(C-82) End-to-end Ultrasound/Photoacoustic Images Dual-Processing & Enhancing Application

Introduction:: Ultrasound (US) and photoacoustic (PA) imaging are non-invasive medical imaging techniques that have become increasingly popular in clinical practice due to their ability to provide valuable information about tissue structure and function. However, the quality of the images obtained from these techniques can be limited by factors such as noise, artifacts and low contrast. To address these challenges, a Matlab-based app that can process and enhance ultrasound and photoacoustic images in real-time was designed and developed.

The app was first developed to process and have a better view of the images, then it was innovated to utilize a range of image processing techniques, including filtering, segmentation, and feature extraction, to enhance the quality of the images obtained from these techniques. In addition, the app incorporates a range of algorithms for noise reduction, contrast enhancement, and image fusion, which can be customized to suit the needs of different imaging applications.

The app, moreover, is being developed to be user-friendly and can be easily integrated into existing imaging workflows. The app is also highly customizable, allowing users to adjust parameters and settings to optimize image quality and enhance the accuracy of their diagnoses.



In this study, key features and capabilities of the app for processing and enhancing ultrasound and photoacoustic will be shown. Consequently, the effectiveness in improving image quality and aiding in the diagnosis of a range of medical conditions will be demonstrated.


Materials and Methods:: We developed an app using MATLAB’s App Designer Tool, which proves a user-friendly interface with customizable features. The app accepts US and PA images with 3D array dta, and allows users to adjust parameters to optimize image quality.

Data processing: The app supports up to one US image and 3 PA images, allowing users to track the lateral view frames by frame through the axial view image. PA images are fused to display multispectral images, with adjustable color weighting.

View Enhancement: The app incorporates several image processing techniques to enhance image quality. The US view offers logscale, background removal, and rescaling options, with the latter reducing noise and allowing for lateral view cropping. In the PA tab, we included Hessian filter function to smooth segments edges, change colormap to change the brightness and contrast of the images, and change angle function to rotate the data.

The app was tested using a range of US and PA images obtained from experimental sources. The images were processed using the app’s various image processing techniques, and the resulting images were compared to the original images to evaluate the effectiveness of the app’s processing algorithms.


Results, Conclusions, and Discussions:: Results


US Rescaling & Background Removal: Fig.3 represents the real data US image after rescaling. The lateral view image was cropped from the region of 26 mm to 48 mm, resulting in a reduction of noise in the top view image. The background removal function was then applied to the data, which effectively reduced the noise, but also removed significant parts of the image.

PA Colormap Change, Background Removal, and Hessian Filter: Fig. 4 shows the reconstructed data of PA. First, the colormap of the images was changed to the maximum of 6000, increasing the contrast of the images. Then, the data’s background was removed, displaying the data with prominent vessels. Lastly, the Hessian filter function was applied to the data, resulting in smoother vessel edges.

PA Angle change: The previous data was from linear array system, in which, the sample was only scanned from one side. In Fig. 5, the data is from ring array system, which provides data from different angles. The change angle function in the app supports users to change the angle to have a better look at both top view and elevational view of the sample.


Conclusions and Discussions:

The developed app using MATLAB’s App Designer Tool provides a user-friendly interface for processing and enhancing US and PA data. The app allows users to upload multiple images and adjust various parameters to optimize image quality.

The results demonstrate the effectiveness of the image processing techniques incorporated into the app. Rescaling and background removal of the US reduced noise and improved lateral view image quality. Changing the colormap and applying background removal and Hessian filter to the PA image resulted in a better contrast and smoother vessel edges. The angle change function provides users with better view from different angles of the samples’ side view.

The developed app can be used by medical professionals to improve the quality of US and PA images, making it a valuable tool for diagnostic and research purposes.

Future work includes integrating more advanced image processing techniques and incorporating machine learning algorithms to automate image processing and enhance analysis accuracy.




Acknowledgements (Optional): :

References (Optional): :