Individual Descriptors Research Articles

Investigating level descriptors and their impact on validity

This article presents a validation study which investigates the relationship between mastery level descriptors and item difficulty in the National Tests of English (NTE) in Norway. The aim is to establish the extent to which the descriptors indicate item difficulty and thus support the argument that the mastery levels are a reflection of the framework of the NTE. This argument strength has direct implications for validity and for the defensibility of moving from criterion-based descriptors to norm-based national data. The study involved a panel of 10 raters assigning level descriptors from 7 content categories to 80 test items, giving a total of 5,600 individual judgements. These judgements are compared to the real test scores of around 46,000 pupils to establish if the level descriptors assigned by raters can predict pupil performance on the tests. The results show a strong correlation between rater judgements and real test scores, meaning that the descriptors offer an indication of item difficulty. However, some individual descriptor categories contain deviations from the expected order. We conclude that, while the level descriptors reflect the test’s framework, and thus support the validity argument, the argument could be strengthened with the revision of some individual descriptors.

Deconvoluting the Individual Effects of Nanoparticle Proximity and Size in Thermocatalysis.

Nanoparticle (NP) size and proximity are two physical descriptors applicable to practically all NP-supported catalysts. However, with conventional catalyst design, independent variation of these descriptors to investigate their individual effects on thermocatalysis remains challenging. Using a raspberry-colloid-templating approach, we synthesized a well-defined catalyst series comprising Pd12Au88 alloy NPs of three distinct sizes and at two different interparticle distances. We show that NP size and interparticle distance independently control activity and selectivity, respectively, in the hydrogenation of benzaldehyde to benzyl alcohol and toluene. Surface-sensitive spectroscopic analysis indicates that the surfaces of smaller NPs expose a greater fraction of reactive Pd dimers, compared to inactive Pd single atoms, thereby increasing intrinsic catalytic activity. Computational simulations reveal how a larger interparticle distance improves catalytic selectivity by diminishing the local benzyl alcohol concentration profile between NPs, thus suppressing its readsorption and consequently, undesired formation of toluene. Accordingly, benzyl alcohol yield is maximized using catalysts with smaller NPs separated by larger interparticle distances, overcoming activity-selectivity trade-offs. This work exemplifies the high suitability of the modular raspberry-colloid-templating method as a model catalyst platform to isolate individual descriptors and establish clear structure-property relationships, thereby bridging the materials gap between surface science and technical catalysts.

Efficient Crowd Anomaly Detection Using Sparse Feature Tracking and Neural Network

Crowd anomaly detection is crucial in enhancing surveillance and crowd management. This paper proposes an efficient approach that combines spatial and temporal visual descriptors, sparse feature tracking, and neural networks for efficient crowd anomaly detection. The proposed approach utilises diverse local feature extraction methods, including SIFT, FAST, and AKAZE, with a sparse feature tracking technique to ensure accurate and consistent tracking. Delaunay triangulation is employed to represent the spatial distribution of features in an efficient way. Visual descriptors are categorised into individual behaviour descriptors and interactive descriptors to capture the temporal and spatial characteristics of crowd dynamics and behaviour, respectively. Neural networks are then utilised to classify these descriptors and pinpoint anomalies, making use of their strong learning capabilities. A significant component of our study is the assessment of how dimensionality reduction methods, particularly autoencoders and PCA, affect the feature set’s performance. This assessment aims to balance computational efficiency and detection accuracy. Tests conducted on benchmark crowd datasets highlight the effectiveness of our method in identifying anomalies. Our approach offers a nuanced understanding of crowd movement and patterns by emphasising both individual and collective characteristics. The visual and local descriptors facilitate high-level analysis by closely relating to semantic information and crowd behaviour. The analysis observed shows that this approach offers an efficient framework for crowd anomaly detection, contributing to improved crowd management and public safety. The proposed model achieves accuracy of 99.5 %, 96.1%, 99.0% and 88.5% in the UMN scenes 1, 2, and 3 and violence in crowds datasets, respectively.

