Reliable Mislabel Detection for Video Capsule Endoscopy Data

by Julia Werner, Julius Oexle, Oliver Bause, Maxime Le Floch, Franz Brinkmann, Hannah Tolle, Jochen Hampe, and Oliver Bringmann
In 48th Annual International Conference of the Engineering in Medicine and Biology Society, 2026.

Image Compression with Bubble-Aware Frame Rate Adaptation for Energy-Efficient Video Capsule Endoscopy

by Oliver Bause, Jörg Gamerdinger, Julia Werner, and Oliver Bringmann
In 48th Annual International Conference of the Engineering in Medicine and Biology Society, 2026.

CapsuleMotion: A Lightweight Real-Time Visual Motion Predictor for Capsule Endoscopy

by Oliver Bause, Julia Werner, and Oliver Bringmann
In 22nd Annual International Conference on Body Sensor Networks, 2026.

Seeing More with Less: Video Capsule Endoscopy with Multi-Task Learning

by Julia Werner, Oliver Bause, Julius Oexle, Maxime Le Floch, Franz Brinkmann, Jochen Hampe, and Oliver Bringmann
In 2025 Applications of Medical AI (AMAI) at MICCAI25, 2025.

Smart Video Capsule Endoscopy: Raw Image-Based Localization for Enhanced GI Tract Investigation

by Oliver Bause*, Julia Werner*, Paul Palomero Bernardo, and Oliver Bringmann (*Equal Contribution)
In 2025 32nd International Conference on Neural Information Processing (ICONIP) - Best Paper Runner-Up Award, 2025.

Systematic Hardware Integration Testing for Smart Video-based Medical Device Prototypes

by Oliver Bause, Julia Werner, and Oliver Bringmann
In 2025 47th Annual International Conference of the IEEE Engineering in Medicine and Biology Society (EMBC2025), 2025.

A Configurable and Efficient Memory Hierarchy for Neural Network Hardware Accelerator

by Oliver Bause, Paul Palomero Bernardo, and Oliver Bringmann
In MBMV 2024; 27. Workshop, pages 31–40, 2024.

Contribution to the Development of a Smart Video Capsule Endoscopy (VCE) System for Enhanced Gastrointestinal Diagnosis

Bachelor’s Thesis / Master’s Thesis / Student Research Project

Implemention and Evaluation of a Hardware-Efficient Image Compression Pipeline in SystemVerilog

Bachelor’s Thesis

Semi-Supervised Bubble Segmentation for Video Capsule Endoscopy

Bachelor’s Thesis

Abstract

Video capsule endoscopy produces vast numbers of frames per examination, particularly of the small bowel, which is difficult to reach by other means, and manually reviewing and labelling this material is impractical. This thesis therefore investigates bubble segmentation in capsule endoscopy frames using semi-supervised learning, so that a small labelled subset suffices to classify the entire dataset. A Python/OpenCV tool is developed to annotate frames and generate bitmasks marking bubble positions; these masks then train a U-Net-based deep CNN in PyTorch, pre-trained on RGB data, whose predictions are evaluated using accuracy and precision scores.

Integration of a Hardware Accelerator for On-Device Localization in an FPGA-Based Video Capsule Prototype

Master’s Thesis

Abstract

In video capsule endoscopy, most of the capsule’s limited energy budget is spent on illumination and the wireless transmission of every captured image, even though many frames show organs irrelevant to the examination. This thesis integrates the resource-efficient UltraTrail AI accelerator into the FPGA of an Ovesco capsule prototype, using a Lattice CrossLink-NX-33 and a SystemVerilog design, so that a compact neural network can localize the capsule within the GI tract on-device and restrict transmission to the organ of interest. The design is verified in simulation and tested on real hardware via a USB-to-SPI adapter and a hardware-in-the-loop camera model, and is extended with improved memory organization, a Hidden Markov Model for more accurate localization, and pipelining of image capture with model execution to further reduce energy consumption.

Evaluation of Energy-Efficient Image Compression Algorithms for Video Capsule Endoscopy

Bachelor’s Thesis

Abstract

Video Capsule Endoscopy transmits image data from a swallowable capsule under tight energy and bandwidth constraints, so compression must reduce data volume while preserving diagnostic image quality at low computational cost. This thesis implements and evaluates three low-complexity approaches — one-dimensional predictive coding (D1), two-dimensional top-left predictive coding (D2), and Run-Length Encoding — in lossless and near-lossless operation on a subset of the Rhode Island Capsule Endoscopy Dataset, together with a temporal preprocessing pipeline based on frame differencing and frame skipping, assessed in Python and on an embedded RISC-V platform. D1 achieved the highest compression ratios at acceptable reconstruction quality and remained computationally feasible on the embedded target, while D2 degraded substantially under quantization due to error propagation, making D1 the most promising approach for future VCE systems.

Hardware-Prototypentwicklung von programmierbaren Echtzeit-IOs in der Amaranth HDL

Master’s Thesis

HW-efficient Implementation of the Viterbi Algorithm for the Video Capsule Endoscopy

Master’s Thesis

HW-Development of a Minimal ML-based Image Processing Pipeline for Video Capsule Endoscopy

Master’s Thesis

Leveraging the Pico PIO API for Generalized Descriptions of State Machine-Based Interfaces

Master’s Thesis