Principal Component Analysis applied to digital image compression

PCA and logistic regression in 2-[18F]FDG PET neuroimaging as an interpretable and diagnostic tool for Alzheimer's disease

Objective.to develop an optimization and training pipeline for a classification model based on principal component analysis and logistic regression using neuroimages from PET with 2-[18F]fluoro-2-deoxy-D-glucose (FDG PET) for the diagnosis of Alzheimer's disease (AD).Approach.as training data, 200 FDG PET neuroimages were used, 100 from the group of patients with AD and 100 from the group of cognitively normal subjects (CN), downloaded from the repository of the Alzheimer's Disease Neuroimaging Initiative (ADNI). Regularization methods L1 and L2 were tested and their respective strength varied by the hyperparameter C. Once the best combination of hyperparameters was determined, it was used to train the final classification model, which was then applied to test data, consisting of 192 FDG PET neuroimages, 100 from subjects with no evidence of AD (nAD) and 92 from the AD group, obtained at the Centro de Diagnóstico por Imagem (CDI).Main results.the best combination of hyperparameters was L1 regularization andC≈ 0.316. The final results on test data were accuracy = 88.54%, recall = 90.22%, precision = 86.46% and AUC = 94.75%, indicating that there was a good generalization to neuroimages outside the training set. Adjusting each principal component by its respective weight, an interpretable image was obtained that represents the regions of greater or lesser probability for AD given high voxel intensities. The resulting image matches what is expected by the pathophysiology of AD.Significance.our classification model was trained on publicly available and robust data and tested, with good results, on clinical routine data. Our study shows that it serves as a powerful and interpretable tool capable of assisting in the diagnosis of AD in the possession of FDG PET neuroimages. The relationship between classification model output scores and AD progression can and should be explored in future studies.

Agroindustrial wastes as a substrate for the cultivation of Eruca sativa Miller seedlings: physical-chemical and phytometric parameters assessment

Evaluation was made of the use of organic substrates obtained from the composting of poultry industry wastes, together with crude glycerin, for the production of arugula seedlings (Eruca sativa Miller). The raw materials included hatchery waste, chicken litter, and flotation tank sludge, in combination with other materials such as tree pruning, sugarcane bagasse, crude glycerin (at 0, 1.5, 3.0, 4.5, and 6.0%), and boiler charcoal. Analysis of the organic substrates included determination of nitrogen, phosphorus, and potassium (NPK), pH, electrical conductivity, functional groups, and carboxylic acids. Physical parameters determined were water retention capacity, solids volume, porosity, density, and granulometry. For the arugula seedlings, determinations were made of the ease of removal of the root ball from the tray, the effect of free drop on the root ball, phytometric parameters, and total phenolic compounds. Decreased concentrations of carboxylic acids, together with the presence of aromatic functional groups, indicated maturation/stabilization of the organic substrates. The phytometric measurements indicated that the use of the organic substrates with addition of 3.0, 4.5, and 6.0% of crude glycerin favored arugula production and led to higher contents of total phenolic compounds in the seedlings, with values of 3657.54, 3602.13, and 3232.92 mg GAE g^-1, respectively. The results demonstrated that the use of these organic substrates with the addition of crude glycerin provided satisfactory development of arugula seedlings.

Applied aerial spectroscopy: A case study on remote sensing of an ancient and semi-natural woodland

An area of ancient and semi-natural woodland (ASNW) has been investigated by applied aerial spectroscopy using an unmanned aerial vehicle (UAV) with multispectral image (MSI) camera. A novel normalised difference spectral index (NDSI) algorithm was developed using principal component analysis (PCA). This novel NDSI was then combined with a simple segmentation method of thresholding and applied for the identification of native tree species as well as the overall health of the woodland. Using this new approach allowed the identification of trees at canopy level, across 7.4 hectares (73,934 m2) of ASNW, as oak (53%), silver birch (37%), empty space (9%) and dead trees (1%). This UAV derived data was corroborated, for its accuracy, by a statistically valid ground-level field study that identified oak (47%), silver birch (46%) and dead trees (7.4%). This simple innovative approach, using a low-cost multirotor UAV with MSI camera, is both rapid to deploy, was flown around 100 m above ground level, provides useable high resolution (5.3 cm / pixel) data within 22 mins that can be interrogated using readily available PC-based software to identify tree species. In addition, it provides an overall oversight of woodland health and has the potential to inform a future woodland regeneration strategy.

The Use of an Unmanned Aerial Vehicle for Tree Phenotyping Studies

A strip of 20th-century landscape woodland planted alongside a 17th to mid-18th century ancient and semi-natural woodland (ASNW) was investigated by applied aerial spectroscopy using an unmanned aerial vehicle (UAV) with a multispectral image camera (MSI). A simple classification approach of normalized difference spectral index (NDSI), derived using principal component analysis (PCA), enabled the identification of the non-native trees within the 20th-century boundary. The tree species within this boundary, classified by NDSI, were further segmented by the machine learning segmentation method of k-means clustering. This combined innovative approach has enabled the identification of multiple tree species in the 20th-century boundary. Phenotyping of trees at canopy level using the UAV with MSI, across 8052 m2, identified black pine (23%), Norway maple (19%), Scots pine (12%), and sycamore (19%) as well as native trees (oak and silver birch, 27%). This derived data was corroborated by field identification at ground-level, over an area of 6785 m2, that confirmed the presence of black pine (26%), Norway maple (30%), Scots pine (10%), and sycamore (14%) as well as other trees (oak and silver birch, 20%). The benefits of using a UAV, with an MSI camera, for monitoring tree boundaries next to a new housing development are demonstrated.

Thermal energy harvesting near-infrared radiation and accessing low temperatures with plasmonic sensors

Near-infrared (NIR) thermal energy harvesting has been demonstrated for gold nanorods (AuNRs), allowing concentration dependent, ppm-level, gas detection of H2, CO, and NO2 at 500 °C without using a white light source. Part-per-million detection capabilities of the gold nanorods are demonstrated with a factor of 11 reduction in collection times in the NIR as compared to measurements made in the visible light region. Decreased collection times are enabled by an increase in S : N ratio, which allowed a demonstration of selectivity through the use of both full spectral and a reduced spectral-based principal component analysis. Furthermore, low temperature thermal imaging spectra have been obtained at sample temperatures ranging from 275-500 °C, showing the possibility of energy harvested gas sensing at lower temperatures. These findings are promising in the area of miniaturizing plasmonic gas sensing technology and integration in areas such as gas turbines.