Current Research Project
By Sidratul Montaha ,Sami Azam ,*,Abul Kalam Muhammad Rakibul Haque Rafid 1,Pronab Ghosh ,Md. Zahid Hasan 1,Mirjam Jonkman and Friso De Boer [https://doi.org/10.3390/biology10121347].
Identification and treatment of breast cancer at an early stage can reduce mortality. Currently, mammography is the most widely used effective imaging technique in breast cancer detection. However, an erroneous mammogram based interpretation may result in false diagnosis rate, as distinguishing cancerous masses from adjacent tissue is often complex and error-prone.
Six pre-trained and fine-tuned deep CNN architectures: VGG16, VGG19, MobileNetV2, ResNet50, DenseNet201, and InceptionV3 are evaluated to determine which model yields the best performance. We propose a BreastNet18 model using VGG16 as foundational base, since VGG16 performs with the highest accuracy. An ablation study is performed on BreastNet18, to evaluate its robustness and achieve the highest possible accuracy. Various image processing techniques with suitable parameter values are employed to remove artefacts and increase the image quality. A total dataset of 1442 preprocessed mammograms was augmented using seven augmentation techniques, resulting in a dataset of 11,536 images.
Our proposed approach based on image processing, transfer learning, fine-tuning, and ablation study has demonstrated a high correct breast cancer classification while dealing with a limited number of complex medical images.
By Sidratul Montaha1 , *Sami Azam2 , A.K.M. Rakibul Haque Rafid3 , Md. Zahid Hasan4 , Asif Karim5 , Ashraful Islam6 [DOI 10.1109/ACCESS.2022.3179577 ].
Identification of brain tumors and accurate grading at an early stage are crucial in cancer diagnosis, as a timely diagnosis can increase the chances of survival. Considering the challenges and risks of tumor biopsies, noninvasive imaging procedures such as Magnetic Resonance Imaging (MRI) are extensively used in analyzing brain tumors. Recent advances in the field of medical imaging with deep learning using three dimensional (3D) MRI is aiding the clinical experts significantly in the diagnosis of brain tumor. In this study, three BraTS MRI datasets named BraTS 2018, BraTS 2019 and BraTS 2020 are employed to classify brain tumor into high-grade glioma (HGG) and low-grade glioma (LGG) where each of the datasets contains four different sequences of 3D MRI brain images named T1-weighted MRI (T1), T1-weighted MRI with contrast enhancement (T1ce), T2-weighted MRI (T2), and Fluid Attenuated Inversion Recovery (FLAIR) for a single patient.
This research is composed of two approaches where, in the first part, we propose a hybrid deep learning model named TimeDistributed-CNN-LSTM (TD-CNNLSTM) combining 3D Convolutional Neural Network (CNN) and Long Short Term Memory (LSTM) where each layer of the architecture is wrapped with a TimeDistributed function. The objective of developing the hybrid model is to consider all the four 3D MRI sequences of each patient as a single input data because every sequence contains necessary information on tumor that can have an impact on improving cancer detection performance. However, interpreting all the MRI sequences together with optimal performance especially in 3D is quite challenging. Therefore, the model is developed with optimal configuration based on highest accuracy performing ablation study for layer architecture and hyperparameters.
In the second part, a 3D CNN model is trained respectively with each of the MRI sequences to compare the performance with TD-CNN-LSTM model. In this regard, for both of the models, BraTS 2018 and BraTS 2019 are combined to increase the number of images and used as train dataset where BraTS 2020 dataset is employed as the test dataset. Moreover, before training the models the datasets is preprocessed to ensure the highest performance. Our results of these two approaches demonstrate that the TD-CNN-LSTM network outperforms 3D CNN achieving the highest test accuracy of 98.90%. Later, to evaluate the performance consistency, the TD-CNN-LSTM model is evaluated with K-fold cross validation using different k values. The approach of putting together all the MRI sequences at a time using hybrid CNN-LSTM approach with good generalization capability to classify brain tumor can be used effectively in future Computer Aided Diagnosis (CAD) based research which can aid radiologists in medical diagnostics.