Meat quality and sensory traits in rabbits fed with two different percentages of bovine colostrum

The nutritional, antimicrobial, and antioxidant properties of bovine colostrum (BC) have encouraged its use in animal nutrition as a functional food in recent years. Nonetheless, the potential implications of BC supplementation on meat quality remain to be thoroughly assessed. To address this, thirty-nine New Zealand White rabbits (n = 13/group) were fed different dietary regimens until slaughter.: commercial standard diet for the control group (C) and C with 2.5% and 5% w/w of BC for BC-2.5 and BC-5 groups, respectively. Rabbits were slaughtered at 91 days of age and meat quality, and sensory characteristics were evaluated at days 2 (48 h after slaughter), 5, and 10 of refrigerated storage at 4 °C.The addition of colostrum in the diet resulted in a reduction of the total viable count, albeit only at the highest concentration and at the final detection, whereas for Lactobacillus spp. and Pseudomonas spp., there was little or no effect. The colour coordinates showed no differences between the groups, but they varied over time according to diet. Some differences between groups emerged in the definition of sensory attributes but did not affect the overall liking and overall scores of individual descriptors. These results indicate that the use of colostrum in rabbit feeding does not significantly impart meat quality and sensory attributes, but the potential of this valuable by-product for the food industry needs further investigation.

Neural network-based prediction of auto-ignition temperature of ternary mixed liquids

Auto-ignition temperature (AIT) is one of the crucial exponents in the design of fire and explosion safety measures. Therefore, in this study, quantitative structure-property relationship approach was used to predict the AIT of ternary hybrid liquids based on molecular structure information. The optimal molecular descriptors were calculated and filtered using Mordred software. Twelve mixing rules were proposed for calculating molecular descriptors of mixtures. A prediction model for the AIT value of binary liquid mixtures was developed, validated and evaluated using a back propagation neural network (BPNN) and a one-dimensional convolutional neural network (1DCNN). The relative contribution and positive and negative correlations between individual molecular descriptors and AIT in the model were interpreted using the shapley additive explanations method. The results show that BPNN and 1DCNN models using mixing rule 1 have the best fitting ability, stability and prediction ability. The determination coefficient of the BPNN and 1DCNN models in the training set were 0.996 and 0.992, the root mean square errors were 3.613 °C and 5.284 °C, the mean absolute errors were 2.483 °C and 4.144 °C, the nash efficiency coefficient was 0.996 and 0.992, respectively, the willmott index was 0.999 and 0.998. and the values of the top three molecular descriptors of relative contribution, SssCH2, SsOH and SsCH3, were negatively correlated with the AIT values. The BPNN and 1DCNN models provide an accurate and reliable method for predicting ternary mixing liquid AIT.

Exploring high thermal conductivity polymers via interpretable machine learning with physical descriptors

The efficient and economical exploitation of polymers with high thermal conductivity (TC) is essential to solve the issue of heat dissipation in organic devices. Currently, the experimental preparation of functional polymers with high TC remains a trial-and-error process due to the multi-degrees of freedom during the synthesis and characterization process. Polymer informatics equips machine learning (ML) as a powerful engine for the efficient design of polymers with desired properties. However, available polymer TC databases are rare, and establishing appropriate polymer representation is still challenging. In this work, we propose a high-throughput screening framework for polymer chains with high TC via interpretable ML and physical feature engineering. The hierarchical down-selection process stepwise optimizes the 320 initial physical descriptors to the final 20 dimensions and then assists the ML models to achieve a prediction accuracy R2 over 0.80, which is superior to traditional graph descriptors. Further, we analyze the contribution of the individual descriptors to TC and derive the explicit equation for TC prediction using symbolic regression. The high TC polymer structures are mostly π-conjugated, whose overlapping p-orbitals enable easy maintenance of strong chain stiffness and large group velocities. Ultimately, we establish the connections between the individual chains and the amorphous state of polymers. Polymer chains with high TC have strong intra-chain interactions, and their corresponding amorphous systems are favorable for obtaining a large radius of gyration and causing enhanced thermal transport. The proposed data-driven framework should facilitate the theoretical and experimental design of polymers with desirable properties.

Applications of machine learning algorithms to support COVID-19 diagnosis using X-rays data information

Due to the rapid spread of the new coronavirus variant during the most recent pandemic, it became difficult to distinguish common cold symptoms from those of other respiratory illnesses and coronavirus infections. X-rays of the thorax and Polymerase Chain Reaction with Reverse Transcription (RT-PCR) are common and efficient methods for preventing the spread of infectious diseases. In recent years, Machine Learning (ML) algorithms have been widely used to aid in the diagnosis of medical images, yielding simpler, more accurate, and quicker results. This study intends to apply texture descriptors to X-rays of COVID-19 patients’ lungs and utilize the extracted features in frameworks designed to accurately evaluate COVID-19 patients. Multiple experiments employing individual texture descriptors and their integration were conducted in order to incorporate these new characteristics into the proposed models. In addition, these frameworks will be compared to the conventional ML models used to aid in the diagnosis of COVID-19. When texture descriptors are used in conjunction with other standard features, the predictive power of the algorithms increases, according to the results. In addition, the accuracy increases when different types of texture descriptors are combined, resulting in enhanced metrics for detecting and diagnosing COVID-19.

