Oliver Amft’s research while affiliated with University of Freiburg and other places

What is this page?


This page lists works of an author who doesn't have a ResearchGate profile or hasn't added the works to their profile yet. It is automatically generated from public (personal) data to further our legitimate goal of comprehensive and accurate scientific recordkeeping. If you are this author and want this page removed, please let us know.

Publications (255)


Fig. 6 Effects of scene-aware re-prompting on segmentation accuracy (top) and variations in relative liver size for Patient 4 based on ground truth annotations (bottom)
Automated multimodel segmentation and tracking for AR-guided open liver surgery using scene-aware self-prompting
  • Article
  • Full-text available

May 2025

·

24 Reads

International Journal of Computer Assisted Radiology and Surgery

·

·

Konstantin Holzapfel

·

[...]

·

Oliver Amft

Purpose We introduce a multimodel, real-time semantic segmentation and tracking approach for Augmented Reality (AR)-guided open liver surgery. Our approach leverages foundation models and scene-aware re-prompting strategies to balance segmentation accuracy and inference time as required for real-time AR-assisted surgery applications. Methods Our approach integrates a domain-specific RGBD model (ESANet), a foundation model for semantic segmentation (SAM), and a semi-supervised video object segmentation model (DeAOT). Models were combined in an auto-promptable pipeline with a scene-aware re-prompting algorithm that adapts to surgical scene changes. We evaluated our approach on intraoperative RGBD videos from 10 open liver surgeries using a head-mounted AR device. Segmentation accuracy (IoU), temporal resolution (FPS), and the impact of re-prompting strategies were analyzed. Comparisons to individual models were performed. Results Our multimodel approach achieved a median IoU of 71% at 13.2 FPS without re-prompting. Performance of our multimodel approach surpasses that of individual models, yielding better segmentation accuracy than ESANet and better temporal resolution compared to SAM. Our scene-aware re-prompting method reaches the DeAOT performance, with an IoU of 74.7% at 11.5 FPS, even when the DeAOT model uses an ideal reference frame. Conclusion Our scene-aware re-prompting strategy provides a trade-off between segmentation accuracy and temporal resolution, thus addressing the requirements of real-time AR-guided open liver surgery. The integration of complementary models resulted in robust and accurate segmentation in a complex, real-world surgical settings.

Download

Automated inflammatory bowel disease detection using wearable bowel sound event spotting

March 2025

·

9 Reads

Introduction Inflammatory bowel disorders may result in abnormal Bowel Sound (BS) characteristics during auscultation. We employ pattern spotting to detect rare bowel BS events in continuous abdominal recordings using a smart T-shirt with embedded miniaturised microphones. Subsequently, we investigate the clinical relevance of BS spotting in a classification task to distinguish patients diagnosed with inflammatory bowel disease (IBD) and healthy controls. Methods Abdominal recordings were obtained from 24 patients with IBD with varying disease activity and 21 healthy controls across different digestive phases. In total, approximately 281 h of audio data were inspected by expert raters and thereof 136 h were manually annotated for BS events. A deep-learning-based audio pattern spotting algorithm was trained to retrieve BS events. Subsequently, features were extracted around detected BS events and a Gradient Boosting Classifier was trained to classify patients with IBD vs. healthy controls. We further explored classification window size, feature relevance, and the link between BS-based IBD classification performance and IBD activity. Results Stratified group K-fold cross-validation experiments yielded a mean area under the receiver operating characteristic curve ≥0.83 regardless of whether BS were manually annotated or detected by the BS spotting algorithm. Discussion Automated BS retrieval and our BS event classification approach have the potential to support diagnosis and treatment of patients with IBD.



