TOPSIS aided ensemble of CNN models for screening COVID-19 in chest X-ray images

Fusing global context with multiscale context for enhanced breast cancer classification

Breast cancer is the second most common type of cancer among women. Prompt detection of breast cancer can impede its advancement to more advanced phases, thereby elevating the probability of favorable treatment consequences. Histopathological images are commonly used for breast cancer classification due to their detailed cellular information. Existing diagnostic approaches rely on Convolutional Neural Networks (CNNs) which are limited to local context resulting in a lower classification accuracy. Therefore, we present a fusion model composed of a Vision Transformer (ViT) and custom Atrous Spatial Pyramid Pooling (ASPP) network with an attention mechanism for effectively classifying breast cancer from histopathological images. ViT enables the model to attain global features, while the ASPP network accommodates multiscale features. Fusing the features derived from the models resulted in a robust breast cancer classifier. With the help of five-stage image preprocessing technique, the proposed model achieved 100% accuracy in classifying breast cancer on the BreakHis dataset at 100X and 400X magnification factors. On 40X and 200X magnifications, the model achieved 99.25% and 98.26% classification accuracy respectively. With a commendable classification efficacy on histopathological images, the model can be considered a dependable option for proficient breast cancer classification.

Amino acid metabolomics and machine learning-driven assessment of future liver remnant growth after hepatectomy in livers of various backgrounds

Accurate assessment of future liver remnant growth after partial hepatectomy (PH) in patients with different liver backgrounds is a pressing clinical issue. Amino acid (AA) metabolism plays a crucial role in liver regeneration. In this study, we combined metabolomics and machine learning (ML) to develop a generalized future liver remnant assessment model for multiple liver backgrounds. The liver index was calculated at 0, 6, 24, 48, 72 and 168 h after 70 % PH in healthy mice and mice with nonalcoholic steatohepatitis or liver fibrosis. The serum levels of 39 amino acids (AAs) were measured using UPLC-MS/MS. The dataset was randomly divided into training and testing sets at a 2:1 ratio, and orthogonal partial least squares regression (OPLS) and minimally biased variable selection in R (MUVR) were used to select a metabolite signature of AAs. To assess liver remnant growth, nine ML models were built, and evaluated using the coefficient of determination (R2), mean absolute error (MAE), and root mean square error (RMSE). The post-Pareto technique for order preference by similarity to the ideal solution (TOPSIS) was employed for ranking the ML algorithms, and a stacking technique was utilized to establish consensus among the superior algorithms. Compared with those of OPLS, the signature AAs set identified by MUVR (Thr, Arg, EtN, Phe, Asa, 3MHis, Abu, Asp, Tyr, Leu, Ser, and bAib) are more concise. Post-Pareto TOPSIS ranking demonstrated that the majority of ML algorithm in combinations with MUVR outperformed those with OPLS. The established SVM-KNN consensus model performed best, with an R2 of 0.79, an MAE of 0.0029, and an RMSE of 0.0035 for the testing set. This study identified a metabolite signature of 12 AAs and constructed an SVM-KNN consensus model to assess future liver remnant growth after PH in mice with different liver backgrounds. Our preclinical study is anticipated to establish an alternative and generalized assessment method for liver regeneration.

GUNet++: guided-U-Net-based compact image representation with an improved reconstruction mechanism

The invention of microscopy- and nanoscopy-based imaging technology opened up different research directions in life science. However, these technologies create the need for larger storage space, which has negative impacts on the environment. This scenario creates the need for storing such images in a memory-efficient way. Compact image representation (CIR) can solve the issue as it targets storing images in a memory-efficient way. Thus, in this work, we have designed a deep-learning-based CIR technique that selects key pixels using the guided U-Net (GU-Net) architecture [Asian Conference on Pattern Recognition, p. 317 (2023)], and then near-original images are constructed using a conditional generative adversarial network (GAN)-based architecture. The technique was evaluated on two microscopy- and two scanner-captured-image datasets and obtained good performance in terms of storage requirements and quality of the reconstructed images.