Completeness of Resection and Long-Term Survival of Patients Undergoing Resection for Pathologic T3 NSCLC: An International Association for the Study of Lung Cancer Analysis

IntroductionCurrently, tumors with different histopathologic characteristics and oncologic outcomes comprise the T3 category of the eight edition TNM classification for lung cancers. To better understand the T3 category, we evaluated completeness of resection and long-term survival in patients undergoing resection for T3 NSCLC. MethodsThe International Association for the Study of Lung Cancer 1999 to 2010 database was queried for patients with pathologic T3N0M0 NSCLC who underwent lobectomy or pneumonectomy. The primary outcome evaluated was overall survival (OS) stratified by T3 descriptors and completeness of resection. ResultsOf 1448 patients with T3N0M0 tumors, 1187 (82.0%) had a single descriptor defining them as T3. T3 tumors with chest wall infiltration (CWI) or parietal pleura infiltration (PL3) had the highest rates of incomplete resection (9.8% and 8.4%, respectively), and those classified as T3 by size only had the lowest rate of incomplete resection (2.9%). Individual T3 descriptors were associated with significant differences in OS (p = 0.005). When tumors with similar survival and complete resection rates were grouped, patients with T3 tumors characterized by size or the presence of a separate nodule (SN) in the same lobe had better 5-year OS than patients with tumors characterized by PL3 or CWI (size/SN 60% versus CWI/PL3 53%, p = 0.017) independent of completeness of resection. ConclusionsSignificant differences in 5-year OS were associated with size, SN, PL3, or CWI T3 descriptors. Subdividing pathologic T3N0M0 tumors according to the presence or absence of CWI or PL3 may increase the prognostic accuracy of tumor staging.

Review Article. Measurement of radiologist reporting times: Assessment of precision, comparison of three different measurement techniques and review of potential applications.

Radiologist reporting times are a key component of radiology department workload assessment, but reliable measurement remains challenging. Currently, there are three contenders for this task: median reporting times (MRTs), extracted directly from a department's radiology information system (RIS); study-ascribed times (SATs), using published tables of individual descriptors derived from a combination of measurement and consensus; and radiology reporting figures (RRFs), using published tables of measured times based on modality and numbers of anatomical areas. We review these techniques, their possible uses and some potential pitfalls. We discuss the level of precision that can realistically be attained in measuring reporting times, and list the strengths and weaknesses of each technique, comparing them in relation to each of eight potential applications. We believe that SATs are challenging for practical use due to their static nature, absent common descriptors and large number. RRFs are more user-friendly but are also static and require ongoing updates; currently, they do not include ultrasound. MRTs cannot currently be extracted from every RIS, but where available they are easy to use and their dynamic nature provides the most objective data. They underestimate the unmeasurable components of a radiologist's work and therefore the total time spent in a reporting session. MRTs are superior to the other methods in flexibility, precision and ease of use. All institutions should have access to this data and we call on vendors of Radiology Information Systems which are currently not capable of providing it to make the necessary modifications.

Explainable Supervised Machine Learning Model To Predict Solvation Gibbs Energy.

Many challenges persist in developing accurate computational models for predicting solvation free energy (ΔGsol). Despite recent developments in Machine Learning (ML) methodologies that outperformed traditional quantum mechanical models, several issues remain concerning explanatory insights for broad chemical predictions with an acceptable speed-accuracy trade-off. To overcome this, we present a novel supervised ML model to predict the ΔGsol for an array of solvent-solute pairs. Using two different ensemble regressor algorithms, we made fast and accurate property predictions using open-source chemical features, encoding complex electronic, structural, and surface area descriptors for every solvent and solute. By integrating molecular properties and chemical interaction features, we have analyzed individual descriptor importance and optimized our model though explanatory information form feature groups. On aqueous and organic solvent databases, ML models revealed the predictive relevance of solutes with increasing polar surface area and decreasing polarizability, yielding better results than state-of-the-art benchmark Neural Network methods (without complex quantum mechanical or molecular dynamic simulations). Both algorithms successfully outperformed previous ΔGsol predictions methods, with a maximum absolute error of 0.22 ± 0.02 kcal mol-1, further validated in an external benchmark database and with solvent hold-out tests. With these explanatory and statistical insights, they allow a thoughtful application of this method for predicting other thermodynamic properties, stressing the relevance of ML modeling for further complex computational chemistry problems.