Fig. 3: Exploration of DCT and NPI combinations regarding their impact on pandemic characteristics. Besides contact tracing, the following NPIs were considered: "Request negative test to exit" (RNT), "Quarantine Contact Person Household" (QCH), "Closure of Bars and Restaurants" (CBR). (a): DCT-optimal behaviour scenario (maximum for DCT adoption, adherence, and compliance). Baselines illustrate scenarios with and without DCT. Switching on and off RNT, QCH, and CBR creates a range of outcomes for pandemic characteristics. (b -d): Realistic behaviour scenario (DCT adoption: 50%, DCT adherence: 70%, DCT compliance: 70%). (b): DCT positively affects pandemic characteristics, even for realistic behaviour. (c): Exploration of RNT in addition to DCT. RNT does not influence pandemic characteristics. (d): Exploration of QCH and CBR in addition to DCT. Both QCH and CBR have a mutually exclusive effect on pandemic characteristics. (a -d): Shaded areas show the results range across active and inactive NPIs (denoted by "±"). All simulations include MCT and ICT. Incubation and clearance periods are not shown, as they are not
Terminology overview and key pathogen model concepts.
Impact of Digital Contact Tracing on Pandemic Control Analysed with Behaviour-driven Agent-based Modelling

January 2025

·

56 Reads

We disentangle the efficacy of individual non-pharmaceutical interventions (NPIs), including digital contact tracing (DCT), with a novel behaviour-driven agent-based model (ABM) to inform ongoing pandemic preparedness efforts. Our model’s Zeitgeber architecture delineates contextual characteristics, including daytime, daily routines, locations, and activities. Our method determines each agent’s current location and behaviour in a realistic environment under the restrictions of NPIs. We model viral load transfer between agents from contact duration, distance, and the infected agent’s infectiousness level. We examine the effects of DCT adoption, adherence, and compliance, both individually and combined with other NPIs, on key pandemic indicators, thus providing novel insight into infection dynamics. DCT combined with other NPIs reduces the total infections up to 52% for realistic behaviour, whereas DCT alone yielded a 43% reduction. Surprisingly however, some NPI combinations do not improve pandemic parameters. Our approach offers fine-grained analysis capabilities on the effectiveness of NPI combinations that cannot be obtained in human studies due to confounding effects. Thus our approach can inform future pandemic control efforts and prioritisation in pandemic preparedness.


Analysis of Melanin Concentration on Reflective Pulse Oximetry Using Monte Carlo Simulations

January 2025

·

39 Reads

IEEE Access

We investigate the impact of melanin concentration on the accuracy of oxygen saturation SpO 2 estimation using reflective photopletsymography (PPG) measurements. Monte Carlo (MC) simulations were used to simulate photon-tissue interaction and detection of light for systole and diastole. We model a reflective pulse oximeter and analyse the perfusion index ( PI ), ratio-of-ratios RoR , and SpO 2 depending on skin colour. We derive calibration models tailored to specific melanin concentrations ranging from 2.55% to 30.5%, as well as a generalised population model, to study reflective pulse oximeter performance. While current pulse oximeters are often calibrated for low melanin concentrations, our results show that calibrating for an appropriate skin colour range, substantially increases pulse oximeter performance. Skin colour-adapted calibration reduced the root mean square error Arms compared to a generalised population model by an average of 44%. SpO 2 estimation error was minimal at a distance of 4mm to 5mm between light source and photodetector of a reflective pulse oximeter. We conclude that skin colour adapted calibration should be applied to make reflective pulse oximeter similarly applicable to any human, independent of pigmentation.


Source-Detector Geometry Analysis of Reflective PPG by Measurements and Simulations

January 2025

·

12 Reads

IEEE Open Journal of Engineering in Medicine and Biology

Goal: We investigate the effect of source-detector geometry, including distance and angle, on the reflective photoplethysmography (PPG) signal. Methods: A porcine skin phantom was used for laboratory measurements and replicated by Monte Carlo simulations. Variations in sensor geometry were analysed. Results: Laboratory measurements and Monte Carlo simulations showed agreement for various geometry settings. With decreasing negative sensor angle, the differential path length factor and the average maximum penetration depth increases. Conclusions: Our analyses highlight the influence of sourcedetector geometry on the PPG DC signal. Based on our analysis of penetration depth and optical path length, the geometry effects can be transferred to the PPG AC signal too. MC simulations provide an important tool to optimise PPG performance.


