Feature Detector-level Fusion Methods in Food Recognition

Chest radiography is the most preferred non-invasive imaging technique for early diagnosis of Tuberculosis (TB). However, lack of radiological expertise in TB detection leads to indiscriminate chest radiograph (CXR) screening. A modest classification approach based on the local image description to detect subtle characteristics of TB using CXRs is highly recommended. In this work, an attempt has been made to classify normal and TB CXR images using Bag of Features (BoF) approach with Speeded-Up Robust Feature (SURF) descriptor. The images are obtained from a public database. Lung fields segmentation is performed using Distance Regularized Level Set (DRLS) formulation. The results of segmentation are validated against the ground truth images using similarity, overlap and area correlation measures. BoF approach with SURF keypoint descriptors is implemented to categorize the images using Multilayer Perceptron (MLP) classifier. The obtained results demonstrate that the DRLS method is able to delineate lung fields from CXR images. The BoF with SURF keypoint descriptor is able to characterize local attributes of normal and TB images. The segmentation results are found to be in high correlation with ground truth. MLP classifier is found to provide high Recall, Specificity (Spec), Accuracy, F-score and Area Under the Curve (AUC) values of 87.7%, 85.9%, 87.8%, 87.6% and 94% respectively between normal and abnormal images. The proposed computer aided diagnostic approach is found to perform better as compared to the existing methods. Thus, the study can be of significant assistance to physicians at the point of care in resource constrained regions.

Early detection of diabetic foot ulcers from thermal images using the bag of features technique

Breast cancer is one of the leading causes of death by cancer among women. The high mortality rates and the occurrence of this cancer worldwide show the importance of the investigation and development of means for the detection and early diagnosis of this disease. Computer-Aided Detection and Diagnosis systems have been developed to improve diagnostic accuracy by radiologists. This work proposes a method for discriminating patterns of malignancy and benignity of masses in digitized mammography images through the analysis of local features. The method comparatively applies the Scale-Invariant Feature Transform (SIFT), Speed Up Robust Feature (SURF), Oriented Fast and Rotated BRIEF (ORB) and Local Binary Pattern (LBP) descriptors for local feature extraction. These features are represented by a Bag of Features (BoF) model, applied to provide new representations of the data and to reduce its dimensionality. Finally, the features are used as input for the Support Vector Machine (SVM), Adaptive Boosting (Adaboost) and Random Forests (RF) classifiers to differentiate malignant and benign masses. The method obtained significant results, reaching 100% sensitivity, 99.65% accuracy and 99.24% specificity for benign and malignant mass classification.

This work addresses the problem of short-term people re-identification, an interesting field with applications in different domains like banks, airports, border crossings etc. For the purpose of this research, an algorithm is presented based on image classification. It is divided in three parts: building a dataset, creating a model, and real-time people re-identification. For every person in dataset, bag of features is used for image classification and features are extracted using MinEigen (Minimum Eigenvalue Algorithm) and SURF (Speeded Up Robust Features). Features are further trained using SVM (Support Vector Machine) and then exported as a model which is used for real time prediction and person recognition.

Food object recognition has gained popularity in recent years. This can perhaps be attributed to its potential applications in fields such as nutrition and fitness. Recognizing food images however is a challenging task since various foods come in many shapes and sizes. Besides having unexpected deformities and texture, food images are also captured in differing lighting conditions and camera viewpoints. From a computer vision perspective, using global image features to train a supervised classifier might be unsuitable due to the complex nature of the food images. Local features on the other hand seem the better alternative since they are able to capture minute intricacies such as interest points and other intricate information. In this paper, two local features namely SURF (Speeded- Up Robust Feature) and MSER (Maximally Stable Extremal Regions) are investigated for food object recognition. Both features are computationally inexpensive and have shown to be effective local descriptors for complex images. Specifically, each feature is firstly evaluated separately. This is followed by feature fusion to observe whether a combined representation could better represent food images. Experimental evaluations using a Support Vector Machine classifier shows that feature fusion generates better recognition accuracy at 86.6%.