Decoding structure-spectrum relationships with physically organized latent spaces

A new semi-supervised machine learning method for the discovery of structure-spectrum relationships is developed and demonstrated using the specific example of interpreting X-ray absorption near-edge structure (XANES) spectra. This method constructs a one-to-one mapping between individual structure descriptors and spectral trends. Specifically, an adversarial autoencoder is augmented with a novel rank constraint (RankAAE). The RankAAE methodology produces a continuous and interpretable latent space, where each dimension can track an individual structure descriptor. As a part of this process, the model provides a robust and quantitative measure of the structure-spectrum relationship by decoupling intertwined spectral contributions from multiple structural characteristics. This makes it ideal for spectral interpretation and the discovery of new descriptors. The capability of this procedure is showcased by considering five local structure descriptors and a database of over fifty thousand simulated XANES spectra across eight first-row transition metal oxide families. The resulting structure-spectrum relationships not only reproduce known trends in the literature, but also reveal unintuitive ones that are visually indiscernible in large data sets. The results suggest that the RankAAE methodology has great potential to assist researchers to interpret complex scientific data, test physical hypotheses, and reveal new patterns that extend scientific insight.

Satellite image matching and registration using affine transformation and hybrid feature descriptors

Image registration (IR) is an image processing technique to determine geometrical transformation that gives the most accurate match between reference and floating images. In existing IR work, the inliers ratio and outlier's ratios are equal. This reduces the image registration accuracy. So, to avoid the outlier's ratio and increase the inliers ratio, hybrid invariant local features descriptor is proposed. This feature descriptor consists of binary-robust-invariant-scalable-key point (BRISK), speed-up-robust-feature (SURF), and feature from accelerated segment test feature (FAST). The hybrid feature descriptor extracts the relevant features from the image. Then, the feature matching step finds the correct correspondences from the two sets of features. Also, affine transformation avoids the false feature matching points. An experimental analysis shows the inliers ratio and repeatability evaluation metrics performance of individual feature descriptor, combined feature descriptor and proposed feature descriptor. The proposed hybrid feature descriptor achieves inliers ratio of 1.913 and repeatability is 0.121.

The Challenge of Assessing Content and Coherence

Content and coherence are the categories most difficult to evaluate fairly when raters use analytic scoring scales. Readers inevitably interpret texts in their own idiosyncratic ways, depending on their knowledge, experience, ethical considerations, and other personal biases that they cannot completely set aside when grading a text. This is also true for descriptors, which are themselves short texts. To make matters worse, due to the very nature of writing but also the lack of consensus among experts in discourse research, writing theory, and writing assessment, descriptors are categorized vaguely and inconsistently. As a result, raters seeking useful evaluation criteria are confronted with descriptors that cover the same concept, such as “relevance”, being categorized in one set of criteria as relating to the content of the written text and in another as belonging to the category of coherence. Nevertheless, the objectivity of the evaluation of written work can be increased. The article examines the relationship between content and coherence, which is reflected in the way the two concepts are defined in the relevant literature, as well as in some descriptors used in two grading scales used in Slovenia. The empirical part of the paper presents a case study involving 46 secondary school teachers, whose responses to a questionnaire confirm the subjectivity of the understanding of individual descriptors and the need for adequate training of teachers in the use of analytic scoring scales, regular standardization in the schools where they work, evaluation of the assessment scales they use and their possible adaptation.

Ultra-short chain fluorocarboxylates exhibit wide ranging reactivity with hydrated electrons

Recent reports demonstrate that technologies generating hydrated electrons (eaq−; e.g., UV-sulfite) are a promising strategy for destruction of per- and polyfluoroalkyl substances, but fundamental rate constants are lacking. This work examines the kinetics and mechanisms of eaq− reactions with ultra-short chain (C2–C4) fluorocarboxylates using experimental and theoretical approaches. Laser flash photolysis (LFP) was used to measure bimolecular rate constants (k2; M−1 s−1) for eaq− reactions with thirteen per-, and for the first time, polyfluorinated carboxylate structures. The measured k2 values varied widely from 5.26 × 106 to 1.30 × 108 M−1s−1, a large range considering the minor structural changes among the target compounds. Molecular descriptors calculated using density functional theory did not reveal correlation between k2 values and individual descriptors when considering the whole dataset, however, semiquantitative correlation manifests when grouping by similar possible initial reduction event such as electron attachment at the α-carbon versus β- or γ-carbons along the backbone. From this, it is postulated that fluorocarboxylate reduction by eaq− occurs via divergent mechanisms with the possibility of non-degradative pathways being prominent. These mechanistic insights provide rationale for contradictory trends between LFP-derived k2 values and apparent degradation rates recently reported in UV-sulfite constant irradiation treatment experiments.