PT-Finder: A multi-modal neural network approach to target identification

Efficient target identification for bioactive compounds, including novel synthetic analogs, is crucial for accelerating the drug discovery pipeline. However, the process of target identification presents significant challenges and is often expensive, which in turn can hinder the drug discovery efforts. To address these challenges machine learning applications have arisen as a promising approach for predicting the targets for novel chemical compounds. These methods allow the exploration of ligand-target interactions, uncovering of biochemical mechanisms, and the investigation of drug repurposing. Typically, the current target identification tools rely on assessing ligand structural similarities. Herein, a multi-modal neural network model was built using a library of proteins, their respective sequences, and active inhibitors. Subsequent validations showed the model to possess accuracy of 82 % and MPRAUC of 0.80. Leveraging the trained model, we developed PT-Finder (Protein Target Finder), a user-friendly offline application that is capable of predicting the target proteins for hundreds of compounds within a few seconds. This combination of offline operation, speed, and accuracy positions PT-Finder as a powerful tool to accelerate drug discovery workflows. PT-Finder and its source codes have been made freely accessible for download at https://github.com/PT-Finder/PT-Finder.

MENet: A Mitscherlich function based ensemble of CNN models to classify lung cancer using CT scans

Lung cancer is one of the leading causes of cancer-related deaths worldwide. To reduce the mortality rate, early detection and proper treatment should be ensured. Computer-aided diagnosis methods analyze different modalities of medical images to increase diagnostic precision. In this paper, we propose an ensemble model, called the Mitscherlich function-based Ensemble Network (MENet), which combines the prediction probabilities obtained from three deep learning models, namely Xception, InceptionResNetV2, and MobileNetV2, to improve the accuracy of a lung cancer prediction model. The ensemble approach is based on the Mitscherlich function, which produces a fuzzy rank to combine the outputs of the said base classifiers. The proposed method is trained and tested on the two publicly available lung cancer datasets, namely Iraq-Oncology Teaching Hospital/National Center for Cancer Diseases (IQ-OTH/NCCD) and LIDC-IDRI, both of these are computed tomography (CT) scan datasets. The obtained results in terms of some standard metrics show that the proposed method performs better than state-of-the-art methods. The codes for the proposed work are available at https://github.com/SuryaMajumder/MENet.

Generalizable disease detection using model ensemble on chest X-ray images

In the realm of healthcare, the demand for swift and precise diagnostic tools has been steadily increasing. This study delves into a comprehensive performance analysis of three pre-trained convolutional neural network (CNN) architectures: ResNet50, DenseNet121, and Inception-ResNet-v2. To ensure the broad applicability of our approach, we curated a large-scale dataset comprising a diverse collection of chest X-ray images, that included both positive and negative cases of COVID-19. The models' performance was evaluated using separate datasets for internal validation (from the same source as the training images) and external validation (from different sources). Our examination uncovered a significant drop in network efficacy, registering a 10.66% reduction for ResNet50, a 36.33% decline for DenseNet121, and a 19.55% decrease for Inception-ResNet-v2 in terms of accuracy. Best results were obtained with DenseNet121 achieving the highest accuracy at 96.71% in internal validation and Inception-ResNet-v2 attaining 76.70% accuracy in external validation. Furthermore, we introduced a model ensemble approach aimed at improving network performance when making inferences on images from diverse sources beyond their training data. The proposed method uses uncertainty-based weighting by calculating the entropy in order to assign appropriate weights to the outputs of each network. Our results showcase the effectiveness of the ensemble method in enhancing accuracy up to 97.38% for internal validation and 81.18% for external validation, while maintaining a balanced ability to detect both positive and negative cases.