Dietary patterns analysis is an emerging area of research. Identifying distinct patterns within a large dietary survey can give a more accurate representation of what people are eating. Furthermore, it allows researchers to analyse relationships between non-communicable diseases (NCD) and complete diets rather than individual food items or nutrients. However, few such studies have been conducted in developing countries including India, where the population has a high burden of diabetes and CVD. We undertook a systematic review of published and grey literature exploring dietary patterns and relationships with diet-related NCD in India. We identified eight studies, including eleven separate models of dietary patterns. Most dietary patterns were vegetarian with a predominance of fruit, vegetables and pulses, as well as cereals; dietary patterns based on high-fat, high-sugar foods and more meat were also identified. There was large variability between regions in dietary patterns, and there was some evidence of change in diets over time, although no evidence of different diets by sex or age was found. Consumers of high-fat dietary patterns were more likely to have greater BMI, and a dietary pattern high in sweets and snacks was associated with greater risk of diabetes compared with a traditional diet high in rice and pulses, but other relationships with NCD risk factors were less clear. This review shows that dietary pattern analyses can be highly valuable in assessing variability in national diets and diet-disease relationships. However, to date, most studies in India are limited by data and methodological shortcomings.

The visual analysis of foods on social media by using food recognition algorithm provides valuable insight from the health, cultural and marketing. Food recognition offers a means to automatically recognise foods as well the useful information such as calories and nutritional estimation by using image processing and machine learning technique. The interest points in food image can be detected effectively by using Maximally Stable Extremal Region (MSER). As MSER used global segmentation and many food images have a complex background, there are numerous irrelevant interest points are detected. These interest points are considered as noises that lead to computation burden in the overall recognition process. Therefore, this research proposes an Extremal Region Selection (ERS) algorithm to improve MSER detection by reducing the number of irrelevant extremal regions by using unsupervised learning based on the k-means algorithm. The performance of ERS algorithm is evaluated based on the classification performance metrics by using classification rate (CR), error rate (ERT), precision (Prec.) and recall (rec.) as well as the number of extremal regions produced by ERS. UECFOOD-100 and UNICT-FD1200 are the two food datasets used to benchmark the proposed algorithm. The results of this research have found that the ERS algorithm by using optimum parameters and thresholds, be able to reduce the number of extremal regions with sustained classification performance.

