Rapid Classification Research Articles

Rapid large-scale building damage level classification after earthquakes using deep learning with Lidar and satellite optical data

ABSTRACT In post-earthquake scenarios, the swift assessment of building damage levels is pivotal for efficient emergency response and recovery planning. Nevertheless, conventional in-situ damage evaluations consume time. Current satellite-based deep learning methods save time but often lack detail, usually classifying damage as either collapsed or intact. This two-level information is not enough for rescue or recovery planning. Light Detection and Ranging (Lidar)-based deep learning methods, which provide three-dimensional (3D) information, could address this issue of damage details. Therefore, this paper proposes a deep learning-based building damage level classification method using both Lidar and satellite data. The proposed method classifies damage into four levels, including no/minor damage, partially collapsed, totally collapsed, and story failure. The developed network builds upon RandLA-Net, incorporating surface normal vectors to enhance accuracy. A colourised Lidar dataset was created for the network. The network underscores the advantage of incorporating surface normal information. A framework is also proposed based on the damage level outcomes of the developed network, which aids in emergency response efforts. Consequently, this paper demonstrates the practical utility of deep learning networks in rapidly assessing detailed building damage levels after earthquakes. Its practical contribution is guiding decision-making during the critical phases of post-earthquake response and recovery.

Investigations on the Performance of a 5 mm CdTe Timepix3 Detector for Compton Imaging Applications

Nuclear power plant decommissioning requires the rapid and accurate classification of radioactive waste in narrow spaces and under time constraints. Photon-counting detector technology offers an effective solution for the quick classification and detection of radioactive hotspots in a decommissioning environment. This paper characterizes a 5 mm CdTe Timepix3 detector and evaluates its feasibility as a single-layer Compton camera. The sensor’s electron mobility–lifetime product and resistivity are studied across bias voltages ranging from −100 V to −3000 V, obtaining values of μeτe = (1.2 ± 0.1) × 10−3 cm2V−1, and two linear regions with resistivities of ρI=(5.8±0.2) GΩ cm and ρII=(4.1±0.1) GΩ cm. Additionally, two calibration methodologies are assessed to determine the most suitable for Compton applications, achieving an energy resolution of 16.3 keV for the 137Cs photopeak. The electron’s drift time in the sensor is estimated to be (122.3 ± 7.4) ns using cosmic muons. Finally, a Compton reconstruction of two simultaneous point-like sources is performed, demonstrating the detector’s capability to accurately locate radiation hotspots with a ∼51 cm resolution.

A self-supervised deep-driven model for automatic weather classification from remote sensing images

ABSTRACT The increasing use of Remote Sensing (RS) technology has resulted in the availability of large volumes of satellite imaging data, necessitating efficient, and scalable solutions for rapid analysis and classification in various interdisciplinary fields. This research focuses on addressing these challenges by examining different training and optimization techniques to achieve the highest classification accuracy on both labelled and unlabelled data from the Large-Scale Cloud Photos Dataset for Meteorology Research (LSCIDMR), which contains 10 high-resolution classes. A key aspect of this study is the implementation of a unique deep learning model based on the ResNet101 architecture, designed to handle the inter-class similarity problem through architectural improvements and imbalanced class distributions using data augmentation techniques. Furthermore, this model integrates a novel approach that combines self-learning with semi-supervised learning to process large amounts of unlabelled data. An iterative self-learning process is employed for continuous refinement, using pseudo-labels generated from the unlabelled data. To ensure higher quality and relevance, confidence-based selection of pseudo-labels is applied. Our model demonstrates superior performance in classification compared to recently reported deep learning algorithms, offering a robust solution for the automated categorization of satellite images. In this study, our ResNet101 model achieved a mean accuracy of 0.917 (±0.02), a mean precision of 0.917 (±0.01), with a mean F1 score of 0.9300 (±0.015) over five folds.

