Search

Abstract Objective To conduct a systematic review of the computer vision applications that detect, diagnose, or analyse tuberculosis (TB) pathology or bacilli using digitised human lung tissue images either through automatic or semi-automatic methods. We categorised the computer vision platform into four technologies: image processing, object/pattern recognition, computer graphics, and deep learning. In this paper, the focus is on image processing and deep learning (DL) applications for either 2D or 3D digitised human lung tissue images. This review is useful for establishing a common practice in TB analysis using human lung tissue as well as identifying opportunities for further research in this space. The review brings attention to the state-of-art techniques for detecting TB, with emphasis on the challenges and limitations of the current techniques. The ultimate goal is to promote the development of more efficient and accurate algorithms for the detection or analysis of TB, and raise awareness about the importance of early detection. Design We searched five databases and Google Scholar for articles published between January 2017 and December 2022 that focus on Mycobacterium tuberculosis detection, or tuberculosis pathology using digitised human lung tissue images. Details regarding design, image processing and computer-aided techniques, deep learning models, and datasets were collected and summarised. Discussions, analysis, and comparisons of state-of-the-art methods are provided to help guide future research. Further, a brief update on the relevant techniques and their performance is provided. Results Several studies have been conducted to develop automated and AI-assisted methods for diagnosing Mtb and TB pathology from digitised human lung tissue images. Some studies presented a completely automated method of diagnosis, while other studies developed AI-assisted diagnostic methods. Low-level focus areas included the development of a novel $$\upmu$$ μ CT scanner for soft tissue image contract, and use of multiresolution computed tomography to analyse the 3D structure of the human lung. High-level focus areas included the investigation the effects of aging on the number and size of small airways in the lungs using CT and whole lung high-resolution $$\upmu$$ μ CT, and the 3D microanatomy characterisation of human tuberculosis lung using $$\upmu$$ μ CT in conjunction with histology and immunohistochemistry. Additionally, a novel method for acquiring high-resolution 3D images of human lung structure and topology is also presented. Conclusion The literature indicates that post 1950s, TB was predominantly studied using animal models even though no animal model reflects the full spectrum of human pulmonary TB disease and does not reproducibly transmit Mtb infection to other animals (Hunter, 2011). This explains why there are very few studies that used human lung tissue for detection or analysis of Mtb. Nonetheless, we found 10 studies that used human tissues (predominately lung) of which five studies proposed machine learning (ML) models for the detection of bacilli and the other five used CT on human lung tissue scanned ex-vivo.

Lung cancer is the leading cause of cancer deaths worldwide. It is a type of cancer that commonly remains undetected due to unpresented symptoms until it has progressed to later stages which motivates the requirement for accurate methods of early detection of lung nodules. Computer-aided diagnosis systems have adapted to aid in detecting and segmenting lung cancer, which can increase a patient's chance of survival. Automatic lung cancer detection and segmentation is a challenging task in aspects of segmentation accuracy. This study provides a comprehensive review of current methods and popular techniques which will aid in further research in lung tumor detection and segmentation. This study presents methods and techniques implemented to solve the challenges associated with lung cancer detection and segmentation and compares the approaches with each other. The methods used to evaluate these techniques and the accuracy rates are also discussed and compared to give insight for future research. Although several combination methods have been proposed over the past decade, an effective and efficient model still needs to be improvised for routine use.

Deep learning approaches, utilizing Bidirectional Encoder Representation from Transformers (BERT) and advanced fine-tuning techniques, have achieved state-of-the-art accuracies in the domain of term extraction from texts. However, BERT presents some limitations in that it primarily captures the semantic context relative to the surrounding text without considering how relevant or central a token is to the overall document content. There has also been research on the application of sequence labeling on contextualized embeddings; however, the existing methods often rely solely on local context for extracting key phrases from texts. To address these limitations, this study proposes a centrality-weighted BERT model for key phrase extraction from text using sequence labelling (CenBERT-SEQ). The proposed CenBERT-SEQ model utilizes BERT to represent terms with various contextual embedding architectures, and introduces a centrality-weighting layer that integrates document-level context into BERT. This layer leverages document embeddings to influence the importance of each term based on its relevance to the entire document. Finally, a linear classifier layer is employed to model the dependencies between the outputs, thereby enhancing the accuracy of the CenBERT-SEQ model. The proposed CenBERT-SEQ model was evaluated against the standard BERT base-uncased model using three Computer Science article datasets, namely, SemEval-2010, WWW, and KDD. The experimental results show that, although the CenBERT-SEQ and BERT-base models achieved higher and close comparable accuracy, the proposed CenBERT-SEQ model achieved higher precision, recall, and F1-score than the BERT-base model. Furthermore, a comparison of the proposed CenBERT-SEQ model to that of related studies revealed that the proposed CenBERT-SEQ model achieved a higher accuracy, precision, recall, and F1-score of 95%, 97%, 91%, and 94%, respectively, than related studies, showing the superior capabilities of the CenBERT-SEQ model in keyphrase extraction from scientific documents.