Cardiovascular diseases (CVDs) remain to be the leading causes of morbidity and mortality in Hong Kong and worldwide. Among different modifiable risk factors, dietary pattern is on the major determinant for CVD and overall mortality. Other than pharmacological therapies for cardiovascular risk factors, such as hypertension, hyperlipidemia and diabetes, maintaining a healthy diet is a more sustainable method in general population to prevent CVDs. Current lifestyle intervention in the West countries focus on high intake of fruit and vegetables with more than 400g per day and limited saturated fats with less than 10% of energy, there is very limited data on impact of dietary pattern on CVDs in Chinese. Prior studies among Chinese in Hong Kong have shown that only half of the local population fell within these recommended ranges for fat, saturated fatty acid and cholesterol intakes.
\n
\nSeveral different dietary patterns have been recommended for CVDs prevention based on: i) food groups, such as Mediterranean diet, the Dietary Approaches to Stop Hypertension (DASH) diet; ii) macronutrients: the low-carbohydrate diet, low glycemic index diet, very-low- fat diet and iii) nutrition or vitamin supplement. However, the effect of different dietary patterns based on modulations of food group, macronutrients and particular micronutrients on vascular structure and function in Chinese subjects is unclear. In the first part of this thesis, the relationships between different dietary pattern and surrogate markers of subclinical atherosclerosis and vascular function in different high risk populations for CVDs were investigated.
\n
\nIn Chapter 3, we compared the assessment of dietary pattern in Chinese using different tool, including Food Frequency Questionnaire (FFQ); Dietary Record; and Dietitian assessment. In this study, we demonstrated that suitable dietary assessments tools should be chosen for the assessment of different dietary pattern, according to characteristics of assessments.
\n
\nIn Chapter 4, the relationship between the fruit intake and subclinical atherosclerosis as measured by carotid intimal thickness (IMT) was investigated in patient with type II diabetes mellitus (DM). Our results showed that high fruit intake was associated with lower burden of carotid atherosclerosis, independent of level of vitamin intake in patients with type II DM.
\n
\nIn Chapter 5, we compared the impact of high carbohydrate diet on arterial stiffness between control subjects without CVDs and patients with high risk for CVDs. Our findings showed that high carbohydrate diet mainly affected patients with established CVDs, and their increased arterial stiffness was associated with an elevation of blood pressure.
\n
\nIn Chapter 6, we determined the effect of dietary vitamin intake on oxidative stress in patients with high risk of CVDs. In those high risk patients for CVDs, we demonstrated that increased dietary intake of vitamin A, beta-carotene and alpha tocopherol were associated with decreased oxidative stress, but these relationships were not observed in those control subjects without CVDs. It is likely attributed to the higher systemic oxidative stress levels in patients with high risk of CVDs.
\n
\nOn the other hand, food intake may also affect the clinical efficacy of cardiovascular therapies. In particularly, it has been well established that herbal intake which is commonly used by Chinese can affect the anticoagulant effect of warfarin on patients with non-valvular atrial fibrillation (AF). Thus, in this second part of the thesis, we investigated the effect of concomitant herbal intake on anticoagulation control in patients with non-valvular AF treated with warfarin. Our results showed that patients with AF treated with warfarin had limited knowledge on potential interaction between herbal substances in foods and warfarin, in which increased herbal substances intake significantly reduced the percentage time of anticoagulant effect within the therapeutic range. Moreover, a single section of education on knowledge of herbal ingredients did not improve their percentage time of therapeutic range for these patients.
\n
\nIn conclusion, these findings suggest that dietary pattern in Chinese might have significant impact of vascular function in patients with type II DM and high risk for CVDs. Moreover, the herbal substances in the diet among Chinese could have significant impact of the therapeutic effects in some of the cardiovascular medications, such as warfarin. Future clinical studies will be needed to confirm these potential beneficial effects of particular diet intake on vascular function in patients with high risks of CVDs as well as potential interaction between herbal substances in Chinese diet and cardiovascular medications.

Several studies have demonstrated that dietary patterns identified by a posteriori and hybrid methods are associated with gastrointestinal (GI) cancer risk and mortality. These studies applied different methods for analyzing dietary data and reported inconsistent findings. This systematic review and meta-analysis were aimed to determine the association between dietary patterns, derived using principal component analysis (PCA) and reduced rank regression (RRR), and GI cancer risk and GI cancer-caused mortality. Articles published up to June 2023 in English were eligible for inclusion. The Medline, SCOPUS, Cochrane Library, CINHAL, PsycINFO, ProQuest, and Web of Sciences databases were used to identify prospective studies. The Preferred Reporting Item for Systematic Review and Meta-analysis Protocol 2020 was used to report results. A total of 28 studies were eligible for inclusion. Varied approaches to deriving dietary patterns were used, including PCA (n = 22), RRR (n = 2), combined PCA and RRR (n = 1), cluster analysis (CA; n = 2) and combined PCA and CA (n = 1). Two dietary patterns, "healthy" and "unhealthy," were derived using PCA and RRR. The healthy dietary pattern was characterized by a higher intake of fruits, whole grains, legumes, vegetables, milk, and other dairy products, whereas the unhealthy dietary pattern was characterized by a higher intake of red and processed meat, alcohol, and both refined and sugar-sweetened beverages. The findings indicated that the PCA-derived healthy dietary pattern was associated with an 8% reduced risk (relative risk [RR], 0.92; 95% CI, 0.87-0.98), and the unhealthy dietary pattern was associated with a 14% increased risk (RR, 1.14; 95% CI, 1.07-1.22) of GI cancers. Similarly, the RRR-derived healthy dietary pattern (RR, 0.83; 95% CI, 0.61-1.12) may be associated with reduced risk of GI cancers. In contrast, the RRR-derived unhealthy dietary pattern (RR, 0.93; 95% CI, 0.57-1.52) had no association with a reduced risk of GI cancers. Similarly, evidence suggested that PCA-derived healthy dietary patterns may reduce the risk of death from GI cancers, whereas PCA-derived unhealthy dietary patterns may increase the risk. Findings from prospective studies on the association of PCA-derived dietary patterns and the risk of GI cancers support the evidence of healthy and unhealthy dietary patterns as either protective or risk-increasing factors for GI cancers and for survivorship, respectively. The findings also suggest that the RRR-derived healthy dietary pattern reduces the risk of GI cancers (albeit with low precision), but no association was found for the RRR-derived unhealthy dietary pattern. Prospective studies are required to further clarify disparities in the association between PCA- and RRR-derived dietary patterns and the risk of GI cancers. Systematic review registration: PROSPERO registration no. CRD42022321644.

