Globally, diabetes affects 537 million people, making it the deadliest and the most common non‐communicable disease. Many factors can cause a person to get affected by diabetes, like excessive body weight, abnormal cholesterol level, family history, physical inactivity, bad food habit etc. Increased urination is one of the most common symptoms of this disease. People with diabetes for a long time can get several complications like heart disorder, kidney disease, nerve damage, diabetic retinopathy etc. But its risk can be reduced if it is predicted early. In this paper, an automatic diabetes prediction system has been developed using a private dataset of female patients in Bangladesh and various machine learning techniques. The authors used the Pima Indian diabetes dataset and collected additional samples from 203 individuals from a local textile factory in Bangladesh. Feature selection algorithm mutual information has been applied in this work. A semi‐supervised model with extreme gradient boosting has been utilized to predict the insulin features of the private dataset. SMOTE and ADASYN approaches have been employed to manage the class imbalance problem. The authors used machine learning classification methods, that is, decision tree, SVM, Random Forest, Logistic Regression, KNN, and various ensemble techniques, to determine which algorithm produces the best prediction results. After training on and testing all the classification models, the proposed system provided the best result in the XGBoost classifier with the ADASYN approach with 81% accuracy, 0.81 F1 coefficient and AUC of 0.84. Furthermore, the domain adaptation method has been implemented to demonstrate the versatility of the proposed system. The explainable AI approach with LIME and SHAP frameworks is implemented to understand how the model predicts the final results. Finally, a website framework and an Android smartphone application have been developed to input various features and predict diabetes instantaneously. The private dataset of female Bangladeshi patients and programming codes are available at the following link: https://github.com/tansin-nabil/Diabetes-Prediction-Using-Machine-Learning .

Glaucoma is an irreversible neurodegenerative disease, where intraocular hypertension is developed due to the increased aqueous humor and blockage of the drainage system between the iris and cornea. As a result, the optic nerve head, which sends visual stimulus from our eyes to the brain, is damaged, causing visual field loss and ultimately blindness. Glaucoma is considered as the sneak thief of vision because it is difficult to diagnose early, and its regular screening is highly recommended to distinguish the neurological disorder. The detection of glaucoma is costly and time-consuming and not only there always remains a good possibility of human error but also this detection method is dependent upon the availability of the resources (experienced ophthalmologists and expensive instruments). In this work, an automatic glaucoma classification technique has been developed by utilizing multiple deep learning approaches. First, a new private dataset of 634 color fundus images has been collected and annotated by two eye specialists, a pediatric ophthalmologist and a glaucoma and refractive surgeon, from Bangladesh Eye Hospital, Bangladesh. Next, various deep learning models (EfficientNet, MobileNet, DenseNet, and GoogLeNet) have been used to detect glaucoma from fundus images. The model with EfficientNet-b3 architecture achieved the best results with test accuracy, F1-score, and ROC AUC of 0.9652, 0.9512, and 0.9574, respectively, for the cropped optic cup and disc fundus photographs. We also constructed a new dataset by segmenting the blood vessels from retinal fundus images employing a U-net model trained on High-Resolution Fundus Image Database. The MobileNet v3 model trained on this dataset achieved a satisfactory test accuracy of 0.8348 and an F1-score of 0.7957. This impressive result suggests that blood vessel segmentation of fundus images can be utilized as an alternative to detect glaucoma automatically.

Object detection, one of the most significant contributions of computer vision and machine learning, plays an immense role in identifying and locating objects in an image or a video. We recognize distinct objects and precisely get their information through object detection, such as their size, shape, and location. This paper developed a low-cost assistive system of obstacle detection and the surrounding environment depiction to help blind people using deep learning techniques. TensorFlow object detection API and SSDLite MobileNetV2 have been used to create the proposed object detection model. The pre-trained SSDLite MobileNetV2 model is trained on the COCO dataset, with almost 328,000 images of 90 different objects. The gradient particle swarm optimization (PSO) technique has been used in this work to optimize the final layers and their corresponding hyperparameters of the MobileNetV2 model. Next, we used the Google text-to-speech module, PyAudio, playsound, and speech recognition to generate the audio feedback of the detected objects. A Raspberry Pi camera captures real-time video where real-time object detection is done frame by frame with Raspberry Pi 4B microcontroller. The proposed device is integrated into a head cap, which will help visually impaired people to detect obstacles in their path, as it is more efficient than a traditional white cane. Apart from this detection model, we trained a secondary computer vision model and named it the “ambiance mode.” In this mode, the last three convolutional layers of SSDLite MobileNetV2 are trained through transfer learning on a weather dataset. The dataset comprises around 500 images from four classes: cloudy, rainy, foggy, and sunrise. In this mode, the proposed system will narrate the surrounding scene elaborately, almost like a human describing a landscape or a beautiful sunset to a visually impaired person. The performance of the object detection and ambiance description modes are tested and evaluated in a desktop computer and Raspberry Pi embedded system. Detection accuracy and mean average precision, frame rate, confusion matrix, and ROC curve measure the model's accuracy on both setups. This low-cost proposed system is believed to help visually impaired people in their day-to-day life.

The coronavirus is considered this century's most disruptive catastrophe and global concern. This disease has prompted extreme social, psychological and economic impacts affecting millions of people around the globe. COVID-19 is transmitted from one infected person's body to another through respiratory droplets. This virus proliferates when people breathe in air-contaminated space with droplets and microscopic airborne particles. This research aims to analyze automatic COVID-19 detection using machine learning techniques to build an intelligent web application. The dataset has been preprocessed by dropping null values, feature engineering, and synthetic oversampling (SMOTE) techniques. Next, we trained and evaluated different classifiers, i.e., logistic regression, random forest, decision tree, k-nearest neighbor, support vector machine (SVM), ensemble models (adaptive boosting and extreme gradient boosting) and deep learning (artificial neural network, convolutional neural network and long short-term memory) techniques. Explainable AI with the LIME framework has been applied to interpret the prediction results. The hybrid CNN-LSTM algorithm with the SMOTE approach performed better than the other models on the employed open-source dataset obtained from the Israeli Ministry of Health website, with 96.34% accuracy and a 0.98 F1 score. Finally, this model was chosen to deploy the proposed prediction system to a website, where users may acquire an instantaneous COVID-19 prognosis based on their symptoms.