Neoadjuvant therapy for resectable non-small cell lung cancer

Treatment perspectives for non-small cell lung cancer (NSCLC) have been significantly expanded by the integration of immune checkpoint inhibitors into multimodal therapy concepts. Currently, combined, immune checkpoint-inhibitor-based therapy concepts are also advancing into early, resectable stages of NSCLC. Neoadjuvant and perioperative chemoimmunotherapy opened up a promising new preoperative treatment approach, but also raises some new questions and challenges. With the expanded perioperative treatment options and the perspective on a further improvement in the absence of recurrence after tumor resection, there is push towards comprehensively collecting therapy-relevant findings for imaging, molecular and histopathological diagnostics at an early stage. All patients with lung carcinoma, regardless of the therapy intention, should be presented to an interdisciplinary tumor board with thoracic oncological expertise. This is regularly given in certified lung cancer centers.A standardized procedure contributes to optimized pre-therapeutic diagnostics and facilitates coordination for the best possible multimodal approach in the interdisciplinary tumor board. In the case of centrally located resectable tumors, for example, neoadjuvant treatment increases the chances of a procedure that is as parenchymal sparing as possible. Some questions cannot yet be answered conclusively. Perioperative systemic therapy with molecular-targeted and immune checkpoint inhibitors is the subject of numerous ongoing studies. The considerable dynamics in newly approved therapies and the development of perioperative therapy concepts require continuous adaptation of diagnostic algorithms and standards. Integration into standard pre-surgical routine makes rapid classification of the relevant findings as well as close coordination between the diagnostic and interventional disciplines essential.

Geographical classification of population: Analysis of amino acid in fingermark residues using UHPLC-QQQ-MS/MS combined with machine learning

ObjectiveTo determine the living regions of individuals based on amino acids in fingermark residues and to establish a rapid and accurate regional classification method using machine learning. Methods: A total of 71 fingermark donors from six different provinces in various regions of China were selected. The content of 18 amino acids in their fingermarks was detected using UHPLC-QQQ-MS/MS. Classification models were established using various machine learning algorithms, and the cross-validation accuracy of 72 combinations, including feature engineering, classification algorithms, and optimization algorithms, was compared. Results: UHPLC-QQQ-MS/MS successfully quantified 16 amino acids. Significant differences in the relative content of amino acids were found between the fingermarks from the eastern and western regions of China, as well as among neighboring provinces. The combination of SFS+SVM+BO was identified as the optimal classification model, achieving an accuracy of 90.14%. Conclusion: The study found regional differences in the relative content of amino acids in fingermarks and established a regional classification method combining UHPLC-QQQ-MS/MS and machine learning. The method developed in this study can be applied to incomplete or distorted fingermarks, and the experimental results can be directly used in police investigations. This research uncovers the multidimensional information carried by fingerprint substances, demonstrating innovation and application value. It not only saves and shortens investigation time and provides investigative leads, but also enables previously unusable physical evidence to play a role again, enhancing the profiling of suspects.

Rapid and non-destructive classification of salinity levels in brined kimchi cabbage using hyperspectral imaging

In this study, we demonstrate the potential of a non-destructive hyperspectral imaging processing method in the near-infrared (NIR) region (874–1734 nm) for classifying the quality of brined kimchi cabbage. The salinity level of brined kimchi cabbage is an important indicator of consumer preference and the quality of kimchi. Hence, we compared the water content and salinity of brined kimchi cabbage via hyperspectral data. We extracted the optimal wavelengths from the hyperspectral image dataset to classify the salinity level of the predicted brined kimchi cabbage, and thus, established a novel approach for classifying kimchi samples into quality-unacceptable and quality-acceptable groups. Standard normal variate and multiplicative scatter correction (MSC) were used for pathlength correction. The Savitzky–Golay first and second derivatives were used for the deconvolution of the raw spectral data. The experimental results confirmed that the decision tree model combined with MSC pathlength correction and Savitzky–Golay first derivative preprocessing was the best classification model. The proposed hyperspectral image–NIR system can be applied to the detection of salinity during industrial kimchi manufacturing.