Optimizing interactions to protein binding sites by integrating docking-scoring strategies into generative AI methods.

The identification and optimization of promising lead molecules is essential for drug discovery. Recently, artificial intelligence (AI) based generative methods provided complementary approaches for generating molecules under specific design constraints of relevance in drug design. The goal of our study is to incorporate protein 3D information directly into generative design by flexible docking plus an adapted protein-ligand scoring function, thereby moving towards automated structure-based design. First, the protein-ligand scoring function RFXscore integrating individual scoring terms, ligand descriptors, and combined terms was derived using the PDBbind database and internal data. Next, design results for different workflows are compared to solely ligand-based reward schemes. Our newly proposed, optimal workflow for structure-based generative design is shown to produce promising results, especially for those exploration scenarios, where diverse structures fitting to a protein binding site are requested. Best results are obtained using docking followed by RFXscore, while, depending on the exact application scenario, it was also found useful to combine this approach with other metrics that bias structure generation into "drug-like" chemical space, such as target-activity machine learning models, respectively.

Motion modeling and blurred image simulation of the star tracker used for deep-space missions

The star tracker is the most accurate attitude measurement device among different types of attitude measurement devices. It is based on captured star images to extract star centroids and then determine attitude. However, images taken by the star tracker under dynamic conditions are often blurred, which limits its dynamic performance. In this paper, we first classify and model motions of the star tracker under dynamic conditions, including rotations and angular vibrations. Then, a motion kernel consisting of three individual descriptors (trajectory, intensity, and point spread function) is proposed to model the motion blurring process and simulate blurred images. Experimental results show that our approach can model a wide variety of blurs including uniform and non-uniform blurs, and it can provide extremely realistic blurred images. These findings are crucial to the subsequent centroid extraction of stars and the dynamic performance of the star tracker.

Evaluating structure-based activity in a high-throughput assay for steroid biosynthesis

Data from a high-throughput human adrenocortical carcinoma assay (HT-H295R) for steroid hormone biosynthesis are available for > 2000 chemicals in single concentration and 654 chemicals in multi-concentration (mc). Previously, a metric describing the effect size of a chemical on the biosynthesis of 11 hormones was derived using mc data referred to as the maximum mean Mahalanobis distance (maxmMd). However, mc HT-H295R assay data remain unavailable for many chemicals. This work leverages existing HT-H295R assay data by constructing structure–activity relationships to make predictions for data-poor chemicals, including: (1) identification of individual structural descriptors, known as ToxPrint chemotypes, associated with increased odds of affecting estrogen or androgen synthesis; (2) a random forest (RF) classifier using physicochemical property descriptors to predict HT-H295R maxmMd binary (positive or negative) outcomes; and, (3) a local approach to predict maxmMd binary outcomes using nearest neighbors (NNs) based on two types of chemical fingerprints (chemotype or Morgan). Individual chemotypes demonstrated high specificity (85–98 %) for modulators of estrogen and androgen synthesis but with low sensitivity. The best RF model for maxmMd classification included 13 predicted physicochemical descriptors, yielding a balanced accuracy (BA) of 71 % with only modest improvement when hundreds of structural features were added. The best two NN models for binary maxmMd prediction demonstrated BAs of 85 and 81 % using chemotype and Morgan fingerprints, respectively. Using an external test set of 6302 chemicals (lacking HT-H295R data), 1241 were identified as putative estrogen and androgen modulators. Combined results across the three classification models (global RF model and two local NN models) predict that 1033 of the 6302 chemicals would be more likely to affect HT-H295R bioactivity. Together, these in silico approaches can efficiently prioritize thousands of untested chemicals for screening to further evaluate their effects on steroid biosynthesis.

Evaluation of the Individual Activity Descriptors of the mMRC Breathlessness Scale: A Mixed Method Study.

