Search

• We analysed over 350 references from all well-famed databases. • We provided a comprehensive survey on deep learning in food category recognition. • This review encompassed current challenges & applications, strengths & limitations. • Fundamental deep learning rules and performance assessment methods were reviewed. • Convolutional neural networks, transfer and semi-supervised learning were reviewed. Integrating artificial intelligence with food category recognition has been a field of interest for research for the past few decades. It is potentially one of the next steps in revolutionizing human interaction with food. The modern advent of big data and the development of data-oriented fields like deep learning have provided advancements in food category recognition. With increasing computational power and ever-larger food datasets, the approach's potential has yet to be realized. This survey provides an overview of methods that can be applied to various food category recognition tasks, including detecting type, ingredients, quality, and quantity. We survey the core components for constructing a machine learning system for food category recognition, including datasets, data augmentation, hand-crafted feature extraction, and machine learning algorithms. We place a particular focus on the field of deep learning, including the utilization of convolutional neural networks, transfer learning, and semi-supervised learning. We provide an overview of relevant studies to promote further developments in food category recognition for research and industrial applications. [ABSTRACT FROM AUTHOR]

Background: The number of patients with Alzheimer's disease is increasing rapidly every year. Scholars often use computer vision and machine learning methods to develop an automatic diagnosis system.Objective: In this study, we developed a novel machine learning system that can make diagnoses automatically from brain magnetic resonance images.Methods: First, the brain imaging was processed, including skull stripping and spatial normalization. Second, one axial slice was selected from the volumetric image, and stationary wavelet entropy (SWE) was done to extract the texture features. Third, a single-hidden-layer neural network was used as the classifier. Finally, a predator-prey particle swarm optimization was proposed to train the weights and biases of the classifier.Results: Our method used 4-level decomposition and yielded 13 SWE features. The classification yielded an overall accuracy of 92.73±1.03%, a sensitivity of 92.69±1.29%, and a specificity of 92.78±1.51%. The area under the curve is 0.95±0.02. Additionally, this method only cost 0.88 s to identify a subject in online stage, after its volumetric image is preprocessed.Conclusion: In terms of classification performance, our method performs better than 10 state-of-the-art approaches and the performance of human observers. Therefore, this proposed method is effective in the detection of Alzheimer's disease. [ABSTRACT FROM AUTHOR]

In order to detect multiple sclerosis (MS) subjects from healthy controls (HCs) in magnetic resonance imaging, we developed a new system based on machine learning. The MS imaging data was downloaded from the eHealth laboratory at the University of Cyprus, and the HC imaging data was scanned in our local hospital with volunteers enrolled from community advertisement. Inter-scan normalization was employed to remove the gray-level difference. We adjust the misclassification costs to alleviate the effect of unbalanced class distribution to the classification performance. We utilized two-level stationary wavelet entropy (SWE) to extract features from brain images. Then, we compared three machine learning based classifiers: the decision tree, k-nearest neighbors (kNN), and support vector machine. The experimental results showed the kNN performed the best among all three classifiers. In addition, the proposed SWE+kNN approach is superior to four state-of-the-art approaches. Our proposed MS detection approach is effective. [ABSTRACT FROM AUTHOR]

This letter researched on the effect of spider-web-plot, which was proposed in the literature “ M. Saritha, K.P. Joseph, A.T. Mathew, Classification of MRI brain images using combined wavelet entropy based spider web plots and probabilistic neural network, Pattern Recognit. Lett. 34 (2013) 2151–2156 ”. We found it was unnecessary to use spider-web-plot as feature transform method. In addition, our simulation results showed that removing the procedure of spider-web-plot yielded the same classification accuracy of 100%. [ABSTRACT FROM AUTHOR]

