Real-Time Contrast-Enhanced Ultrasound Parametric Imaging in the Prostate

Introduction

Good morning.
My name is Peter Franking.
I'm a physicist working at Bruco Swiss in Geneva.
I'm a physicist.
And today I'd like to discuss with you some of the solutions we developed for the detection, improving the detection of contrast enhanced ultrasound in prostate cancer.

This presentation is entitled Realtime Contrast Enhanced Ultrasound parametric Imaging in the prostate and in prostate cancer.

Need for Accurate Imaging Methods

There is an urgent need for an accurate and reliable imaging methods, basically to first of all to detect and localize prostate cancer.
And this is important for guiding targeted biopsies.
Secondly, also to characterize the lesion and especially to reduce false positives.
Then another important point is the staging and grading to address the degree of aggressiveness of the lesion, and of course for therapy monitoring, assessing effectiveness of a treatment.

Current Criteria in Contrast Enhanced Ultrasound

Just as a reminder, in contrast enhanced ultrasound, currently accepted criteria to evaluate a suspicious lesion against normal tissue is to look at early and fast wash in, followed by a rapid wash out after a bo injection, of course, to look at the increase in contrast enhancement and to look for diffuse enhancement or abnormal perfusion patterns.

Challenges in Detection

However, the effective time window for detecting suspicious lesions after a bo bo injection is relatively short on the order of 20 to 30 seconds during a contrast examination.
Since prostate cancer is a multifocal disease, sometimes it can be very difficult to detect different lesions at the same time.

Developed Solution: Sonar Prostate Life

What we developed was a software solution called Sonar Prostate Life, and it combines a real time video capture unit with a robust smoothing algorithm.
Data are processed in two steps in real time.
First, we apply a real time maximum intensity projection imaging on the video data during the wash in phase.
And this is basically done to enhance vascular morphology.
Next, after the initial wash in phase, a parametric image is constructed on a pixel by pixel basis of the wash in rate parameter.
And this is based on linear linearized video data.

Time Intensity Curve and Parameters

Just as a reminder, below you see a time intensity curve.
So we have echo power as a function of time, and this is a typical bolus time intensity curve.
And based on this time intensity curve, different parameters can be calculated such as the time to peak or mean transit time.
In this work, we are particularly interested in the washing rate, which is basically the steepest slope of the time intensity curve.
Next, what we do is to apply a dynamic auto-scaling to normalize all the washing rate values.
And as a consequence, the perfusion information is summarized in a single parametric image.
That means that we have a spatial distribution of the machine rate values.

Software Screen Capture

So here you see a screen capture of the software that we developed.
So this is a typical ultrasound image used for contrast enhanced ultrasound examination.
So basically the whole screen is split up in two parts.
On the left we have the contrast image, which in this case, since we are looking before contrast agents injection is completely black due to the suppression of tissue.
On the right side, we have the standard BM mode image.
As you can appreciate, we see here the prostate gland, and this allows for and to have anatomical inflammation at the same time.

Regions of Interest and Color Palette

So for our analysis, we, the user has to identify different regions of interest.
First of all, the limitation region of interest, which basically defines the area that we use to process the data.
During the processing, we use a two, basically, a palette which consists of two parts.
First, a gray skill part, which will be applied during the wash in phase during the MIP analysis.
And secondly, we use a kind of jet palette, which color rises the washing rate parameters, where a blue collar in the identify or indicates a low rate value and red colors indicate high washing rate values.

Demonstration of the Process

So I will start the movie now, and as you can see, the contrast agent appears in the prostate.
We get the wash in and once for a specific pixel, the time intensity curve reaches its maximum, the washing rate value is calculated, and the corresponding pixel displays the washing rate value using this jet collar map.
So, as you can see now, we have in one single image a summary of the perfusion kinetics in the prostate.

Overlay on B-Mode Image

So now what we can do is to overlay this parametric image also on the B mode image, as you see on the right side, using a transparency overlay.
So now we can combine functional information and anatomical information.
And based on this image, we can see that particularly in the peripheral zone of the prostate, there are two suspicious areas, both on the right side of the right peripheral zone as well on the left peripheral zone.

Correlation with Histopathology

So the next thing is to correlate this with histopathology.
And in this particular example, you see that there is a good correspondence between the washing rate parametric image, and the histopathology.
In this particular case, the pathologist had identified two regions, which correspond very well with the washing rate parametric image, both on the right side as well as on the left side.

Conclusion

So in conclusion, this solution, sono prostate life, may be a new imaging method for the detection or the improved detection of prostate cancer.
By contrast enhanced ultrasound, it emphasizes differences in perfusion kinetics in real time, and could be considered for guiding targeted biopsies.
Moreover, a good agreement was found between washing rate parametric maps and histopathology, and a first pilot study done at the academic medical center in the Netherlands was very successful.
So thank you for your attention.