A Review of Convolutional Neural Networks in Cancer Image Classification

Abstract. Cancer image analysis is a crucial element in the field of modern oncology, offering insights into tumor characteristics that are vital for accurate diagnosis and treatment planning. Conventional diagnostic techniques are frequently constrained by discrepancies in interpretation and delays in processing. In recent years, deep learning techniques, in particular convolutional neural networks (CNNs), have demonstrated remarkable capabilities in processing large-scale medical images, enabling the rapid detection and classification of cancerous tissue with high accuracy and precision. This paper reviews the application of CNNs in cancer image classification, introduces common CNN models, and discusses advances such as transfer learning and multimodal data fusion. Furthermore, this paper examines the current challenges confronting convolutional neural networks (CNNs) in the domain of cancer classification. It proposes promising avenues for future research and underscores the potential of CNNs in enhancing cancer diagnosis and even revolutionizing the healthcare system.

Dual-task vision transformer for rapid and accurate intracerebral hemorrhage CT image classification

Intracerebral hemorrhage (ICH) is a severe and sudden medical condition caused by the rupture of blood vessels in the brain, leading to permanent damage to brain tissue and often resulting in functional disabilities or death in patients. Diagnosis and analysis of ICH typically rely on brain CT imaging. Given the urgency of ICH conditions, early treatment is crucial, necessitating rapid analysis of CT images to formulate tailored treatment plans. However, the complexity of ICH CT images and the frequent scarcity of specialist radiologists pose significant challenges. Therefore, we collect a dataset from the real world for ICH and normal classification and three types of ICH image classification based on the hemorrhage location, i.e., Deep, Subcortical, and Lobar. In addition, we propose a neural network structure, dual-task vision transformer (DTViT), for the automated classification and diagnosis of ICH images. The DTViT deploys the encoder from the Vision Transformer (ViT), employing attention mechanisms for feature extraction from CT images. The proposed DTViT framework also incorporates two multilayer perception (MLP)-based decoders to simultaneously identify the presence of ICH and classify the three types of hemorrhage locations. Experimental results demonstrate that DTViT performs well on the real-world test dataset.

Classification and Identification of Foodborne Bacteria in Beef by Utilising Surface-Enhanced Raman Spectroscopy Coupled with Chemometric Methods.

The detection and classification of foodborne pathogenic bacteria is crucial for food safety monitoring, consequently requiring rapid, accurate and sensitive methods. In this study, the surface-enhanced Raman spectroscopy (SERS) technique coupled with chemometrics methods was used to detect and classify six kinds of foodborne pathogenic bacteria, including Salmonella typhimurium (S. typhimurium), Escherichia coli (E. coli) O157:H7, Staphylococcus aureus (S. aureus), Listeria monocytogenes (L. monocytogenes), Listeria innocua (L. innocua), and Listeria welshimeri (L. welshimeri). First, silver nanoparticles (AgNPs) with different particle sizes were prepared as SERS-enhanced substrates by changing the concentration of sodium citrate, and the volume ratio of silver nanosol to bacterial solution was optimised to obtain the optimal SERS signal. Then, principal component analysis (PCA) and hierarchical cluster analysis (HCA) were used to classify the SERS spectra of six bacteria at three classification levels (Gram type level, genus level and species level), and appropriate classification models were established. Finally, these models were validated on 540 spectra using linear discriminant analysis (LDA), achieving an average accuracy of 95.65%. Overall, it was concluded that the SERS technique combined with chemometrics methods could achieve the rapid detection and classification identification of foodborne pathogenic bacteria, providing an effective means for food safety monitoring.

STRPsearch: fast detection of structured tandem repeat proteins.