PurposeThe modified-Medical Research Council (mMRC) breathlessness scale consists of five grades that contain of a description of different activities. It has wide utility in the assessment of disability due to breathlessness but was originally developed before the advent of modern psychometric methodology and, for example contains more than one activity per grade. We conducted an evaluation of the mMRC structure.Patients and MethodsCognitive debriefing was conducted with COPD patients to elicit their understanding of each mMRC activity. In a cross-sectional study, patients completed the mMRC scale (grades 0–4) and an MRC-Expanded (MRC-Ex) version consisting of 10-items, each containing one mMRC activity. Each activity was then given a 4-point response scale (0 “not at all” to 4 “all of the time”) and all 10 items were given to 203 patients to complete Rasch analysis and assess the pattern of MRC item severity and its hierarchical structure.ResultsCognitive debriefing with 36 patients suggested ambiguity with the term “strenuous exercise” and perceived severity differences between mMRC activities. 203 patients completed the mMRC-Ex. Strenuous exercise was located third on the ascending severity scale. Rasch identified the mildest term was “walking up a slight hill” (logit −2.76) and “too breathless to leave the house” was the most severe (logit 3.42).ConclusionThis analysis showed that items that were combined into a single mMRC grade may be widely separated in terms of perceived severity when assessed individually. This suggests that mMRC grades as a measure of individual disability related to breathlessness contain significant ambiguity due to the combination of activities of different degrees of perceived severity into a single grade.

An Image Retrieval Framework Design Analysis Using Saliency Structure and Color Difference Histogram

This paper focuses on retrieving plant leaf images based on different features that can be useful in the plant industry. Various images and their features can be used to identify the type of leaf and its disease. For this purpose, a well-organized computer-assisted plant image retrieval approach is required that can use a hybrid combination of the color and shape attributes of leaf images for plant disease identification and botanical gardening in the agriculture sector. In this research work, an innovative framework is proposed for the retrieval of leaf images that uses a hybrid combination of color and shape features to improve retrieval accuracy. For the color features, the Color Difference Histograms (CDH) descriptor is used while shape features are determined using the Saliency Structure Histogram (SSH) descriptor. To extract the various properties of leaves, Hue and Saturation Value (HSV) color space features and First Order Statistical Features (FOSF) features are computed in CDH and SSH descriptors, respectively. After that, the HSV and FOSF features of leaf images are concatenated. The concatenated features of database images are compared with the query image in terms of the Euclidean distance and a threshold value of Euclidean distance is taken for retrieval of images. The best results are obtained at the threshold value of 80% of the maximum Euclidean distance. The system’s effectiveness is also evaluated with different performance metrics like precision, recall, and F-measure, and their values come out to be respectively 1.00, 0.96, and 0.97, which is better than individual feature descriptors.

Histologic Cell Shape Descriptors for the Retinal Pigment Epithelium in Age-Related Macular Degeneration: A Comparison to Unaffected Eyes.

PurposePhenotype alterations of the retinal pigment epithelium (RPE) are a main characteristic of age-related macular degeneration (AMD). Individual RPE cell shape descriptors may help to delineate healthy from AMD-affected cells in early disease stages.MethodsTwenty-two human RPE flatmounts (7 eyes with AMD [early, 3; geographic atrophy, 1; neovascular, 3); 15 unaffected eyes [8 aged ≤51 years; 7 aged >80 years)] were imaged at the fovea, perifovea, and near periphery (predefined sample locations) using a laser-scanning confocal fluorescence microscope. RPE cell boundaries were manually marked with computer assistance. For each cell, 11 shape descriptors were calculated and correlated with donor age, cell autofluorescence (AF) intensity, and retinal location. Statistical analysis was performed using an ensemble classifier based on logistic regression.ResultsIn AMD, RPE was altered at all locations (most pronounced at the fovea), with area, solidity, and form factor being the most discriminatory descriptors. In the unaffected macula, aging had no significant effect on cell shape factors; however, with increasing distance to the fovea, area, solidity, and convexity increased while form factor decreased. Reduced AF in AMD was significantly associated with decreased roundness and solidity.ConclusionsAMD results in an altered RPE with enlarged and deformed cells that could precede clinically visible lesions and thus serve as early biomarkers for AMD onset. Our data may also help guide the interpretation of RPE morphology in in vivo studies utilizing high-resolution single-cell imaging.Translational RelevanceOur histologic RPE cell shape data have the ability to identify robust biomarkers for the early detection of AMD-affected cells, which also could serve as a basis for automated segmentation of RPE sheets.