Real-time monitoring of rock stability during the mining process is critical. This paper first proposed a RIME algorithm (CCRIME) based on vertical and horizontal crossover search strategies to improve the quality of the solutions obtained by the RIME algorithm and further enhance its search capabilities. Then, by constructing a binary version of CCRIME, the key parameters of FKNN were optimized using a binary conversion method. Finally, a discrete CCRIME-based BCCRIME was developed, which uses an S-shaped function transformation approach to address the feature selection issue by converting the search result into a real number that can only be zero or one. The performance of CCRIME was examined in this study from various perspectives, utilizing 30 benchmark functions from IEEE CEC2017. Basic algorithm comparison tests and sophisticated variant algorithm comparison experiments were also carried out. In addition, this paper also used collected microseismic and blasting data for classification prediction to verify the ability of the BCCRIME-FKNN model to process real data. This paper provides new ideas and methods for real-time monitoring of rock mass stability during deep well mineral resource mining. [ABSTRACT FROM AUTHOR]

According to the World Health Organisation, falling is a major health problem with potentially fatal implications. Each year, thousands of people die as a result of falls, with seniors making up 80% of these fatalities. The automatic detection of falls may reduce the severity of the consequences. Our study focuses on developing a vision-based fall detection system. Our work proposes a new feature descriptor that results in a new fall detection framework. The body geometry of the subject is analyzed and patterns that help to distinguish falls from non-fall activities are identified in our proposed method. An AlphaPose network is employed to identify 17 keypoints on the human skeleton. Thirteen keypoints are used in our study, and we compute two additional keypoints. These 15 keypoints are divided into five segments, each of which consists of a group of three non-collinear points. These five segments represent the left hand, right hand, left leg, right leg and craniocaudal section. A novel feature descriptor is generated by extracting the distances from the segmented parts, angles within the segmented parts and the angle of inclination for every segmented part. As a result, we may extract three features from each segment, giving us 15 features per frame that preserve spatial information. To capture temporal dynamics, the extracted spatial features are arranged in the temporal sequence. As a result, the feature descriptor in the proposed approach preserves the spatio-temporal dynamics. Thus, a feature descriptor of size [ m × 15 ] is formed where m is the number of frames. To recognize fall patterns, machine learning approaches such as decision trees, random forests, and gradient boost are applied to the feature descriptor. Our system was evaluated on the UPfall dataset, which is a benchmark dataset. It has shown very good performance compared to the state-of-the-art approaches. [ABSTRACT FROM AUTHOR]

The variable magnification levels in histopathology images make it difficult to accurately categorize tumor regions in breast cancer histology. In this study, a novel architecture for accurate image interpretation MagNet is presented. With specific modules like Separable Dilation Convolution (SDC), Separable Dilation Skip Block (SDSB), and Point-wise Reformation Block (PRB), MagNet uses a Parallel U-Net (PU-Net) infrastructure. SDC in the PU-Net encoder ensures downsampled generalized feature representations by capturing characteristic attributes at varying magnifications. Using feature upsampling, attribute mapping merging, and PRB for precise feature capture, the decoder improves reconstruction. While PRB combines data from several decoder levels, SDSB creates vital links between the encoder and decoder layers. MagNet requires less processing of histopathology images and improves multi-magnification feature maps. MagNet performs exceptionally well, constantly outperforming rivals in terms of accuracy (0.98), F1 score (0.97), sensitivity (0.96), and specificity (0.97). The effectiveness of MagNet and its revolutionary potential in cancer diagnostics are shown by these quantitative data. [ABSTRACT FROM AUTHOR]

[Display omitted] Tetragonal ratio (c / a) is a critical structural parameter that heavily decides plenty of performances of tetragonal martensitic materials. Nevertheless, the knowledge about c/a remains limited and there is still no definitive strategy to tailor it. In this work, the machine learning method, combined with high-throughput ab-initio calculation, was introduced as an attempt to investigate c/a. After training, a high-precision random forest model of predicting c / a was established. Combining a five-step descriptor screening and visualization analyses, the most pertinent parameters deciding c / a , i.e. , mean melting point ( T ¯ M) and mean volume ( V ¯ ) of constituent elements, were identified. Furthermore, a simple relation between c/a and these two parameters, i.e. , c / a = m T ¯ M / V ¯ + b where the m and b are constants, was proposed. Surprisingly, this relation exhibits an exceptional generalization performance in a wide range of untrained materials. This work is expected to provide an explicit, efficient and general route to tailor c/a and hence promote the design of advanced martensitic materials. [ABSTRACT FROM AUTHOR]