Structured Tandem Repeats Proteins (STRPs) constitute a subclass of tandem repeats characterized by repetitive structural motifs. These proteins exhibit distinct secondary structures that form repetitive tertiary arrangements, often resulting in large molecular assemblies. Despite highly variable sequences, STRPs can perform important and diverse biological functions, maintaining a consistent structure with a variable number of repeat units. With the advent of protein structure prediction methods, millions of 3D-models of proteins are now publicly available. However, automatic detection of STRPs remains challenging with current state-of-the-art tools due to their lack of accuracy and long execution times, hindering their application on large datasets. In most cases, manual curation remains the most accurate method for detecting and classifying STRPs, making it impracticable to annotate millions of structures. We introduce STRPsearch, a novel tool for the rapid identification, classification, and mapping of STRPs. Leveraging manually curated entries from RepeatsDB as the known conformational space of STRPs, STRPsearch employs the latest advances in structural alignment for a fast and accurate detection of repeated structural motifs in proteins, followed by an innovative approach to map units and insertions through the generation of TM-score profiles. STRPsearch is highly scalable, efficiently processing large datasets, and can be applied to both experimental structures and predicted models. Additionally, it demonstrates superior performance compared to existing tools, offering researchers a reliable and comprehensive solution for STRP analysis across diverse proteomes. STRPsearch is coded in Python. All scripts and associated documentation are available from: https://github.com/BioComputingUP/STRPsearch. Supplementary data are available at Bioinformatics online.

Correction to: Rapid source classification and distance estimation for compact binary mergers with PyCBC live

Correction to: Rapid source classification and distance estimation for compact binary mergers with PyCBC live

A Rapid Classification of Cross-Contaminations in Aviation Oil Using Impedance-Driven Supervised Machine Learning

A Rapid Classification of Cross-Contaminations in Aviation Oil Using Impedance-Driven Supervised Machine Learning

CBMAFF-Net: An Intelligent NMR-Based Nontargeted Screening Method for New Psychoactive Substances.

With the proliferation and rapid evolution of new psychoactive substances (NPSs), traditional database-based search methods face increasing challenges in identifying NPS seizures with complex compositions, thereby complicating their regulation and early warning. To address this issue, CBMAFF-Net (CNN BiLSTM Multistep Attentional Feature Fusion Network) is proposed as an intelligent screening method to rapidly classify unknown confiscated substances using 13C nuclear magnetic resonance (NMR) and 1H NMR data. Initially, we utilize the synergy of a convolutional neural network (CNN) and bidirectional long short-term memory network (BiLSTM) to extract the global and local features of the NMR data. These features are sequentially fused through a weighted approach guided by an attention mechanism, thoroughly capturing the essential NPS information. We evaluated the model on a generated simulated data set, where it performed with 99.8% accuracy and a 99.8% F1 score. Additionally, testing on 42 actual seizure cases yielded a recognition accuracy of 97.6%, significantly surpassing the performance of conventional database-based similarity search algorithms. These findings suggest that the proposed method holds substantial promise for the rapid screening and classification of NPSs.

Automatic assessment of nailfold capillaroscopy software: a pilot study.

Capillaroscopy is a simple method of nailfold capillary imaging, used to diagnose diseases from the systemic sclerosis spectrum. However, the assessment of the capillary image is time-consuming and subjective. This makes it difficult to use for a detailed comparison of studies assessed by various physicians. This pilot study aimed to validate software used for automatic capillary counting and image classification as normal or pathological. The study was based on the assessment of 200 capillaroscopic images obtained from patients suffering from systemic sclerosis or scleroderma spectrum diseases and healthy people. Dinolite MEDL4N Pro was used to perform capillaroscopy. Each image was analysed manually and described using working software. The neural network was trained using the fast.ai library (based on PyTorch). The ResNet-34 deep residual neural network was chosen; 10-fold cross-validation with the validation and test set was performed, using the Darknet-YoloV3 state of the art neural network in a GPU-optimized (P5000 GPU) environment. For the calculation of 1 mm capillaries, an additional detection mechanism was designed. The results obtained under neural network training were compared to the results obtained in manual analysis. The sensitivity of the automatic tool relative to manual assessment in classification of correct vs. pathological images was 89.0%, specificity 89.4% for the training group, in validation 89.0% and 86.9% respectively. For the average number of capillaries in 1 mm the precision of real images detected within the region of interest was 96.48%. The pilot software for fully automatic capillaroscopic image assessment can be a useful tool for the rapid classification of a normal and altered capillaroscopy pattern. In addition, it allows one to quickly calculate the number of capillaries. In the future, the tool will be developed and will make it possible to obtain full imaging characteristics independent of the experience of the examiner.