By Sidratul Montaha1 , *Sami Azam2 , A.K.M. Rakibul Haque Rafid3 , Sayma Islam4 , Pronab Ghosh5 , Mirjam jok6.
The complex feature characteristics and low contrast of cancer lesions, a high degree of inter-class resemblance between malignant and benign lesions, and the presence of various artifacts including hairs make automated melanoma recognition in dermoscopy images quite challenging. To date, various computer-aided solutions have been proposed to identify and classify skin cancer.
In this work, a deep learning model with a shallow architecture is proposed to classify the lesions into benign and malignant. To achieve effective training while limiting overfitting problems due to limited training data, image preprocessing and data augmentation processes are introduced. After this, the ‘box blur’ down-scaling method is employed, which adds efficiency to our study by reducing the overall training time and space complexity significantly.
Classify cardiovascular disease based on enhanced paper-based ECG image
Heart disease is a major cause of death that can be life-threatening if not identified and treated early. Electrocardiograms are important in classifying cardiovascular disease (CVD), and physicians often examine paper-based ECG images (2D) to diagnose cardiac disease.
An automated cardiovascular prognosis system can help accurately classify heart disease at an early stage. This study aims to classify cardiac disease in paper-based ECG images into five categories using deep learning methods with the highest possible accuracy and the lowest possible time complexity.
This study consists of two methods. Five deep learning models, InceptionV3, ResNet50, MobileNetV2, VGG19, and DenseNet201, were employed in the first method. In the second method, an Integrated Deep Learning Model (InRes-106) has been introduced, combining InceptionV3 and ResNet50 as a Deep Coevolutionary Neural Network (DCNN) capable of extracting hidden and high-level features from images.
The dataset contains a small amount of complicated ECG images, our proposed method, based on multiple image pre-processing methods, model fine-tuning, and ablation studies, can effectively diagnose high absolute cardiac disease.
Lung Disease Classification using Modified Compact Convolutional Transformer
Early identification and adequate treatment can assist to prevent lung and COVID disorders from becoming chronic, severe, and life-threatening, lowering the death rate by identifying at an early stage. Furthermore, X-ray images are commonly advised since the methods are less expensive, quicker, and expose patients to less radiation.
As a result, an automated and effective method including deep learning techniques might be a potential alternative for diagnosing lung disorders using X-ray images. The dataset used in this research which contains an uneven amount of images in different classes, which may result in poor model performance. Initially, the quantity of data in the classes is balanced by creating synthetic images based on the pattern and characteristics of the original images using Deep Convolutional
Generative Adversarial Network (DCGAN). Afterwards, un-wanted regions from x-ray images are removed, the brightness and contrast of the images are enhanced, and the chest abnormality are highlighted with different artifacts removal, noise reduction, and enhancement techniques.
Efficient diagnosis of cardiac disease using ensemble and traditional machine learning models on enhanced paper-based ECG images.
Heart disease is the most common serious illness that affects human health and is also a cause of mortality, so it has become a major threat to human health. This is why automated diagnosis methods are needed to identify heart diseases accurately and effectively. The electrocardiogram is mostly usable by the physician and the medical field to detect heart diseases. The essential goal of this work is to identify different cardiac diseases by classifying ECG images into five classes using the traditional and ensemble machine learning techniques with the highest accuracy.
This study experiments with paper-based ECG images based on segmentation, feature extraction, and feature selection as well. In addition, it continued into two stages; in the first stage, after feature extraction and feature selection, traditional model as Decision Tree, KNN, SVM, LightGBM, MLP, Naive Bayes, and Logistic Regression and ensemble model were applied. In the second stage, after performing the segmentation, feature extraction, and feature selection; again, run the tradition and the ensemble model. Before training the model, several image pre-processing techniques have been applied to remove the image patterns and enrich the image quality.