI 2021, 2021
Citation: Baumgartner, Michael and Jaeger Paul, et al. "nnDetection: a self-configuring method for medical object detection." Medical Image Computing and Computer Assisted Intervention–MICCAI 2021: 24th International Conference, Strasbourg, France, September 27–October 1, 2021, Proceedings, Part V 24. Springer International Publishing, 2021.
Simultaneous localisation and categorization of objects in medical images, also referred to as medical object detection, is of high clinical relevance because diagnostic decisions often depend on rating of objects rather than e.g. pixels. For this task, the cumbersome and iterative process of method configuration constitutes a major research bottleneck. Recently, nnU-Net has tackled this challenge for the task of image segmentation with great success. Following nnU-Net’s agenda, in this work we systematize and automate the configuration process for medical object detection. The resulting self-configuring method, nnDetection, adapts itself without any manual intervention to arbitrary medical detection problems while achieving results en par with or superior to the state-of-the-art. We demonstrate the effectiveness of nnDetection on two public benchmarks, ADAM and LUNA16, and propose 11 further medical object detection tasks on public data sets for comprehensive method evaluation. Code is at this URL.
Accurate Detection of Mediastinal Lesions with nnDetection
Published in Lesion Segmentation in Surgical and Diagnostic Applications. CuRIOUS KiPA MELA 2022, 2023
Citation: Baumgartner, M., et al. (2023). Accurate Detection of Mediastinal Lesions with nnDetection. In: Xiao, Y., Yang, G., Song, S. (eds) Lesion Segmentation in Surgical and Diagnostic Applications. CuRIOUS KiPA MELA 2022 2022 2022. Lecture Notes in Computer Science, vol 13648. Springer, Cham.
The accurate detection of mediastinal lesions is one of the rarely explored medical object detection problems. In this work, we applied a modified version of the self-configuring method nnDetection to the Mediastinal Lesion Analysis (MELA) Challenge 2022. By incorporating automatically generated pseudo masks, training high capacity models with large patch sizes in a multi GPU setup and an adapted augmentation scheme to reduce localization errors caused by rotations, our method achieved an excellent FROC score of 0.9922 at IoU 0.10 and 0.9880 at IoU 0.3 in our cross-validation experiments. The submitted ensemble ranked third in the competition with a FROC score of 0.9897 on the MELA challenge leaderboard.
Taming Detection Transformers for Medical Object Detection
Published in BVM Workshop 2023, 2023
Citation: Ickler, Marc K. and Baumgartner, Michael, et al. "Taming Detection Transformers for Medical Object Detection." BVM Workshop. Wiesbaden: Springer Fachmedien Wiesbaden, 2023.
The accurate detection of suspicious regions in medical images is an error-prone and time-consuming process required by many routinely performed diagnostic procedures. To support clinicians during this difficult task, several automated solutions were proposed relying on complex methods with many hyperparameters. In this study, we investigate the feasibility of DEtection TRansformer (DETR) models for volumetric medical object detection. In contrast to previous works, these models directly predict a set of objects without relying on the design of anchors or manual heuristics such as non-maximum-suppression to detect objects. We show by conducting extensive experiments with three models, namely DETR, Conditional DETR, and DINO DETR on four data sets (CADA, RibFrac, KiTS19, and LIDC) that these set prediction models can perform on par with or even better than currently existing methods. DINO DETR, the best-performing model in our experiments demonstrates this by outperforming a strong anchor-based one-stage detector, Retina U-Net, on three out of four data sets.
Deep-learning based detection of vessel occlusions on CT-angiography in patients with suspected acute ischemic stroke
Published in Nature Communications, 2023
Citation: Brugnara, Gianluca and Baumgartner, Michael and Scholze, Edwin David, et al. "Deep-learning based detection of vessel occlusions on CT-angiography in patients with suspected acute ischemic stroke." Nature Communications 14.1 (2023): 4938.
Swift diagnosis and treatment play a decisive role in the clinical outcome of patients with acute ischemic stroke (AIS), and computer-aided diagnosis (CAD) systems can accelerate the underlying diagnostic processes. Here, we developed an artificial neural network (ANN) which allows automated detection of abnormal vessel findings without any a-priori restrictions and in <2 minutes. Pseudo-prospective external validation was performed in consecutive patients with suspected AIS from 4 different hospitals during a 6-month timeframe and demonstrated high sensitivity (≥87%) and negative predictive value (≥93%). Benchmarking against two CE- and FDA-approved software solutions showed significantly higher performance for our ANN with improvements of 25–45% for sensitivity and 4–11% for NPV (p ≤ 0.003 each). We provide an imaging platform URL for online processing of medical imaging data with the developed ANN, including provisions for data crowdsourcing, which will allow continuous refinements and serve as a blueprint to build robust and generalizable AI algorithms.
talks
Retina U-Net for aneurysm detection in MR images
Published:
Presented our winning solution (Task 1: Detection Task) of the Aneurysm Detection And segMentation Challenge (2020) challenge. Our solution used a Retina U-Net architecture with a Feature Pyramid Network to detection aneurysms in the provided MR images. More info can be found in our short description.
Advanced Deep Learning Tutorial: Medical Object Detection
Published:
The remarkable rise of deep learning has led to an overwhelming amount of new papers coming up by the week. This tutorial intends to filter out the research most relevant for the medical image computing (MIC) community and present it in a structured and understandable form. It will cover recent developments related to common tasks in the community (e.g. segmentation, detection), but will also discuss methods that are currently gaining traction. Basic knowledge of neural networks and deep learning is recommended.
What’s Next: Technological Advancement for AI in Radiology On the horizon
Published:
Presented selected topics about recent technological advancements in the field of artificial intelligence in radiology. The following areas were presented:
- Transformers: Closing the Gap between Natural Language Processing (NLP) and Computer Vision (CV) in Medical Image Computing
- Semi-Supervised Learning: How to learn from unlabelled Data?
- Federated Learning: Big Data without Data Transfer
- Generalisation to new data sets via self-configuring methods
Self-configuring Methods for Deep Learning-based Biomedical Image Analysis
Published:
Presented our work on self-configuring methods for deep learning-based biomedical image segmentation (nnU-Net) and object detection (nnDetection).
nnDetection: A Self-configuring Method for Medical Object Detection (Abstract)
Published:
Presented our submitted abstract about nnDetection: A Self-configuring Method for Medical Object Detection. Presentation was awarded the Best Presentation Award.
Accurate Detection of Mediastinal Lesions with nnDetection
Published:
Presented our solution (3rd Rank) of the Mediastinal Lesion Detection Detection (2022) challenge. More info can be found in our short description.
Taming Detection Transformers for Medical Object Detection
Published:
Presented our work on DEtection TRansformers (DETR) for medical object detection. More info can be found in our publication. Our paper ranked 3rd for the Best Paper Award.