Magnetic resonance imaging (MRI) is increasingly used in the diagnosis of Alzheimer's disease (AD) in order to identify abnormalities in the brain. Indeed, cortical atrophy, a powerful biomarker for AD, can be detected using structural MRI (sMRI), but it cannot detect impairment in the integrity of the white matter (WM) preceding cortical atrophy. The early detection of these changes is made possible by the novel MRI modality known as diffusion tensor imaging (DTI). In this study, we integrate DTI and sMRI as complementary imaging modalities for the early detection of AD in order to create an effective computer‐assisted diagnosis tool. The fused Bag‐of‐Features (BoF) with Speeded‐Up Robust Features (SURF) and modified AlexNet convolutional neural network (CNN) are utilized to extract local and deep features. This is applied to DTI scalar metrics (fractional anisotropy and diffusivity metric) and segmented gray matter images from T1‐weighted MRI images. Then, the classification of local unimodal and deep multimodal features is first performed using support vector machine (SVM) classifiers. Then, the majority voting technique is adopted to predict the final decision from the ensemble SVMs. The study is directed toward the classification of AD versus mild cognitive impairment (MCI) versus cognitively normal (CN) subjects. Our proposed method achieved an accuracy of 98.42% and demonstrated the robustness of multimodality imaging fusion.

The analysis of dietary patterns has recently drawn considerable attention as a method of investigating the association between the overall whole diet and the risk of colorectal cancer. However, the results have yielded conflicting findings. Here, we carried out a meta-analysis to identify the association between dietary patterns and the risk of colorectal cancer. A total of 40 studies fulfilled the inclusion criteria and were included in this meta-analysis. The highest category of 'healthy' dietary pattern compared with the lowest category was apparently associated with a decreased risk for colorectal cancer [odds ratio (OR)=0.75; confidence interval (CI): 0.68-0.83; P<0.00001]. An increased risk of colorectal cancer was shown for the highest compared with the lowest category of a 'western-style' dietary pattern (OR=1.40; CI: 1.26-1.56; P<0.00001). There was an increased risk of colorectal cancer in the highest compared with the lowest category of 'alcohol-consumption' pattern (OR=1.44; CI: 1.13-1.82; P=0.003). The results of this meta-analysis indicate that a 'healthy' dietary pattern may decrease the risk of colorectal cancer, whereas 'western-style' and 'alcohol-consumption' patterns may increase the risk of colorectal cancer.

