Conversational assessment using artificial intelligence is as clinically useful as depression scales and preferred by users.

BACKGROUND: Depression is prevalent, chronic, and burdensome. Due to limited screening access, depression often remains undiagnosed. Artificial intelligence (AI) models based on spoken responses to interview questions may offer an effective, efficient alternative to other screening methods. OBJECTIVE: The primary aim was to use a demographically diverse sample to validate an AI model, previously trained on human-administered interviews, on novel bot-administered interviews, and to check for algorithmic biases related to age, sex, race, and ethnicity. METHODS: Using the Aiberry app, adults recruited via social media (N = 393) completed a brief bot-administered interview and a depression self-report form. An AI model was used to predict form scores based on interview responses alone. For all meaningful discrepancies between model inference and form score, clinicians performed a masked review to determine which one they preferred. RESULTS: There was strong concurrent validity between the model predictions and raw self-report scores (r = 0.73, MAE = 3.3). 90 % of AI predictions either agreed with self-report or with clinical expert opinion when AI contradicted self-report. There was no differential model performance across age, sex, race, or ethnicity. LIMITATIONS: Limitations include access restrictions (English-speaking ability and access to smartphone or computer with broadband internet) and potential self-selection of participants more favorably predisposed toward AI technology. CONCLUSION: The Aiberry model made accurate predictions of depression severity based on remotely collected spoken responses to a bot-administered interview. This study shows promising results for the use of AI as a mental health screening tool on par with self-report measures.

Computer image analysis with artificial intelligence: a practical introduction to convolutional neural networks for medical professionals.

Artificial intelligence tools, particularly convolutional neural networks (CNNs), are transforming healthcare by enhancing predictive, diagnostic, and decision-making capabilities. This review provides an accessible and practical explanation of CNNs for clinicians and highlights their relevance in medical image analysis. CNNs have shown themselves to be exceptionally useful in computer vision, a field that enables machines to 'see' and interpret visual data. Understanding how these models work can help clinicians leverage their full potential, especially as artificial intelligence continues to evolve and integrate into healthcare. CNNs have already demonstrated their efficacy in diverse medical fields, including radiology, histopathology, and medical photography. In radiology, CNNs have been used to automate the assessment of conditions such as pneumonia, pulmonary embolism, and rectal cancer. In histopathology, CNNs have been used to assess and classify colorectal polyps, gastric epithelial tumours, as well as assist in the assessment of multiple malignancies. In medical photography, CNNs have been used to assess retinal diseases and skin conditions, and to detect gastric and colorectal polyps during endoscopic procedures. In surgical laparoscopy, they may provide intraoperative assistance to surgeons, helping interpret surgical anatomy and demonstrate safe dissection zones. The integration of CNNs into medical image analysis promises to enhance diagnostic accuracy, streamline workflow efficiency, and expand access to expert-level image analysis, contributing to the ultimate goal of delivering further improvements in patient and healthcare outcomes.

Deep Transfer Learning-Based Approach for Glucose Transporter-1 (GLUT1) Expression Assessment.

Glucose transporter-1 (GLUT-1) expression level is a biomarker of tumour hypoxia condition in immunohistochemistry (IHC)-stained images. Thus, the GLUT-1 scoring is a routine procedure currently employed for predicting tumour hypoxia markers in clinical practice. However, visual assessment of GLUT-1 scores is subjective and consequently prone to inter-pathologist variability. Therefore, this study proposes an automated method for assessing GLUT-1 scores in IHC colorectal carcinoma images. For this purpose, we leverage deep transfer learning methodologies for evaluating the performance of six different pre-trained convolutional neural network (CNN) architectures: AlexNet, VGG16, GoogleNet, ResNet50, DenseNet-201 and ShuffleNet. The target CNNs are fine-tuned as classifiers or adapted as feature extractors with support vector machine (SVM) to classify GLUT-1 scores in IHC images. Our experimental results show that the winning model is the trained SVM classifier on the extracted deep features fusion Feat-Concat from DenseNet201, ResNet50 and GoogLeNet extractors. It yields the highest prediction accuracy of 98.86%, thus outperforming the other classifiers on our dataset. We also conclude, from comparing the methodologies, that the off-the-shelf feature extraction is better than the fine-tuning model in terms of time and resources required for training.

