No abstract

Lung ultrasound (US) imaging has the potential to be an effective point-of-care test for detection of COVID-19, due to its ease of operation with minimal personal protection equipment along with easy disinfection. The current state-of-the-art deep learning models for detection of COVID-19 are heavy models that may not be easy to deploy in commonly utilized mobile platforms in point-of-care testing. In this work, we develop a lightweight mobile friendly efficient deep learning model for detection of COVID-19 using lung US images. Three different classes including COVID-19, pneumonia, and healthy were included in this task. The developed network, named as Mini-COVIDNet, was bench-marked with other lightweight neural network models along with state-of-the-art heavy model. It was shown that the proposed network can achieve the highest accuracy of 83.2% and requires a training time of only 24 min. The proposed Mini-COVIDNet has 4.39 times less number of parameters in the network compared to its next best performing network and requires a memory of only 51.29 MB, making the point-of-care detection of COVID-19 using lung US imaging plausible on a mobile platform. Deployment of these lightweight networks on embedded platforms shows that the proposed Mini-COVIDNet is highly versatile and provides optimal performance in terms of being accurate as well as having latency in the same order as other lightweight networks. The developed lightweight models are available at ht. tps://github.com/navchetan-awasthi/Mini-COVIDNet.

Several algorithms exist to solve the photoacoustic image reconstruction problem depending on the expected reconstructed image features. These reconstruction algorithms promote typically one feature, such as being smooth or sharp, in the output image. Combining these features using a guided filtering approach was attempted in this work, which requires an input and guiding image. This approach act as a postprocessing step to improve commonly used Tikhonov or total variational regularization method. The result obtained from linear backprojection was used as a guiding image to improve these results. Using both numerical and experimental phantom cases, it was shown that the proposed guided filtering approach was able to improve (as high as 11.23 dB) the signal-to-noise ratio of the reconstructed images with the added advantage being computationally efficient. This approach was compared with state-of-the-art basis pursuit deconvolution as well as standard denoising methods and shown to outperform them.

As limited data photoacoustic tomographic image reconstruction problem is known to be ill-posed, the iterative reconstruction methods were proven to be effective in terms of providing good quality initial pressure distribution. Often, these iterative methods require a large number of iterations to converge to a solution, in turn making the image reconstruction procedure computationally inefficient. In this work, two variants of vector polynomial extrapolation techniques were deployed to accelerate two standard iterative photoacoustic image reconstruction algorithms, including regularized steepest descent and total variation regularization methods. It is shown using numerical and experimental phantom cases that these extrapolation methods that are proposed in this work can provide significant acceleration (as high as 4.7 times) along with added advantage of improving reconstructed image quality.

The methods available for solving the inverse problem of photoacoustic tomography promote only one feature-either being smooth or sharp-in the resultant image. The fusion of photoacoustic images reconstructed from distinct methods improves the individually reconstructed images, with the guided filter based approach being state-of-the-art, which requires that implicit regularization parameters are chosen. In this work, a deep fusion method based on convolutional neural networks has been proposed as an alternative to the guided filter based approach. It has the combined benefit of using less data for training without the need for the careful choice of any parameters and is a fully data-driven approach. The proposed deep fusion approach outperformed the contemporary fusion method, which was proved using experimental, numerical phantoms and in-vivo studies. The improvement obtained in the reconstructed images was as high as 95.49% in root mean square error and 7.77 dB in signal to noise ratio (SNR) in comparison to the guided filter approach. Also, it was demonstrated that the proposed deep fuse approach, trained on only blood vessel type images at measurement data SNR being 40 dB, was able to provide a generalization that can work across various noise levels in the measurement data, experimental setups as well as imaging objects.

Cardiovascular diseases (CVDs) remain the leading cause of death worldwide. An effective management and treatment of CVDs highly relies on accurate diagnosis of the disease. As the most common imaging technique for clinical diagnosis of the CVDs, US imaging has been intensively explored. Especially with the introduction of deep learning (DL) techniques, US imaging has advanced tremendously in recent years. Photoacoustic imaging (PAI) is one of the most promising new imaging methods in addition to the existing clinical imaging methods. It can characterize different tissue compositions based on optical absorption contrast and thus can assess the functionality of the tissue. This paper reviews some major technological developments in both US (combined with deep learning techniques) and PA imaging in the application of diagnosis of CVDs.

Lightweight segmentation models are becoming more popular for fast diagnosis on small and low cost medical imaging devices. This study focuses on the segmentation of the left ventricle (LV) in cardiac ultrasound (US) images. A new lightweight model [LV network (LVNet)] is proposed for segmentation, which gives the benefits of requiring fewer parameters but with improved segmentation performance in terms of Dice score (DS). The proposed model is compared with state-of-the-art methods, such as UNet, MiniNetV2, and fully convolutional dense dilated network (FCdDN). The model proposed comes with a postprocessing pipeline that further enhances the segmentation results. In general, the training is done directly using the segmentation mask as the output and the US image as the input of the model. A new strategy for segmentation is also introduced in addition to the direct training method used. Compared with the UNet model, an improvement in DS performance as high as 5% for segmentation with papillary (WP) muscles was found, while showcasing an improvement of 18.5% when the papillary muscles are excluded. The model proposed requires only 5% of the memory required by a UNet model. LVNet achieves a better trade-off between the number of parameters and its segmentation performance as compared with other conventional models. The developed codes are available at ht. tps://github.com/navchetanawasthi/Left_Ventricle_Segmentation.

No abstract