• More than 245 references from top journals are analyzed. • A systematic theoretical analysis of deep learning in dentistry is provided. • Extensive works on deep learning systems of dentistry are well summarized. • Future challenges of deep learning methods in dentistry are analyzed. Oral diseases have a significant impact on human health, often going unnoticed in their early stages. Deep learning, a promising field in artificial intelligence, has shown remarkable success in various domains, especially dentistry. This paper aims to provide an overview of recent research on deep learning applications in dentistry, with a focus on dental imaging. Deep learning algorithms perform well in difficult tasks such as image segmentation and recognition, enabling accurate identification of oral conditions and abnormalities. Integration of deep learning with other oral health data offers a holistic understanding of the relationship between oral and systemic health. However, there are still many challenges that need to be addressed. [ABSTRACT FROM AUTHOR]

Locally advanced cervical cancer (LACC) is frequently treated with neoadjuvant chemotherapy (NACT), which includes paclitaxel and platinum. However, the development of severe chemotherapy toxicity is a barrier to successful NACT. Phosphatidylinositol 3-kinase (PI3K)/serine/threonine kinase (AKT) pathway is related to the occurrence of chemotherapeutic toxicity. In this research work, we employ a random forest (RF) machine learning model to forecast NACT toxicity (neurological, gastrointestinal, and hematological reactions). Twenty-four single nucleotide polymorphisms (SNPs) in the PI3K/AKT pathway from 259 LACC patients were used to construct a dataset. Following the data preprocessing, the RF model was trained. The Mean Decrease in Impurity approach was adopted to evaluate the relevance of 70 selected genotypes' importance by comparing chemotherapy toxicity grades 1–2 vs. 3. In the Mean Decrease in Impurity analysis, neurological toxicity was much more likely to occur in LACC patients with homozygous AA in Akt2 rs7259541 than in those with AG or GG genotypes. The CT genotype of PTEN rs532678 and the CT genotype of Akt1 rs2494739 increased the risk of neurological toxicity. The top three loci were rs4558508, rs17431184, and rs1130233, which were attributed to an elevated risk of gastrointestinal toxicity. LACC patients who had heterozygous AG in Akt2 rs7259541 exhibited an obviously greater risk of hematological toxicity than those who had AA or GG genotypes. And the CT genotype for Akt1 rs2494739 and the CC genotype in PTEN rs926091 showed a tendency to increase the risk of suffering from hematological toxicity. Akt2 rs7259541 and rs4558508, Akt1 rs2494739 and rs1130233, PTEN rs532678, rs17431184, and rs926091 polymorphisms are associated with different toxic effects during the chemotherapy treatment of LACC. [Display omitted] • Chemotherapy toxicity is predicted by SNPs & genotypes from MDI-based feature importance analysis. • The RF prediction model improves clinical treatment plans by reducing chemotherapeutic side effects. • This methodology is applicable to a variety of features with additional genotype datasets. [ABSTRACT FROM AUTHOR]

The PI3K/Akt pathway involves in regulating resistance to platinum-based neoadjuvant chemotherapy (NACT) in locally advanced cervical cancer (LACC) patients. Single nucleotide polymorphisms (SNPs) reflect the basic genetic variation between individuals. Random forest (RF) is one of the machine-learning models that can predict drug sensitivity with high accuracy. We applied the RF model for genomic prediction of NACT sensitivity in LACC patients. A total of 259 LACC patients were separated to two groups (i) effective and (ii) ineffective NACT group, depending on the NACT response. The 24 SNPs in four genes (PTEN, PIK3CA, Akt1, and Akt2) were genotyped by the Sequenom MassArray system in these patients. We implemented the SNPs as the feature to train the RF model, calculated the feature importance using mean decreases in impurity based on the model, and further analyzed the importance of each SNP. The importance analysis indicated that the top three SNPs (rs4558508, rs1130233, and rs7259541) and the last six loci (rs892120, rs62107593, rs34716810, rs10416620, rs41275748, and rs41275746) were all located in Akt. The patients carrying heterozygous GA in Akt2 rs4558508 had a considerably higher risk of chemoresistance than those carrying GG or AA genotype. The RF model could accurately predict the response to platinum-based NACT of LACC patients. The variables of Akt2 rs4558508 and rs7259541, and Akt1 rs1130233 were major polymorphic loci for NACT inefficiency. The LACC patients carrying heterozygous GA of Akt2 rs4558508 had a significantly increased risk of chemoresistance. Akt was an important gene in PI3K/Akt pathway that could predict the response of platinum-based NACT. The study applied the basis for an individualized approach to LACC patient therapy. [ABSTRACT FROM AUTHOR]