Using Personalized Avatars as an Adjunct to an Adult Weight Loss Management Program: Randomized Controlled Feasibility Study.

BACKGROUND: Obesity is a global public health concern. Interventions rely predominantly on managing dietary intake and increasing physical activity; however, sustained adherence to behavioral regimens is often poor. The lack of sustained motivation, self-efficacy, and poor adherence to behavioral regimens are recognized barriers to successful weight loss. Avatar-based interventions achieve better patient outcomes in the management of chronic conditions by promoting more active engagement. Virtual representations of self can affect real-world behavior, acting as a catalyst for sustained weight loss behavior. OBJECTIVE: We evaluated whether a personalized avatar, offered as an adjunct to an established weight loss program, can increase participant motivation, sustain engagement, optimize service delivery, and improve participant health outcomes. METHODS: A feasibility randomized design was used to determine the case for future development and evaluation of avatar-based technology in a randomized controlled trial. Participants were recruited from general practitioner referrals to a 12-week National Health Service weight improvement program. The main outcome measure was weight loss. Secondary outcome measures were quality-of-life and self-efficacy. Quantitative data were subjected to descriptive statistical tests and exploratory comparison between intervention and control arms. Feasibility and acceptability were assessed through interviews and analyzed using framework approach. Health Research Authority ethics approval was granted. RESULTS: Overall, 10 men (n=7, 70% for routine care and avatar and n=3, 30% for routine care) and 33 women (n=23, 70% for intervention and n=10, 30% for routine care) were recruited. Participants' initial mean weight was greater in the intervention arm than in the routine care arm (126.3 kg vs 122.9 kg); pattern of weight loss was similar across both arms of the study in T0 to T1 period but accelerated in T1 to T2 period for intervention participants, suggesting that access to the self-resembling avatar may promote greater engagement with weight loss initiatives in the short-to-medium term. Mean change in participants' weight from T0 to T2 was 4.5 kg (95% CI 2.7-6.3) in the routine care arm and 5.3 kg (95% CI 3.9-6.8) in the intervention arm. Quality-of-life and self-efficacy measures demonstrated greater improvement in the intervention arm at both T1 (105.5 for routine care arm and 99.7 for intervention arm) and T2 (100.1 for routine care arm and 81.2 for intervention arm). Overall, 13 participants (n=11, 85% women and n=2, 15% men) and two health care professionals were interviewed about their experience of using the avatar program. CONCLUSIONS: Participants found using the personalized avatar acceptable, and feedback reiterated that seeing a future self helped to reinforce motivation to change behavior. This feasibility study demonstrated that avatar-based technology may successfully promote engagement and motivation in weight loss programs, enabling participants to achieve greater weight loss gains and build self-confidence. TRIAL REGISTRATION: ISRCTN Registry 17953876; https://doi.org/10.1186/ISRCTN17953876.

Social distancing enhanced automated optimal design of physical spaces in the wake of the COVID-19 pandemic.

As the COVID-19 pandemic unfolds, manually enhanced ad-hoc solutions have helped the physical space designers and decision makers to cope with the dynamic nature of space planning. Due to the unpredictable nature by which the pandemic is unfolding, the standard operating procedures also change, and the protocols for physical interaction require continuous reconsideration. Consequently, the development of an appropriate technological solution to address the current challenge of reconfiguring common physical environments with prescribed physical distancing measures is much needed. To do this, we propose a design optimization methodology which takes the dimensions, as well as the constraints and other necessary requirements of a given physical space to yield optimal redesign solutions on the go. The methodology we propose here utilizes the solution to the well-known mathematical circle packing problem, which we define as a constrained mathematical optimization problem. The resulting optimization problem is solved subject to a given set of parameters and constraints - corresponding to the requirements on the social distancing criteria between people and the imposed constraints on the physical spaces such as the position of doors, windows, walkways and the variables related to the indoor airflow pattern. Thus, given the dimensions of a physical space and other essential requirements, the solution resulting from the automated optimization algorithm can suggest an optimal set of redesign solutions from which a user can pick the most feasible option. We demonstrate our automated optimal design methodology by way of a number of practical examples, and we discuss how this framework can be further taken forward as a design platform that can be implemented practically.

