Sadia Sultana Chowa

The identification and early treatment of retinal disease can help to prevent loss of vision. Early diagnosis allows a greater range of treatment options and results in better outcomes. Optical coherence tomography (OCT) is a technology used by ophthalmologists to detect and diagnose certain eye conditions. In this paper, human retinal OCT images are classified into four classes using deep learning. Several image preprocessing techniques are employed to enhance the image quality. An augmentation technique, called generative adversarial network (GAN), is utilized in the Drusen and DME classes to address data imbalance issues, resulting in a total of 130,649 images. A lightweight optimized compact convolutional transformers (OCCT) model is developed by conducting an ablation study on the initial CCT model for categorizing retinal conditions. The proposed OCCT model is compared with two transformer-based models: vision Transformer (ViT) and Swin Transformer. The models are trained and evaluated with 32 × 32 sized images of the GAN-generated enhanced dataset. Additionally, eight transfer learning models are presented with the same input images to compare their performance with the OCCT model. The proposed model’s stability is assessed by decreasing the number of training images and evaluating the performance. The OCCT model’s accuracy is 97.09%, and it outperforms the two transformer models. The result further indicates that the OCCT model sustains its performance, even if the number of images is reduced.

Autism spectrum disorder (ASD) is a complex neurodevelopmental condition that presents significant diagnostic challenges due to its varied symptoms and nature. This study aims to improve ASD classification using advanced deep learning techniques applied to neuroimaging data. We developed an automated system leveraging the ABIDE-1 dataset and a novel lightweight quantized one-dimensional (1D) Convolutional Neural Network (Q-CNN) model to analyze fMRI data. Our approach employs the NIAK pipeline with multiple brain atlases and filtering methods. Initially, the Regions of Interest (ROIs) are converted into feature vectors using tangent space embedding to feed into the Q-CNN model. The proposed 1D-CNN is quantized through Quantize Aware Training (QAT). As the quantization method, int8 quantization is utilized, which makes it both robust and lightweight. We propose a federated learning (FL) framework to ensure data privacy, which allows decentralized training across different data centers without compromising local data security. Our findings indicate that the CC200 brain atlas, within the NIAK pipeline’s filt-global filtering methods, provides the best results for ASD classification. Notably, the ASD classification outcomes have achieved a significant test accuracy of 98% using the CC200 and filt-global filtering techniques. To the best of our knowledge, this performance surpasses previous studies in the field, highlighting a notable enhancement in ASD detection from fMRI data. Furthermore, the FL-based Q-CNN model demonstrated robust performance and high efficiency on a Raspberry Pi 4, underscoring its potential for real-world applications. We exhibit the efficacy of the Q-CNN model by comparing its inference time, power consumption, and storage requirements with those of the 1D-CNN, quantized CNN, and the proposed int8 Q-CNN models. This research has made several key contributions, including the development of a lightweight int8 Q-CNN model, the application of FL for data privacy, and the evaluation of the proposed model in real-world settings. By identifying optimal brain atlases and filtering methods, this study provides valuable insights for future research in the field of neurodevelopmental disorders.

This study proposes a novel approach for breast tumor classification from ultrasound images into benign and malignant by converting the region of interest (ROI) of a 2D ultrasound image into a 3D representation using the point-e system, allowing for in-depth analysis of underlying characteristics. Instead of relying solely on 2D imaging features, this method extracts 3D mesh features that describe tumor patterns more precisely. Ten informative and medically relevant mesh features are extracted and assessed with two feature selection techniques. Additionally, a feature pattern analysis has been conducted to determine the feature’s significance. A feature table with dimensions of 445 × 12 is generated and a graph is constructed, considering the rows as nodes and the relationships among the nodes as edges. The Spearman correlation coefficient method is employed to identify edges between the strongly connected nodes (with a correlation score greater than or equal to 0.7), resulting in a graph containing 56,054 edges and 445 nodes. A graph attention network (GAT) is proposed for the classification task and the model is optimized with an ablation study, resulting in the highest accuracy of 99.34%. The performance of the proposed model is compared with ten machine learning (ML) models and one-dimensional convolutional neural network where the test accuracy of these models ranges from 73 to 91%. Our novel 3D mesh-based approach, coupled with the GAT, yields promising performance for breast tumor classification, outperforming traditional models, and has the potential to reduce time and effort of radiologists providing a reliable diagnostic system.

Purpose: An automated computerized approach can aid radiologists in the early diagnosis of breast cancer. In this study, a novel method is proposed for classifying breast tumors into benign and malignant, based on the ultrasound images through a Graph Neural Network (GNN) model utilizing clinically significant features. Method: Ten informative features are extracted from the region of interest (ROI), based on the radiologists’ diagnosis markers. The significance of the features is evaluated using density plot and T test statistical analysis method. A feature table is generated where each row represents individual image, considered as node, and the edges between the nodes are denoted by calculating the Spearman correlation coefficient. A graph dataset is generated and fed into the GNN model. The model is configured through ablation study and Bayesian optimization. The optimized model is then evaluated with different correlation thresholds for getting the highest performance with a shallow graph. The performance consistency is validated with k-fold cross validation. The impact of utilizing ROIs and handcrafted features for breast tumor classification is evaluated by comparing the model’s performance with Histogram of Oriented Gradients (HOG) descriptor features from the entire ultrasound image. Lastly, a clustering-based analysis is performed to generate a new filtered graph, considering weak and strong relationships of the nodes, based on the similarities. Results: The results indicate that with a threshold value of 0.95, the GNN model achieves the highest test accuracy of 99.48%, precision and recall of 100%, and F1 score of 99.28%, reducing the number of edges by 85.5%. The GNN model’s performance is 86.91%, considering no threshold value for the graph generated from HOG descriptor features. Different threshold values for the Spearman’s correlation score are experimented with and the performance is compared. No significant differences are observed between the previous graph and the filtered graph. Conclusion: The proposed approach might aid the radiologists in effective diagnosing and learning tumor pattern of breast cancer.