Research

Deep learning has been used extensively for medical image analysis applications, assuming the training and test data adhere to the same probability distributions. However, a common challenge arises when dealing with medical images generated by different systems or even the same system with varying parameter settings. Such images often contain diverse textures and noise patterns, violating the assumption. Consequently, models trained on data from one machine or setting usually struggle to perform effectively on data from another. To address this issue in ultrasound images, we proposed a Generative Adversarial Network (GAN) based model in this paper. We formulated image harmonization and denoising tasks as an image-to-image translation task, wherein we modified the texture pattern and reduced noise in Carotid ultrasound images while keeping the image content (the anatomy) unchanged. The performance was evaluated using feature distribution and pixel-space similarity metrics. In addition, blood-to-tissue contrast and influence on computed risk markers (Gray scale median, GSM) were evaluated. The results showed that domain adaptation was achieved in both tasks (histogram correlation 0.920 and 0.844), as compared to no adaptation (0.890 and 0.707), and that the anatomy of the images was retained (structure similarity index measure of the arterial wall 0.71 and 0.80). In addition, the image noise level (contrast) did not change in the image harmonization task (-34.1 vs 35.2 dB) but was improved in the noise reduction task (-23.5 vs -46.7 dB). The model outperformed the CycleGAN in both tasks. Finally, the risk marker GSM increased by 7.6 (p<0.001) in task 1 but not in task 2. We conclude that domain translation models are powerful tools for ultrasound image improvement while retaining the underlying anatomy but that downstream calculations of risk markers may be affected. 

ArXiv, 2024

https://arxiv.org/abs/2407.05163

IEEE Access 

https://ieeexplore.ieee.org/abstract/document/8519726/

This paper presents the analysis of real-life medical big data obtained from a hospital in central China from 2013 to 2015 for risk assessment of cerebral infarction disease. We propose a new recurrent convolutional neural network (RCNN)-based disease risk assessment multimodel by utilizing structured and unstructured text data from the hospital. In the proposed model, the convolutional layer becomes a bidirectional recurrent neural network by utilizing the intra-layer recurrent connection within the convolutional layer. Each neuron within convolutional layer receives feedforward and recurrent inputs from the previous unit and neighborhood, respectively. In addition to step-by-step recurrent operation, the region of context capture increases, thereby facilitating fine-grain feature extraction. Furthermore, we use a data parallelism approach over multimodel data during training and testing of the proposed model. Results show that the data parallelism approach leads to fast conversion speed. The RCNN-based model works differently from the traditional convolutional neural network and other typical methods. The proposed model exhibits a prediction accuracy of 96.02%, which is higher than those of typical existing methods. 

Computer Methods and Programs in Biomedicine

https://www.sciencedirect.com/science/article/abs/pii/S0169260719311708

Nowadays computer-aided disease diagnosis from medical data through deep learning methods has become a wide area of research. Existing works of analyzing clinical text data in the medical domain, which substantiate useful information related to patients with disease in large quantity, benefits early-stage disease diagnosis. However, benefits of analysis not achieved well when the traditional rule-based and classical machine learning methods used; which are unable to handle the unstructured clinical text and only a single method is not able to handle all challenges related to the analysis of the unstructured text, Moreover, the contribution of all words in clinical text is not the same in the prediction of disease. Therefore, there is a need to develop a neural model which solve the above clinical application problems, is an interesting topic which needs to be explored. Thus considering the above problems, first, this paper present self-attention based recurrent convolutional neural network (RCNN) model using real-life clinical text data collected from a hospital in Wuhan, China. This model automatically learns high-level semantic features from clinical text by using bi-direction recurrent connection within convolution. Second, to deal with other clinical text challenges, we combine the ability of RCNN with the self-attention mechanism. Thus, self-attention gets the focus of the model on essential convolve features which have effective meaning in the clinical text by calculating the probability of each convolve feature through softmax. The proposed model is evaluated on real-life hospital dataset and used measurement metrics as Accuracy and recall. Experiment results exhibit that the proposed model reaches up to accuracy 95.71%, which is better than many existing methods for cerebral infarction disease. This article presented the self-attention based RCNN model by combining the RCNN with self-attention mechanism for prediction of cerebral infarction disease. The obtained results show that the presented model better predict the cerebral infarction disease risk compared to many existing methods. The same model can also be used for the prediction of other disease risks. 