Background: Considering that Alzheimer's disease (AD) is untreatable, early diagnosis of AD from the healthy elderly controls (HC) is pivotal. However, computer-aided diagnosis (CAD) systems were not widely used due to its poor performance.Objective: Inspired from the eigenface approach for face recognition problems, we proposed an eigenbrain to detect AD brains. Eigenface is only for 2D image processing and is not suitable for volumetric image processing since faces are usually obtained as 2D images.Methods: We extended the eigenbrain to 3D. This 3D eigenbrain (3D-EB) inherits the fundamental strategies in either eigenface or 2D eigenbrain (2D-EB). All the 3D brains were transferred to a feature space, which encoded the variation among known 3D brain images. The feature space was named as the 3D-EB, and defined as eigenvectors on the set of 3D brains. We compared four different classifiers: feed-forward neural network, support vector machine (SVM) with linear kernel, polynomial (Pol) kernel, and radial basis function kernel.Results: The 50x10-fold stratified cross validation experiments showed that the proposed 3D-EB is better than the 2D-EB. SVM with Pol kernel performed the best among all classifiers. Our "3D-EB + Pol-SVM" achieved an accuracy of 92.81% ± 1.99% , a sensitivity of 92.07% ± 2.48% , a specificity of 93.02% ± 2.22% , and a precision of 79.03% ± 2.37% . Based on the most important 3D-EB U1, we detected 34 brain regions related with AD. The results corresponded to recent literature.Conclusions: We validated the effectiveness of the proposed 3D-EB by detecting subjects and brain regions related to AD. [ABSTRACT FROM AUTHOR]

Background: Within the past decade, computer scientists have developed many methods using computer vision and machine learning techniques to detect Alzheimer's disease (AD) in its early stages.Objective: However, some of these methods are unable to achieve excellent detection accuracy, and several other methods are unable to locate AD-related regions. Hence, our goal was to develop a novel AD brain detection method.Methods: In this study, our method was based on the three-dimensional (3D) displacement-field (DF) estimation between subjects in the healthy elder control group and AD group. The 3D-DF was treated with AD-related features. The three feature selection measures were used in the Bhattacharyya distance, Student's t-test, and Welch's t-test (WTT). Two non-parallel support vector machines, i.e., generalized eigenvalue proximal support vector machine and twin support vector machine (TSVM), were then used for classification. A 50 × 10-fold cross validation was implemented for statistical analysis.Results: The results showed that "3D-DF+WTT+TSVM" achieved the best performance, with an accuracy of 93.05 ± 2.18, a sensitivity of 92.57 ± 3.80, a specificity of 93.18 ± 3.35, and a precision of 79.51 ± 2.86. This method also exceled in 13 state-of-the-art approaches. Additionally, we were able to detect 17 regions related to AD by using the pure computer-vision technique. These regions include sub-gyral, inferior parietal lobule, precuneus, angular gyrus, lingual gyrus, supramarginal gyrus, postcentral gyrus, third ventricle, superior parietal lobule, thalamus, middle temporal gyrus, precentral gyrus, superior temporal gyrus, superior occipital gyrus, cingulate gyrus, culmen, and insula. These regions were reported in recent publications.Conclusions: The 3D-DF is effective in AD subject and related region detection. [ABSTRACT FROM AUTHOR]