DBU-Net: Dual branch U-Net for tumor segmentation in breast ultrasound images

Breast ultrasound medical images often have low imaging quality along with unclear target boundaries. These issues make it challenging for physicians to accurately identify and outline tumors when diagnosing patients. Since precise segmentation is crucial for diagnosis, there is a strong need for an automated method to enhance the segmentation accuracy, which can serve as a technical aid in diagnosis. Recently, the U-Net and its variants have shown great success in medical image segmentation. In this study, drawing inspiration from the U-Net concept, we propose a new variant of the U-Net architecture, called DBU-Net, for tumor segmentation in breast ultrasound images. To enhance the feature extraction capabilities of the encoder, we introduce a novel approach involving the utilization of two distinct encoding paths. In the first path, the original image is employed, while in the second path, we use an image created using the Roberts edge filter, in which edges are highlighted. This dual branch encoding strategy helps to extract the semantic rich information through a mutually informative learning process. At each level of the encoder, both branches independently undergo two convolutional layers followed by a pooling layer. To facilitate cross learning between the branches, a weighted addition scheme is implemented. These weights are dynamically learned by considering the gradient with respect to the loss function. We evaluate the performance of our proposed DBU-Net model on two datasets, namely BUSI and UDIAT, and our experimental results demonstrate superior performance compared to state-of-the-art models.

Microstructural segmentation using a union of attention guided U-Net models with different color transformed images

Metallographic images or often called the microstructures contain important information about metals, such as strength, toughness, ductility, corrosion resistance, which are used to choose the proper materials for various engineering applications. Thus by understanding the microstructures, one can determine the behaviour of a component made of a particular metal, and can predict the failure of that component in certain conditions. Image segmentation is a powerful technique for determination of morphological features of the microstructure like volume fraction, inclusion morphology, void, and crystal orientations. These are some key factors for determining the physical properties of metal. Therefore, automatic micro-structure characterization using image processing is useful for industrial applications which currently adopts deep learning-based segmentation models. In this paper, we propose a metallographic image segmentation method using an ensemble of modified U-Nets. Three U-Net models having the same architecture are separately fed with color transformed imaged (RGB, HSV and YUV). We improvise the U-Net with dilated convolutions and attention mechanisms to get finer grained features. Then we apply the sum-rule-based ensemble method on the outcomes of U-Net models to get the final prediction mask. We achieve the mean intersection over union (IoU) score of 0.677 on a publicly available standard dataset, namely MetalDAM. We also show that the proposed method obtains results comparable to state-of-the-art methods with fewer number of model parameters. The source code of the proposed work can be found at  https://github.com/mb16biswas/attention-unet .

Monkeypox detection from skin lesion images using an amalgamation of CNN models aided with Beta function-based normalization scheme

We have recently been witnessing that our society is starting to heal from the impacts of COVID-19. The economic, social and cultural impacts of a pandemic cannot be ignored and we should be properly equipped to deal with similar situations in future. Recently, Monkeypox has been concerning the international health community with its lethal impacts for a probable pandemic. In such situations, having appropriate protocols and methodologies to deal with the outbreak efficiently is of paramount interest to the world. Early diagnosis and treatment stand as the only viable option to tackle such problems. To this end, in this paper, we propose an ensemble learning-based framework to detect the presence of the Monkeypox virus from skin lesion images. We first consider three pre-trained base learners, namely Inception V3, Xception and DenseNet169 to fine-tune on a target Monkeypox dataset. Further, we extract probabilities from these deep models to feed into the ensemble framework. To combine the outcomes, we propose a Beta function-based normalization scheme of probabilities to learn an efficient aggregation of complementary information obtained from the base learners followed by the sum rule-based ensemble. The framework is extensively evaluated on a publicly available Monkeypox skin lesion dataset using a five-fold cross-validation setup to evaluate its effectiveness. The model achieves an average of 93.39%, 88.91%, 96.78% and 92.35% accuracy, precision, recall and F1 scores, respectively. The supporting source codes are presented in https://github.com/BihanBanerjee/MonkeyPox.