This paper describes a simulator of ultrasound-guided prostate biopsies for cancer diagnosis. When performing biopsy series, the clinician has to move the ultrasound probe and to mentally integrate the real-time bi-dimensional images into a three-dimensional (3D) representation of the anatomical environment. Such a 3D representation is necessary to sample regularly the prostate in order to maximize the probability of detecting a cancer if any. To make the training of young physicians easier and faster we developed a simulator that combines images computed from three-dimensional ultrasound recorded data to haptic feedback. The paper presents the first version of this simulator.
Deep Dive into BiopSym: a simulator for enhanced learning of ultrasound-guided prostate biopsy.
This paper describes a simulator of ultrasound-guided prostate biopsies for cancer diagnosis. When performing biopsy series, the clinician has to move the ultrasound probe and to mentally integrate the real-time bi-dimensional images into a three-dimensional (3D) representation of the anatomical environment. Such a 3D representation is necessary to sample regularly the prostate in order to maximize the probability of detecting a cancer if any. To make the training of young physicians easier and faster we developed a simulator that combines images computed from three-dimensional ultrasound recorded data to haptic feedback. The paper presents the first version of this simulator.
Prostate cancer is the first cancer of men in many countries. When a cancer is suspected biopsies are collected in the gland, most often based on ultrasound (US) guidance, for further anatomo-pathologic analysis. The US probe is inserted into the rectum of the patient and a mechanical guide is attached to the probe for needle insertion (cf. fig. 1). Based on prostate exploration and image understanding the clinician repeatedly positions the probe and collect a sample in order to execute a predefined sampling scheme. The clinician has to face several difficulties: (1) the cancer, if any, is generally not visible in the US images; (2) the biopsy schemes are 3D whilst US images are 2D;
(3) the prostate may be moved and deformed by the US probe. Such difficulties limit the gesture accuracy and degrade the sensitivity of biopsies. We have developed a simulator for US-guided prostate biopsies (1) to enable relevant learning, (2) to make quantitative evaluation of operators possible and (3) to allow the quantitative comparisons of protocols. In addition to the previous clinical difficulties, the training phase of this intervention also poses some problems. A large part of the expertise concerning biopsies relies on the coordination between vision (US 2D images), haptic perception (holding the probe) and cognition (mental representation of the gland). This coordination is necessary to link visual and haptic feedback to anatomical knowledge, and take decision concerning samples location. Simple observation of clinicians at work is not sufficient to provide residents with all these aspects of knowledge and to train their links in situation. In particular, the cognitive processes involved for the treatment visual information and for the control of actions (coordination of 2D and 3D) are far from being easy to elaborate [1], [2].
Simulation is a possible answer to this training difficulty. Simulators may involve role plays, virtual systems, or mixed systems including a virtual and a robotic part [3]. In any case, we argue that simulation alone is not sufficient to manage relevant training. Focusing on training implies to define some pedagogical features of the training system. Even in the case of standardized patients1 , these latter are aware of the kind of reaction and feedback they have to give residents during and at the end of the training session [4]. In the following sections we give details on the way we address these issues in the case of Biopsym.
The simulator includes a haptic device, the Omni Phantom from Sensable Devices Inc., which allows the operator to move the virtual US probe with respect to the anatomy of the virtual patient (see fig. 2). The system generates US images corresponding to the position of the probe. A haptic feedback renders both the constraint related to the motion of the probe in the rectum and the biopsy gun percussion when a sample is collected.
Based on the motions of the virtual probe transmitted by the operator using the haptic device, the system generates corresponding US images. The haptic feedback renders both the constraint related to the motion of the probe in the rectum and the biopsy gun percussion when a sample is collected.
Two approaches can be used for US image simulation. One consists in generating US images [5] based on the modelling of US wave propagation and interaction with the human tissues. Such an approach generally requires complex models and is time consuming or images may have a limited realism when the model is simplified for computational efficiency. Another approach consists in using databases of recorded US exams and producing new images by exploiting this database. Interpolation may be necessary [6] when the image position does not correspond to recorded data. We have selected this second type of approach based on a 3D database. In this case, the generation of an image simply corresponds to slicing the volume in the direction of interest.
One major issue related to US images concerns the tissue deformations that may result from the image acquisition itself (probe pressure) or from external actions (needle insertions for instance). This issue not yet handled with this version of the simulator will be discussed in section 4. However, even without this function, the simulator may already help the trainee to navigate in the prostate volume, to visualize the prostate appearance and to correlate his/her probe 3D motions to corresponding 2D slices. We design targeted exercises for this training objective.
Biopsym uses open source libraries (see fig. 3) and is multi-platform.
3D US volumes have been acquired during real biopsy sessions at La Pitié Salpétrière Hospital for more than two years using a GE Voluson 730 system with a 3D endorectal US probe. By mid-june 2008, 87 patients were included. The database provides information about the prostate size, the patient age, the PSA level, etc. The SQL encoded database also includes information about operator
…(Full text truncated)…
This content is AI-processed based on ArXiv data.