A Study of Deep Learning-Based Face Recognition Models for Sibling Identification.

Accurate identification of siblings through face recognition is a challenging task. This is predominantly because of the high degree of similarities among the faces of siblings. In this study, we investigate the use of state-of-the-art deep learning face recognition models to evaluate their capacity for discrimination between sibling faces using various similarity indices. The specific models examined for this purpose are FaceNet, VGGFace, VGG16, and VGG19. For each pair of images provided, the embeddings have been calculated using the chosen deep learning model. Five standard similarity measures, namely, cosine similarity, Euclidean distance, structured similarity, Manhattan distance, and Minkowski distance, are used to classify images looking for their identity on the threshold defined for each of the similarity measures. The accuracy, precision, and misclassification rate of each model are calculated using standard confusion matrices. Four different experimental datasets for full-frontal-face, eyes, nose, and forehead of sibling pairs are constructed using publicly available HQf subset of the SiblingDB database. The experimental results show that the accuracy of the chosen deep learning models to distinguish siblings based on the full-frontal-face and cropped face areas vary based on the face area compared. It is observed that VGGFace is best while comparing the full-frontal-face and eyes-the accuracy of classification being with more than 95% in this case. However, its accuracy degrades significantly when the noses are compared, while FaceNet provides the best result for classification based on the nose. Similarly, VGG16 and VGG19 are not the best models for classification using the eyes, but these models provide favorable results when foreheads are compared.

The role of mobile phones as a possible pathway for pathogen movement, a cross-sectional microbial analysis.

INTRODUCTION: Mobile phones are used the world over, including in healthcare settings. This study aimed to investigate the viable microbial colonisation of mobile phones used by healthcare personnel. METHODS: Swabs collected on the same day from 30 mobile phones belonging to healthcare workers from three separate paediatric wards of an Australian hospital were cultured on five types of agar plate, then colonies from each phone were pooled, extracted and sequenced by shotgun metagenomics. Questionnaires completed by staff whose phones were sampled assisted in the analysis and interpretation of results. RESULTS AND DISCUSSION: All phones sampled cultured viable bacteria. Overall, 399 bacterial operational taxonomic units were identified from 30 phones, with 1432 cumulative hits. Among these were 58 recognised human pathogenic and commensal bacteria (37 Gram-negative, 21 Gram-positive). The total number of virulence factor genes detected was 347, with 1258 cumulative hits. Antibiotic resistance genes (ARGs) were detected on all sampled phones and overall, 133 ARGs were detected with 520 cumulative hits. The most important classes of ARGs detected encoded resistance to beta-lactam, aminoglycoside and macrolide antibiotics and efflux pump mediated resistance mechanisms. CONCLUSION: Mobile phones carry viable bacterial pathogens and may act as fomites by contaminating the hands of their users and indirectly providing a transmission pathway for hospital-acquired infections and dissemination of antibiotic resistance. Further research is needed, but meanwhile adding touching mobile phones to the five moments of hand hygiene is a simple infection control strategy worth considering in hospital and community settings. Additionally, the implementation of practical and effective guidelines to decontaminate mobile phone devices would likely be beneficial to the hospital population and community at large.

Facial Age Synthesis Using Sparse Partial Least Squares (The Case of Ben Needham).

