Fraunhofer Institute for Computer Graphics Research IGD
Fraunhofer Institute for Computer Graphics Research IGD

ICON VirtoFusion: Virtual Autopsy with Automated Sensor Fusion

The ICON VirtoFusion project is developing an automated system for the postmortem digitization and examination of deceased persons. Together with the University of Zurich, Fraunhofer IGD is working on an integrated workflow that captures, processes, and combines imaging data from different modalities.

Fraunhofer IGD contributes expertise in the development, design, and prototypical implementation of customized sensor and hardware solutions, software development, system integration, as well as automated data acquisition and processing. Its expertise also includes the calibration of sensors and hardware components, the registration and overlay of multimodal data, and 3D reconstruction. The University of Zurich complements this expertise with its knowledge of forensic medicine and postmortem imaging.

The technologies used include computed tomography (CT), color-calibrated surface capture, and multispectral imaging. Sensors and hardware components are to be calibrated automatically, with the resulting data transferred into a common coordinate system and spatially overlaid. Initial automated analyses and correlations of the acquired sensor data are intended to support the examination process and reveal relevant relationships at an early stage. The goal is to establish a fully integrated, automated workflow from data acquisition and data fusion through to analysis.

Pedro Santos and Hasan Kutlu from Fraunhofer IGD, and Dominic Gascho and Prof. Dr. Michael Thali from the University of Zurich (from left to right), together with the members of the ICON VirtoFusion Advisory Board joining remotely. The board supports the project with scientific, application-oriented and strategic expertise.
© University of Zurich (UZH)
© University of Zurich (UZH)
© University of Zurich (UZH)

Until now, data from different imaging modalities have been handled separately: sensors are operated individually, images are processed separately, and the resulting datasets are subsequently correlated with one another. In particular, the spatial overlay of CT, surface, color, and multispectral data is complex and time-consuming. The ICON VirtoFusion project therefore combines existing imaging techniques into an integrated overall system: sensors are to be controlled centrally, while their data are automatically registered, analyzed, processed, visualized, and displayed together. Analyzing correlations between sensor data could generate new insights and support the validation of forensic findings. This reduces manual processing steps and potential sources of error while making findings easier to compare.

For forensic pathologists, the system is intended to simplify the technical preparation of data and provide a basis for expert analysis; the automated correlation of different sensor data may also provide additional indications. Jointly registered image data also open up new opportunities for research and teaching. Existing scanning systems and real-world forensic application environments at the University of Zurich are available for testing and validation.

ICON VirtoFusion began on July 1, 2026, and is scheduled to run for three years. Following the requirements analysis, the sensors will be integrated step by step, while calibration, data overlay, analysis, and visualization will be automated and a demonstrator developed. Toward the end of the project, an extended testing and validation phase is planned at the University of Zurich.

The project is funded internally as part of the Fraunhofer "CONNECT Science: ICON - International Cooperation and Networking" program.

Project Partner & Advisors

Project Partner from University of Zurich

Advisors:

Frequently Asked Questions about ICON VirtoFusion

Expand all Close all

  • In the context of ICON VirtoFusion, a virtual autopsy refers to the digital examination of deceased persons using imaging technologies. These include computed tomography, 3D surface capture, and other optical imaging techniques.

  • Sensor fusion refers to combining data from different measurement and imaging systems. In ICON VirtoFusion, CT, surface, color, and multispectral data, among others, are to be automatically registered and spatially combined.

  • Not conclusively. In addition to automated data acquisition and fusion, ICON VirtoFusion aims to develop methods for the automated analysis of image and sensor data. These methods can identify anomalies and correlations and provide initial indications for further examination. The results serve as a basis for expert analysis by forensic pathologists. The final assessment of findings and determination of the cause of death remain medical and professional responsibilities.

  • In the long term, jointly registered image data could help document forensic findings in a transparent and spatially coherent manner and make use of correlations between separately acquired sensor data to generate additional insights from the digital records. The aim is for the results to be usable in court proceedings. Whether, and to what extent, this will ultimately be possible remains open. Admissibility in court has not yet been demonstrated as an outcome of the project.