FHIR up Ubicomp Data: Mastering Usability, Common Metrics, FAIRness, and Privacy

November 2024

·

12 Reads

IEEE Pervasive Computing

We propose and analyze a schema to harmonize data and metadata encoding for wearable and ubiquitous devices based on the Fast Healthcare Interoperability Resources standard. Currently, the lack of dataset encoding standards contributes to interpretation errors and proprietary implementations. Our proposed schema describes sensor devices, assessment data, and their characteristics. We develop and evaluate our schema regarding usability, common metric support, FAIRness, and privacy, using four stress monitoring datasets.


EghiFit: Smartphone based Behaviour Monitoring and Health Recommendation in a Weight Loss Intervention Study

November 2024

·

17 Reads

Background Current health recommender systems lack interactivity that relates to the current situation. Methods We designed and implemented an intervention study for obese patients that incorporates context information obtained from smartphone and smartwatch sensors, gamification, as well as joint goal management of patients and health coaches. We developed a health behaviour recommendation system comprising of a smartphone application, cloud platform for data management, and a data dashboard for coaches. Results We conducted a three months long study and analysed data from eight patients, focusing on system function, patient adherence, satisfaction and overall impact of the proposed system on changing health-related habits. Along with data analysis, we also provide patient feedback collected during interview round after the end of the study. Conclusion Patients could successfully implement the goals using the EghiFit app. Challenges regarding the data collection, recommendation synthesis, and patient engagement persist. Furthermore, reliable sensor data processing on current smartphone platforms is difficult due to system restrictions. Future research should further integrate sensor data, gaming, and health behaviour intervention design using smart devices.


AI and Health: Using Digital Twins to Foster Healthy Behavior

September 2024

·

27 Reads

·

1 Citation

This workshop brings researchers together to discuss and explore how artificial intelligence (AI) can be used to improve general health. During our workshop at the MuC conference, we will focus on three main areas: developing ethical AI health recommendations, exploring how smart technologies in our homes can influence our health habits, and understanding how different types of feedback can change our health behaviors. The workshop aims to be a space where various research areas meet, encouraging a shared understanding and creating new ways to use AI to encourage healthy living. By focusing on real-world applications of AI and digital twins, we seek to guide our discussions toward strategies that have a direct and positive impact on individual and societal health.


Patient Adherence and Challenges in a Weight Loss Study: Smartphone Data Stream and Gamification

September 2024

·

16 Reads

This paper presents findings from our implementation of a context aware health guidance system for obese individuals, with a focus on smartphone- and smartwatch-based health monitoring and participant adherence. To aid participants in weight loss, our system utilizes data from wearables and smartphones, integrating nutrition tracking and gamification elements into a smartphone application and a web-based health dashboard for health coaches. Eight participants completed a 120-day field study to evaluate the system and examine user adherence to health monitoring and the effectiveness of gamification in a weight loss program. Data on steps, sleep, heart rate, weather, and manually logged meals were collected. Adherence varied across data types, with step counts being the most consistently collected, while sleep and heart rate data were limited due to inconsistent smartwatch usage.


Citations (61)


... Abdominal Sounds have been used extensively to diagnose issues in the GI tract [18,20,28]. Wearable devices and automated sound processing methods for monitoring abdominal sounds have recently attracted attention from the research community [6,28,33,37]. Although abdominal sounds are within hearing range, they come from inside the body, and as such, their signal strength is significantly weakened by the layers of muscle, fat, and abdomen lining. ...

Reference:

GutIO: Toward Sensing and Inducing Gut Feelings with Abdominal Sounds
Multi-scale Bowel Sound Event Spotting in Highly Imbalanced Wearable Monitoring Data: Algorithm Development and Validation (Preprint)
  • Citing Article
  • July 2023

JMIR AI

... For wearable rehabilitation robots to function reliably, accurate sensing of body motion is essential. This is particularly important in upper-limb rehabilitation, where joint movements are complex and unpredictable [1], [2]. Biomechanical models paired with joint angle measurements, provide a means for detecting upper limb motion. ...

Where to mount the IMU? Validation of joint angle kinematics and sensor selection for activities of daily living

Frontiers in Computer Science

... In a previous study, we introduced a pipeline for real-time organ tracking and registration in AR-guided liver surgery, validated in a phantom study [12]. The pipeline processes RGBD data from a head-mounted AR device, eliminating the need for markers and additional sensors. ...