Rapid Raman spectroscopy analysis assisted with machine learning: a case study on Radix Bupleuri.

Radix Bupleuri has been widely used for its plentiful pharmacological effects. But it is hard to evaluate their safety and efficacy because the concentrations of components are tightly affected by the surrounding environment. Thus, Radix Bupleuri samples from different regions and varieties were collected. Based on the experimental and computational Raman spectrum, machine learning is emphasized for certain obscured characteristics; for example, linear discriminant analysis (LDA), support vector machine (SVM), eXtreme gradient boosting (XGBoost) and light gradient boosting machine (LightGBM). After dimension reduction by LDA, models of SVM, XGBoost and LightGBM were trained for classification and regression prediction of Bupleurum production regions. Support vector classifiers achieved the best accuracy of 98% and an F1 score above 0.96 on the test set. Support vector regression has a good fitting performance with an R2 score above 0.90 and a relatively low mean square error. However, complex models were prone to overfitting, resulting in poor generalization ability. Among these machine learning models, the typical LDA-SVM models, consistent with the high-performance liquid chromatography results, demonstrate great performance and stability. We envision that this rapid classification and regression technique can be extended to predictions for other herbs. © 2024 Society of Chemical Industry.

Combining microscale ATR-FTIR and chemometrics to interpret degradation characteristics of earlywood, latewood, and compression wood in waterlogged archaeological pine wood

Interpreting the degradation characteristics of waterlogged archaeological wood (WAW) is crucial for the conservation of wooden cultural heritage. Generally, multidisciplinary diagnostic methods, including physical, micromorphological, and chemical approaches, are employed to evaluate the preservation state of WAW. However, primarily focused on the sample level, this methodology limits the understanding of the variability in degradation from a detailed perspective. In this paper, we adopted the in-situ microscale attenuated total reflectance Fourier transform infrared (ATR-FTIR) method to investigate the degradation variability in waterlogged archaeological Masson pine (Pinus massoniana) wood excavated from the ancient Chinese shipwreck Nanhai No. 1. Specifically, spectra of earlywood (EW), latewood (LW), and compression wood (CW) were extracted and combined with chemometrics to achieve rapid classification of their degradation levels. The micromorphological features of wood cell walls in conjunction with the ratios of lignin (A1509) and carbohydrate (A1370) peak areas were used to estimate the degradation levels. Unlike recent wood, moderate degradation in CW and severe degradation in EW and LW were classified in archaeological samples. The degradation levels were effectively determined through principal component analysis (PCA) and sparse partial least squares discriminant analysis (sPLSDA). The results suggest that chemometric analysis is a promising method to discern the variable degradation levels of archaeological wood at the tissue level. The methodologies developed in this study provide detailed insights into the degradation characteristics in WAW and improve the accuracy of evaluating the preservation state.

Rapid Classification of Sugarcane Nodes and Internodes Using Near-Infrared Spectroscopy and Machine Learning Techniques.

Accurate and rapid discrimination between nodes and internodes in sugarcane is vital for automating planting processes, particularly for minimizing bud damage and optimizing planting material quality. This study investigates the potential of visible-shortwave near-infrared (Vis-SWNIR) spectroscopy (400-1000 nm) combined with machine learning for this classification task. Spectral data were acquired from the sugarcane cultivar Khon Kaen 3 at multiple orientations, and various preprocessing techniques were employed to enhance spectral features. Three machine learning algorithms, linear discriminant analysis (LDA), K-Nearest Neighbors (KNNs), and artificial neural networks (ANNs), were evaluated for their classification performance. The results demonstrated high accuracy across all models, with ANN coupled with derivative preprocessing achieving an F1-score of 0.93 on both calibration and validation datasets, and 0.92 on an independent test set. This study underscores the feasibility of Vis-SWNIR spectroscopy and machine learning for rapid and precise node/internode classification, paving the way for automation in sugarcane billet preparation and other precision agriculture applications.

