Hermann Baumgartl

Hermann Baumgartl

Auto-generated profile, based on publicly available information

Hermann Baumgartl

Department of Business Information Systems, Aalen University, Aalen, Germany

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
18 February 2020 in Progress in Additive Manufacturing
Verified
Journal Article
Open access
PDF file
HBHermann Baumgartl
Hermann Baumgartl
JTJosef Tomas
Josef Tomas
RBRicardo Buettner
Ricardo Buettner
MMMarkus Merkel
Markus Merkel
Introduction: There is neither strong evidence on effective treatments for patients with chronic back pain (CBP) and depressive disorder nor sufficiently available mental health care offers. Objective: The aim is to assess the effectiveness of internet- and mobile-based interventions (IMI) as a scalable approach for treating depression in a routine care setting. Methods: This is an observer-masked, multicenter, pragmatic randomized controlled trial with a randomization ratio of 1:1.Patients with CBP and diagnosed depressive disorder (mild to moderate severity) were recruited from 82 orthopedic rehabilitation clinics across Germany. The intervention group (IG) received a guided depression IMI tailored to CBP next to treatment-as-usual (TAU; including medication), while the control group (CG) received TAU. The primary outcome was observer-masked clinician-rated Hamilton depression severity (9-week follow-up). The secondary outcomes were: further depression outcomes, pain-related outcomes, health-related quality of life, and work capacity. Biostatistician blinded analyses using regression models were conducted by intention-to-treat and per protocol analysis. Results: Between October 2015 and July 2017, we randomly assigned 210 participants (IG, n = 105; CG, n = 105), mostly with only a mild pain intensity but substantial pain disability. No statistically significant difference in depression severity between IG and CG was observed at the 9-week follow-up (β = –0.19, 95% CI –0.43 to 0.05). Explorative secondary depression (4/9) and pain-related (4/6) outcomes were in part significant (p < 0.05). Health-related quality of life was significantly higher in the IG. No differences were found in work capacity. Conclusion: The results indicate that an IMI for patients with CBP and depression in a routine care setting has limited impact on depression. Benefits in pain and health-related outcomes suggest that an IMI might still be a useful measure to improve routine care.

17 December 2020 in PLOS ONE
Verified
Journal Article
Open access
PDF file
TRThilo Rieg
Thilo Rieg
JFJanek Frick
Janek Frick
HBHermann Baumgartl
Hermann Baumgartl
RBRicardo Buettner
Ricardo Buettner
We present the results from a white-box machine learning approach to detect cardiac arrhythmias using electrocardiographic data. A C5.0 is trained to recognize four classes using common features. The four classes are (i) atrial fibrillation and atrial flutter, (ii) tachycardias (iii), sinus bradycardia and (iv) sinus rhythm. Data from 10,646 subjects, 83% of whom have at least one arrhythmia and 17% of whom exhibit a normal sinus rhythm, are used. The C5.0 is trained using 10-fold cross-validation and is able to achieve a balanced accuracy of 95.35%. By using the white-box machine learning approach, a clear and comprehensible tree structure can be revealed, which has selected the 5 most important features from a total of 24 features. These 5 features are ventricular rate, RR-Interval variation, atrial rate, age and difference between longest and shortest RR-Interval. The combination of ventricular rate, RR-Interval variation and atrial rate is especially relevant to achieve classification accuracy, which can be disclosed through the tree. The tree assigns unique values to distinguish the classes. These findings could be applied in medicine in the future. It can be shown that a white-box machine learning approach can reveal granular structures, thus confirming known linear relationships and also revealing nonlinear relationships. To highlight the strength of the C5.0 with respect to this structural revelation, the results of further white-box machine learning and black-box machine learning algorithms are presented.

1 January 2019
Verified
Conference Paper
Open access
PDF file
RBRicardo Buettner
Ricardo Buettner
HBHermann Baumgartl
Hermann Baumgartl
Using convolutional neural networks we extend the work by Dugdale's group on socially relevant multi-agent systems in crisis and emergency situations by giving the artificial agent the ability to precisely recognize escape signs, doors and stairs for escape route planning. We build an efficient recognition module consisting of three blocks of a depth-wise separable convolutional layer, a max-pooling layer, and a batchnormalization layer before dense, dropout and classifying the image. A rigorous evaluation based on the MCIndoor20000 dataset shows excellent performance values (e.g. over 99.81 percent accuracy). In addition, our module architecture is 78 times smaller than the MCIndoor20000 benchmark-making it suitable for embedding in operational drones and robots.

1 February 2022 in IEEE Access
Verified
Journal Article
Open access
JGJan Gross
Jan Gross
RBRicardo Buettner
Ricardo Buettner
HBHermann Baumgartl
Hermann Baumgartl
With the growing availability and complexity of time-series sequences, scalable and robust machine learning approaches are required that overcome the sampling challenge of quantitatively sufficient training data. Following the research trend towards the deep learning-based analysis of time-series encoded as images, this study proposes a time-series imaging workflow that overcomes the challenge of quantitatively limited sensor data across domains (i.e., medicine and engineering). After systematically identifying the three relevant dimensions that affect the performance of the deep learning-based analysis of visualized time-series data, we performed a benchmarking evaluation with a total of 24 unique convolutional neural network models. Following a two-level transfer learning investigation, we reveal that fine-tuning the mid-level features results in the best classification performance. As a result, we present an optimized representation of the VGG16 network, which outperforms previous studies in the field. Our approach is accurate, robust, and manifests internal and external validity. By only using the raw time-series data, our model does not require manual feature engineering, being of high practical relevance. As the post-hoc analysis of our results reveals that our model allows automated extraction of meaningful features based on the trend of the underlying time-series data, our study also adds to explainable artificial intelligence. Furthermore, our proposed workflow reduces the sequence length of the input data while preserving all information. Especially with the hurdle of long-term dependencies in sequential time-series data, we overcome related work’s limitation of the vanishing gradients problem and contribute to the sequential learning theory in artificial intelligence.

27 October 2020
Verified
Book Chapter
Closed access
PDF file
HBHermann Baumgartl
Hermann Baumgartl
SBSamuel Bayerlein
Samuel Bayerlein
RBRicardo Buettner
Ricardo Buettner