Forests play a pivotal role in mitigating climate change as well as contributing to the socio-economic activities of many countries. Therefore, it is of paramount importance to monitor forest cover. Traditional machine learning classifiers for segmenting images lack the ability to extract features such as the spatial relationship between pixels and texture, resulting in subpar segmentation results when used alone. To address this limitation, this study proposed a novel hybrid approach that combines deep neural networks and machine learning algorithms to segment an aerial satellite image into forest and non-forest regions. Aerial satellite forest image features were first extracted by two deep neural network models, namely, VGG16 and ResNet50. The resulting features are subsequently used by five machine learning classifiers including Random Forest (RF), Linear Support Vector Machines (LSVM), k-nearest neighbor (kNN), Linear Discriminant Analysis (LDA), and Gaussian Naive Bayes (GNB) to perform the final segmentation. The aerial satellite forest images were obtained from a deep globe challenge dataset. The performance of the proposed model was evaluated using metrics such as Accuracy, Jaccard score index, and Root Mean Square Error (RMSE). The experimental results revealed that the RF model achieved the best segmentation results with accuracy, Jaccard score, and RMSE of 94%, 0.913 and 0.245, respectively; followed by LSVM with accuracy, Jaccard score and RMSE of 89%, 0.876, 0.332, respectively. The LDA took the third position with accuracy, Jaccard score, and RMSE of 88%, 0.834, and 0.351, respectively, followed by GNB with accuracy, Jaccard score, and RMSE of 88%, 0.837, and 0.353, respectively. The kNN occupied the last position with accuracy, Jaccard score, and RMSE of 83%, 0.790, and 0.408, respectively. The experimental results also revealed that the proposed model has significantly improved the performance of the RF, LSVM, LDA, GNB and kNN models, compared to their performance when used to segment the images alone. Furthermore, the results showed that the proposed model outperformed other models from related studies, thereby, attesting its superior segmentation capability.

The advent of modern remote sensors alongside the development of advanced parallel computing has significantly transformed both the theoretical and real implementation aspects of remote sensing. Several algorithms for detecting objects of interest in remote sensing images and subsequent classification have been devised, and these include template matching based methods, machine learning and knowledge-based methods. Knowledge-driven approaches have received much attention from the remote sensing fraternity. They do, however, face challenges in terms of sensory gap, duality of expression, vagueness and ambiguity, geographic concepts expressed in multiple modes, and semantic gap. This paper aims to review and provide an up-to-date survey on machine learning and knowledge driven approaches towards remote sensing forest image analysis. It is envisaged that this work will assist researchers in coming up with efficient models that accurately detect and classify forest images. There is a mismatch between what domain experts expect from remote sensing data and what remote sensing science produces. Such a mismatch or disparity can be reduced or alleviated by adopting an ontology paradigm methodology. Ontologies should be used to support the future of remote sensing in forest object classification. The paper is presented in five parts: (1) a review of methods used for forest image detection and classification; (2) challenges faced by object detection methods; (3) analysis of segmentation techniques employed; (4) feature extraction and classification; and (5) performance of the state-of-the-art methods employed in forest image detection and classification.

Most existing approaches to image classification neglect the concept of semantics, resulting in two major shortcomings. Firstly, categories are treated as independent even when they have a strong semantic overlap. Secondly, the features used to classify images into different categories can be the same. It has been demonstrated that the integration of ontologies and semantic relationships greatly improves image classification accuracy. In this study, a hybrid ontological bagging algorithm and an ensemble technique of convolutional neural network (CNN) models have been developed to improve forest image classification accuracy. The ontological bagging approach learns discriminative weak attributes over multiple learning instances, and the bagging concept is adopted to minimize the error propagation of the classifiers. An ensemble of ResNet50, VGG16, and Xception models is used to generate a set of features for the classifiers trained through an ontology to perform the image classification process. To the authors’ best knowledge, there are no publicly available datasets for forest-type images; hence, the images used in this study were obtained from the internet. Obtained images were put into eight categories, namely: orchards, bare land, grassland, woodland, sea, buildings, shrubs, and logged forest. Each category comprised 100 images for training and 19 images for testing; thus, in total, the dataset contained 800 images for training and 152 images for testing. Our ensemble deep learning approach with an ontology model was successfully used to classify forest images into their respective categories. The classification was based on the semantic relationship between image categories. The experimental results show that our proposed model with ontology outperformed other baseline classifiers without ontology with 96% accuracy and the lowest root-mean-square error (RMSE) of 0.532 compared to 88.8%, 86.2%, 81.6%, 64.5%, and 63.8% accuracy and 1.048, 1.094, 1.530, 1.678, and 2.090 RMSE for support-vector machines, random forest, k-nearest neighbours, Gaussian naive Bayes, and decision trees, respectively.