Granular characterisation of coal spoil dump using unmanned aerial vehicle data to enhance stability analysis

Open pit mining operations generate significant spoil dumps that need to be characterised for stability to identify potentially unstable slopes. However, the current subjective practice for spoil characterisation often involves tedious and risky field work. To this end, this study demonstrated the use of periodically acquired unmanned aerial vehicle (UAV)-based images over a coal mine spoil dump in New South Wales, Australia. A granular approach that captures the variability of each truck offload pile on a dump was adopted through morphology-based segmentation and ensemble algorithm-based classification which consolidates predictions from multiple classifiers. Overall accuracy of over 90% in the material characterisation based on the classification framework was achieved. The two-dimensional classification outcome was then transformed into three-dimensional (3D) block models using a point-based interpolation approach for stability analysis. The factor of safety derived from the granular approach offered improved assessment of failure risk compared to the conventional approaches, which treat the entire dump as a uniform category. This rapid classification and assessment method proposed in this study will help reduce the uncertainty associated with the variability of spoil dumps in slope stability assessments, thereby enhancing the safety and efficiency of mining operations.

A depth-based land-use semantic extraction method for landscape images

Integrated aerial and ground-based observations are essential for rapid land-use classification in modern land surveys, requiring efficient analysis of large volumes of landscape images. These images often contain background noise that impedes classification accuracy and exhibit spatial compression of the main scene, which causes semantic spatial distortion as depth increases. To address these challenges, this paper proposes a depth-based land-use semantic extraction (DLSE) method that effectively reduces background noise and corrects spatial distortions. The DLSE method follows three main steps: (1) depth estimation per pixel using a multi-resolution neural network to delineate the main scene range, filtering semantic noise; (2) conversion of perspective projection to orthographic projection per pixel to correct spatial distortion; and (3) improved land-use semantic extraction through MobileViT-enhanced feature extraction in SegNet. Experimental results demonstrate that DLSE achieves a 96.84% accuracy with more detailed outputs and operates at 23 fps, positioning it as an efficient tool for automated land surveys and land-use decision-making.

Classification of Vaginal Cleanliness Grades through Surface‐Enhanced Raman Spectral Analysis via The Deep‐Learning Variational Autoencoder–Long Short‐Term Memory Model

In this study, it is aimed to establish a novel method based on a deep‐learning‐guided surface‐enhanced Raman spectroscopy (SERS) technique to achieve rapid and accurate classification of vaginal cleanliness levels. We proposed a variational autoencoder (VAE) approach to enhance spectral quality, coupled with a deep learning algorithm long short‐term memory (LSTM) neural network to analyze SERS spectra produced by vaginal secretions. The performance of various machine learning (ML) algorithms is assessed using multiple evaluation metrics. Finally, the reliability of the optimal model is tested using blind test data (N = 10/group for each cleanliness level). The data quality of the SERS fingerprints of four types of vaginal secretions is significantly improved after VAE decoding and reconstruction. The signal‐to‐noise ratio of the generated spectra increased from the original 2.58–11.13. Among all algorithms, the VAE–LSTM algorithm demonstrates the best prediction ability and time efficiency. Additionally, blind test datasets yielded an overall accuracy of 85%. In this study, it is concluded that the deep‐learning‐guided SERS technique holds significant potential in rapidly distinguishing between different levels of vaginal cleanliness through human vaginal secretion samples. This contributes to the efficient diagnosis of vaginal cleanliness levels in clinical settings.