Background: Dietary pattern assessment by healthcare providers leads to a better understanding of usual intake and evaluation of nutritional status, systemic health, and disease. Interprofessional team members can use such information to provide interventions leading to improved health outcomes. Objective: The aim was to explore the dietary patterns of adults seen in a dental clinic using the Rapid Eating and Activity Assessment for Patients (REAP) tool. Methods: This was a cross-sectional study of data from 220 adult patients (aged 18-89 years) who had a diet evaluation completed in a dental school clinic. Demographic information and REAP responses were obtained from the electronic health record and reported using frequency distributions. Results: The study sample (N=220) was 50.0% male (n= 110). The median (IQR) age and BMI were 56.0 years (IQR=48.2, 66.0) and 28.0 kg/m2 (IQR= 24.3, 32.8 kg/m2), respectively; 73.5% had a BMI considered overweight or obese. Approximately one-third reported usually/sometimes eating sweets more than twice/day (n=74, 35.9%) and drinking more than 16 ounces of SSBs (n=74, 34.1%). Most reported usually/sometimes eating less than 3 servings of whole grains (n=165, 75.0%), 2-3 servings of fruit (n=155, 71.1%), or 3-4 servings of vegetables (n=121, 70.8%) daily. Seventy-two percent (n=156) indicated they were willing to make dietary changes. Conclusion: This study revealed that the dietary patterns of adults seen in a dental school clinic did not align with the Dietary Guidelines for Americans recommendations. Diet assessment is feasible in this setting and may be an important part of interprofessional education.

Although several nutrients and foods have been suggested to be preventive for constipation, all previous studies have examined a single nutrient or food in each analysis. In contrast, analysis of dietary patterns may provide new insights into the influence of diet on functional constipation. We conducted a cross-sectional examination of the association between dietary pattern and functional constipation in 3,770 Japanese female dietetic course students aged 18-20 y from 53 institutions in Japan. Diet was assessed with a validated self-administered diet history questionnaire with 148 food items, from which 30 food groups were created and entered into a factor analysis. Functional constipation was defined using the Rome I criteria, which has previously been used in several epidemiologic studies on constipation. The prevalence of functional constipation was 26.0% (n=979). Four dietary patterns were identified: (1) "Healthy", (2) "Japanese traditional", (3) "Western," and 4) "Coffee and dairy products." After adjustment for several confounding factors, the "Japanese traditional" pattern, characterized by a high intake of rice, miso soup, and soy products and a low intake of bread and confectionaries, was associated with a significantly lower prevalence of functional constipation. In comparison with the lowest quintile, the multivariate adjusted odds ratio (95% confidence interval) was 0.52 (0.41-0.66) in the highest quintile (p for trend < 0.0001). Other dietary patterns were not associated with functional constipation. The Japanese traditional dietary pattern, characterized by a high intake of rice and a low intake of bread and confectionaries, may be beneficial in preventing functional constipation in young Japanese women.

In this paper, design and implementation the feature extraction method of Speeded-Up Robust Features (SURF) and Support Vector Machine (SVM) classification method into the traffic signs recognition application. The output of this application is the meaning of the traffic sign with two languages, indonesia and english. In the SURF method, the smallest large number of keypoints will affect the accuracy level to recognize a image. Based on the results, accuracy of this traffic signs detection has a high accuracy rate of 96%, when taking this image right in the green box displayed on the smartphone screen and taken when the brightness level of the light on 4106 lux up to 10896 lux.

To investigate associations between dietary patterns and suboptimal iron status in premenopausal women living in Auckland, New Zealand. Premenopausal women (n = 375; 18-44 years) were included in this cross-sectional analysis. Suboptimal iron status was defined as serum ferritin <20 μg/L. Participants completed a 144-item iron food frequency questionnaire (FeFFQ) and a questionnaire on dietary practices to assess dietary intake over the past month. Factor analysis was used to determine dietary patterns from the FeFFQ. Logistic regression was used to determine associations between these dietary patterns and iron status. Seven dietary patterns were identified: refined carbohydrate and fat; Asian; healthy snacks; meat and vegetable; high tea and coffee; bread and crackers; and milk and yoghurt. Logistic regression suggested that following a "meat and vegetable" dietary pattern reduced the risk of suboptimal iron status by 41 % (95 % CI: 18, 58 %; P = 0.002) and following a "milk and yoghurt" pattern increased the risk of suboptimal iron status by 50 % (95 % CI: 15, 96 %; P = 0.003). These results suggest that dietary patterns characterized by either a low intake of meat and vegetables or a high intake of milk and yoghurt are associated with an increased risk of suboptimal iron status. Dietary pattern analysis is a novel and potentially powerful tool for investigating the relationship between diet and iron status.