Gesture control: The hand for the contactless control of the MRI during interventional procedures

Congress:

ECR 2018

Poster Number:

C-1819

Type:

Scientific Exhibit

Keywords:

Workforce, Intraoperative, Computer Applications-General, MR, MR physics, Interventional non-vascular, Computer applications

Authors:

F. V. Güttler, A. Heinrich, F. Herrmann, U. K. M. Teichgräber; Jena/DE

DOI:

10.1594/ecr2018/C-1819

DOI-Link:

https://dx.doi.org/10.1594/ecr2018/C-1819

Fig. 1: Experimental setup with a 1.5T Magnetom Avanto: MR-compatible camera with...

Fig. 2: Experimental setup with a 1.5T Magnetom Avanto: The hand gestures presented to...

Fig. 3: The software was able to detect different hand gestures.

Aims and objectives

The objective of this proof of concept study is to allow interventionalists and surgeons the direct and touchless hand gesture based control of an MRI during (real time) interventional procedures, i.e. MR guided prostate biopsy.

Methods and materials

Gesture recognition software for controlling a 1.5T MRI (Magnetom Avanto, Siemens, Germany) was implemented based on a MR-compatible camera (MR CAM Model 12M, MRC Systems, Germany) and Matlab R2014b (MathWorks, USA) (Fig. 1). A MR compatible in-room display shows the MRI user interface via virtual desktop mirroring. The hand gestures presented to the MR camera were analysed by the software running on a separate desktop computer in the anteroom (Fig. 2). The commands to control the MRI console were transmitted by an automation software script...

Results

The software was able to detect different hand gestures (Fig. 3): to enable the gesture recognition, to move the FOV left/right or up/down (position of hand decides move direction and speed) and tilt MRI slice orientation by 90 degrees (sagittal, coronal, transverse). Additionally, the software allows to navigate through the already acquired images. The system was able to identify all given hand gestures. All test persons were able to control the MRI with gesture recognition to the tumor suspicious areas in average 182 ± 68...

Conclusion

This proof of concept study presents the implementation and evaluation of a camera based computer vision system for gesture recognition, allowing interoperative control of a MRI using simple hand signs. A gesture control of MRI allows new workflow concepts for a sterile and efficient work.

Personal information

Felix Güttler Department of Radiology, University Hospital Jena, Friedrich Schiller University, Germany Am Klinikum 1 07747 Jena, Germany Fon: +49 3641 9 - 324 831 Fax: +49 3641 9 - 324 832 e-mail: [email protected]

References

Güttler, F., Heinrich, A., Hench, T., & Teichgräber, U. (2016). Gestensteuerung: Die Hand zur berührungslosen Steuerung des MRT. In RöFo-Fortschritte auf dem Gebiet der Röntgenstrahlen und der bildgebenden Verfahren (Vol. 188, No. S 01, p. WISS104_6). Sural, S., et al. (2002). Segmentation and histogram generation using the HSV color space for image retrieval. Proceedings. International Conference on Image Processing 2: 589-592, Rochester, USA. Jolesz, F. A., et al. (2001). Integration of interventional MRI with computer-assisted surgery. J. Magn. Reson. Imaging 13(1): 69-77. Weiss, C. R.,...