MR Prostate
Automatic prostate segmentation and lesion classification

In many countries, while the number of imaging examinations is steadily increasing, the number of experts is not growing proportionally. This results in an ever increasing workload for the radiologist. The effect of this is that shorter reading times can have a negative impact on the quality of the reports. Advanced image processing tools like MR Prostate support clinicians in analysing and processing medical images. MR Prostate automatically prepares valuable information for clinicians in order to facilitate the subsequent image interpretation workflow.

Available on syngo.via and AI-Rad Companion.

Overview MR Prostate

Siemens Healthineers MR Prostate supports Prostate reading and reporting. By deploying artificial intelligence, MR Prostate assists in the interpretation of multiparametric MR Prostate. The deep neural networks are designed to aid the detection and classification of prostate lesions and generate a pre-populated report.

The MR Prostate workflow supports optimal reading and reporting of the prostate examination. The images are sorted and displayed in such a way that all relevant images of the same orientation are side by side and move synchronously when the user scrolls through the images and standard tools support the analyses and diagnosis of the prostate.

MR Prostate performs an automatic segmentation of the prostate gland. Moreover, the software analyzes T2 weighted and diffusion weighted image series and provide an automatic detection and classification of regions suspicious for cancer which has been trained on PI-RADS >= 3 and positive biopsy labels.

MR Prostate supports in reading and reporting of prostate MRI by providing an automated segmentation and volumetric assessment of the prostate, the automatic detection and classification of lesions and a pre-populated report

MR Prostate aims also to provide support in the planning of targeted biopsies of the prostate gland using ultrasound-guided MR-US fusion biopsy. It allows the radiologist to communicate the location of lesions and the prostate segmentation in MR Prostate images to a urologist in order to plan the biopsies.

Results of the prostate evaluation are exported in a RTSS format that can be imported into MR-US fusion biopsy systems.

The findings use the latest PI-RADS prostate reporting guidelines: Level of Suspicion (LOS) between 60 and 80 is score 3, LOS > 80 and lesion <1.5cm score 4 and LOS >80 and lesion >1.5cm score 5.

Moreover, for diffusion data, the median value and 10th percentile of ADC are displayed. The location of the finding is added in a prostate pictogram, which is also part of the final report. The final report can be sent to PACS or to referring physicians.

Next to on-premise access through syngo.via, MR prostate evaluation and biopsy support is also available via the cloud and edge-based AI-Rad Companion.

Features

Automatic prostate segmentation and volume estimation, with the option for manual adjustments
Automatic determination of location and size of lesions
Calculation of PSA density based on the input of the patientʼs PSA value by the clinical user
Export in a suitable format for reading and archiving in PACS, as well as in a second format that can be imported by ultrasound system (e.g. RTStruct), allowing the urologist to perform targeted MR-US fusion biopsy

Scientific Outcomes

Researchers from Universities and hospitals in Skåne (Sweden) concluded that our deep learning algorithm performs similarly to radiologists in the assessment of prostate volume on MRI in a study published in European Radiology. See link below.