Image-based Live Tracking and Registration for AR-Guided Liver Surgery Using Hololens2: A Phantom Study
  • Citing Conference Paper
  • November 2023

... Since AudioSet only provides weak audio labels, no pretraining could be applied to the EffUNet decoder, therefore He initialisation (29) was used. Our previous experiments showed that transfer learning can improve BS spotting performance (30). ...

Segment-Based Spotting of Bowel Sounds Using Pretrained Models in Continuous Data Streams
  • Citing Article
  • May 2023

IEEE Journal of Biomedical and Health Informatics

... The DT in [7] considers the whole body but focuses on a specific goal, i.e., 3D pose reconstruction utilising monocular camera videos as input. Other DTs have been developed for the production of simulated data for future AI network training: electromyography signals [8], and gait analysis [9] algorithm testing.In [10], leveraged artificial intelligence, machine learning technology and DT paradigm to identify a real personalized motion axis of the tibiotalar joint. 3D models of distal extremities were generated using computed tomography data of normal patients. ...

Co-simulation of human digital twins and wearable inertial sensors to analyse gait event estimation

... However, recent advancements in artificial intelligence (AI) have significantly reshaped the methodologies employed by dietitians for assessing dietary intake and influenced how the general public manages their dietary habits [2], [3]. Over the past few years, there has been a marked increase in the development of AI algorithms and applications specifically designed for automatic food and ingredient recognition [4]- [11], food segmentation [12], food volume estimation [13]- [15], and recipe retrieval [16]- [20] or generation [21]- [23], as well as machinelearning-enabled smart sensors to detect eating events [24]- [26] and behaviors [27], [28] in free-living settings. In addition to the algorithmic and hardware advancements, the swift progression of AI-based technological solutions for dietary assessment can be partly attributed to the curation of largescale datasets by the community, for instance, Food2K [8], a dataset that encompasses over 1 million food images across 2,000 categories, and Recipe1M+ [29], which comprises over 1 million recipes and 13 million food images. ...

Proximity-based Eating Event Detection in Smart Eyeglasses with Expert and Data Models
  • Citing Conference Paper
  • December 2022

... Some studies have demonstrated measures of accuracy, including waveform similarity and amplitude comparison, that were consistent with estimates of upper limb joint kinematics. In 2022, Uhlenberg et al. (2022) introduced a framework based on OpenSense to estimate joint angle accuracy in combined ADLs with three different sensor fusion algorithms. The authors considered four ADLs with 10 participants and concluded that the most accurate estimations were obtained using the Madgwick and Mahony filters. ...

IMUAngle: Joint Angle Estimation with Inertial Sensors in Daily Activities
  • Citing Conference Paper
  • December 2022

... Another promising long-term solution could be the contact-free detection of the Temporalis muscle contraction by proximity sensors. This approach was already tested for dietary monitoring and eating event detection by Saphala et al. [31,32] and Selamat et al. [33] and could be adapted to our interface with little effort. ...

Non-contact temporalis muscle monitoring to detect eating in free-living using smart eyeglasses
  • Citing Conference Paper
  • September 2022

... Among these, pose and video data are especially valuable for sensor-based HAR, as they allow the integration of existing Pose2IMU and Video2IMU methods [42,49,50,54,56,90,94,95] to generate synchronized virtual accelerometer and gyroscope signals. In future work, we plan to jointly simulate these sensor modalities, leveraging LLM-based activity generation to create richly annotated, multi-modal datasets. ...

Comparison of Surface Models and Skeletal Models for Inertial Sensor Data Synthesis
  • Citing Conference Paper
  • September 2022

... We decreased the thickness d of epidermis according to literature [12] and added the seventh layer muscle according to [5]. A MC simulation framework was implemented and validated [13]. The MC simulation included the simulation of light intensities based on parameter configurations (see Fig. 1). ...

Simulation framework for reflective PPG signal analysis depending on sensor placement and wavelength
  • Citing Conference Paper
  • November 2022