Biopsym: a learning environment for trans-rectal ultrasound guided prostate biopsies.

This paper describes a learning environment for image-guided prostate biopsies for cancer diagnosis; it is based on an ultrasound probe simulator virtually exploring real datasets obtained from patients. The aim is to make the training of young physicians easier and faster with a tool that combines lectures, biopsy simulations and recommended exercises to master this medical gesture. It is designed particularly to help improve the acquisition of the three-dimensional representation of the prostate required for practicing biopsy sequences. The simulator uses haptic feedback to compute the position of the virtual probe from three-dimensional (3D) ultrasound recorded data. This paper presents the current version of this learning environment.

Estimation of urinary stone composition by automated processing of CT images.

The objective of this article was developing an automated tool for routine clinical practice to estimate urinary stone composition from CT images based on the density of all constituent voxels. A total of 118 stones for which the composition had been determined by infrared spectroscopy were placed in a helical CT scanner. A standard acquisition, low-dose and high-dose acquisitions were performed. All voxels constituting each stone were automatically selected. A dissimilarity index evaluating variations of density around each voxel was created in order to minimize partial volume effects: stone composition was established on the basis of voxel density of homogeneous zones. Stone composition was determined in 52% of cases. Sensitivities for each compound were: uric acid: 65%, struvite: 19%, cystine: 78%, carbapatite: 33.5%, calcium oxalate dihydrate: 57%, calcium oxalate monohydrate: 66.5%, brushite: 75%. Low-dose acquisition did not lower the performances (P < 0.05). This entirely automated approach eliminates manual intervention on the images by the radiologist while providing identical performances including for low-dose protocols.

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.

Mapping of transrectal ultrasonographic prostate biopsies: quality control and learning curve assessment by image processing.

OBJECTIVE: Mapping of transrectal ultrasonographic (TRUS) prostate biopsies is of fundamental importance for either diagnostic purposes or the management and treatment of prostate cancer, but the localization of the cores seems inaccurate. Our objective was to evaluate the capacities of an operator to plan transrectal prostate biopsies under 2-dimensional TRUS guidance using a registration algorithm to represent the localization of biopsies in a reference 3-dimensional ultrasonographic volume. METHODS: Thirty-two patients underwent a series of 12 prostate biopsies under local anesthesia performed by 1 operator using a TRUS probe combined with specific third-party software to verify that the biopsies were indeed conducted within the planned targets. RESULTS: The operator reached 71% of the planned targets with substantial variability that depended on their localization (100% success rate for targets in the middle and right parasagittal parts versus 53% for targets in the left lateral base). Feedback from this system after each series of biopsies enabled the operator to significantly improve his dexterity over the course of time (first 16 patients: median score, 7 of 10 and cumulated median biopsy length in targets of 90 mm; last 16 patients, median score, 9 of 10 and a cumulated median length of 121 mm; P = .046). CONCLUSIONS: In addition to being a useful tool to improve the distribution of prostate biopsies, the potential of this system is above all the preparation of a detailed "map" of each patient showing biopsy zones without substantial changes in routine clinical practices.

[Image fusion: use in the control of the distribution of prostatic biopsies]. / Fusion d'images: application au contrôle de la distribution des biopsies prostatiques.

Prostate biopsies are performed under 2D TransRectal UltraSound (US) guidance by sampling the prostate according to a predefined pattern. Modern image processing tools allow better control of biopsy distribution. We evaluated the accuracy of a single operator performing a pattern of 12 ultrasound-guided biopsies by registering 3D ultrasound control images acquired after each biopsy. For each patient, prostate image alignment was performed automatically with a voxel-based registration algorithm allowing visualization of each biopsy trajectory in a single ultrasound reference volume. On average, the operator reached the target in 60% of all cases. This study shows that it is difficult to accurately reach targets in the prostate using 2D ultrasound. In the near future, real-time fusion of MRI and US images will allow selection of a target in previously acquired MR images and biopsy of this target by US guidance.