Future Generation Computer Systems

https://www.sciencedirect.com/science/article/abs/pii/S0167739X18332291

Magnetic Resonance Imaging (MRI) is dominant modality for infant brain analysis. Segmenting the whole infant MRI brain into number of tissues such as Cerebrospinal fluid (CSF), White matter (WM), and Gray Matter (GM) are highly desirable in the clinical environment. However, traditional methods tend to be degrading due to low contrast between GM and WM in isointense phase (about 6–8 months of early life). Recently, Convolutional Neural Network (CNN) emerged as a robust intelligent approach to examine medical image. The UNet model is among the preferred CNN models that have been widely used for medical imaging applications and achieved excellent results. The UNet architecture is a combination of convolutional, pooling, and up-sampling layers. Recently, 3D-UNet architecture used to exploit 3D-contextual information of volumetric data in many applications. However, CNN faces challenge to distinguish the similar brain tissues. In this paper, we present a variant of 3D-UNet to extract the volumetric contextual information of medical data. We propose a novel combined architecture of dense connection, residual connection, and inception module. The proposed architecture contains three stages, namely the densely connected stage, a residual inception stage, and an up-sampling stage. Our proposed approach provides state-of-art results in comparison to other existing approaches. This suggested approach achieves dice scores of 0.95, 0.905, and 0.92 in CSF, WM, and GM tissues respectively. 

ArXiv

https://arxiv.org/abs/2406.00696

In this study, we proposed a model for skin disease classification using a Bilinear Convolutional Neural Network (BCNN) with a Constrained Triplet Network (CTN). BCNN can capture rich spatial interactions between features in image data. This computes the outer product of feature vectors from two different CNNs by a bilinear pooling. The resulting features encode second-order statistics, enabling the network to capture more complex relationships between different channels and spatial locations. The CTN employs the Triplet Loss Function (TLF) by using a new loss layer that is added at the end of the architecture called the Constrained Triplet Loss (CTL) layer. This is done to obtain two significant learning objectives: inter-class categorization and intra-class concentration with their deep features as often as possible, which can be effective for skin disease classification. The proposed model is trained to extract the intra-class features from a deep network and accordingly increases the distance between these features, improving the model’s performance. The model achieved a mean accuracy of 93.72%. 

Future Generation Computer Systems

https://www.sciencedirect.com/science/article/abs/pii/S0167739X19334600 

Convolution and recurrent neural network have obtained remarkable performance in natural language processing(NLP). Moreover, from the attention mechanism perspective convolution neural network(CNN) is applied less than recurrent neural network(RNN). Because RNN can learn long-term dependencies and gives better results than CNN. But CNN has its own advantage, can extract high-level features by using its local fix size context at the input level. Thus, this paper proposed a new model based on RNN with CNN-based attention mechanism by using the merits of both architectures together in one model. In the proposed model, first, CNN learns the high-level features of sentence from input representation. Second, we used attention mechanism to get the attention of the model on the features which contribute much in the prediction task by calculating the attention score from features context generated from CNN filters. Finally, these features context from CNN with attention score are commonly used at the RNN to process them sequentially. To validate the model we experiment on three benchmark datasets. Experiment results and their analysis demonstrate the effectiveness of the model. 

1) Disease Prediction Using Multimodal Healthcare Data (Sept 2019-Aug 2021)

Deanship of Scientific Research funded this project at King Saud University, Riyadh, Saudi Arabia, through the Research Group Project under grant number RGP-049.

Description: In this project, we worked on multimodal medical data from 2013 to 2015, consisting of structured and unstructured data, to address the following problem in disease prediction applications. 1) How to extract main characteristics usually not mentioned in patient medical records that affect chronic diseases for risk assessment. 2) Difficulty extracting fine-grain disease-related features from unstructured text data using traditional CNN? These phenomena limit the accurate prediction of disease risk assessment.