Automatic facial age progression (AFAP) has been an active area of research in recent years. This is due to its numerous applications which include searching for missing. This study presents a new method of AFAP. Here, we use an active appearance model (AAM) to extract facial features from available images. An aging function is then modelled using sparse partial least squares regression (sPLS). Thereafter, the aging function is used to render new faces at different ages. To test the accuracy of our algorithm, extensive evaluation is conducted using a database of 500 face images with known ages. Furthermore, the algorithm is used to progress Ben Needham's facial image that was taken when he was 21 months old to the ages of 6, 14, and 22 years. The algorithm presented in this study could potentially be used to enhance the search for missing people worldwide.

The use of thermographic imaging to evaluate therapeutic response in human tumour xenograft models.

Non-invasive methods to monitor tumour growth are an important goal in cancer drug development. Thermographic imaging systems offer potential in this area, since a change in temperature is known to be induced due to changes within the tumour microenvironment. This study demonstrates that this imaging modality can be applied to a broad range of tumour xenografts and also, for the first time, the methodology's suitability to assess anti-cancer agent efficacy. Mice bearing subcutaneously implanted H460 lung cancer xenografts were treated with a novel vascular disrupting agent, ICT-2552, and the cytotoxin doxorubicin. The effects on tumour temperature were assessed using thermographic imaging over the first 6 hours post-administration and subsequently a further 7 days. For ICT-2552 a significant initial temperature drop was observed, whilst for both agents a significant temperature drop was seen compared to controls over the longer time period. Thus thermographic imaging can detect functional differences (manifesting as temperature reductions) in the tumour response to these anti-cancer agents compared to controls. Importantly, these effects can be detected in the first few hours following treatment and therefore the tumour is observable non-invasively. As discussed, this technique will have considerable 3Rs benefits in terms of reduction and refinement of animal use.

Elastic-plastic contact law for simulation of tablet crushing using the biharmonic equation.

This work presents a technique for shape modelling of cylindrical and spherical tablets subject to compression. This technique is based on the use of partial differential equations (PDEs), the biharmonic equation in particular. The deformation of the compressed elastic-plastic tablet of both shapes was obtained using the existing contact models found in literature. The mathematical properties of the biharmonic equation have been exploited to achieve simple mathematical expressions characterising the shape of the distorted tablet. Thus, the height, radius and contact area of both configurations due to uniaxial compression are represented by analytic expressions relating the coefficients associated with the solution of the biharmonic equation. The results obtained from the PDE-based simulation are compared with the theoretical ones. It is found that the analytic solution of the elliptic PDE can be utilised to represent the physical changes of the deformed object.

Method of modelling the compaction behaviour of cylindrical pharmaceutical tablets.

The mechanisms involved for compaction of pharmaceutical powders have become a crucial step in the development cycle for robust tablet design with required properties. Compressibility of pharmaceutical materials is measured by a force-displacement relationship which is commonly analysed using a well known method, the Heckel model. This model requires the true density and compacted powder mass value to determine the powder mean yield pressure. In this paper, we present a technique for shape modelling of pharmaceutical tablets based on the use of partial differential equations (PDEs). This work also presents an extended formulation of the PDE method to a higher dimensional space by increasing the number of parameters responsible for describing the surface in order to generate a solid tablet. Furthermore, the volume and the surface area of the parametric cylindrical tablet have been estimated numerically. Finally, the solution of the axisymmetric boundary value problem for a finite cylinder subject to a uniform axial load has been utilised in order to model the displacement components of a compressed PDE-based representation of a tablet. The Heckel plot obtained from the developed model shows that the model is capable of predicting the compaction behaviour of pharmaceutical materials since it fits the experimental data accurately.

Modelling of oedemous limbs and venous ulcers using partial differential equations.

BACKGROUND: Oedema, commonly known as tissue swelling, occurs mainly on the leg and the arm. The condition may be associated with a range of causes such as venous diseases, trauma, infection, joint disease and orthopaedic surgery. Oedema is caused by both lymphatic and chronic venous insufficiency, which leads to pooling of blood and fluid in the extremities. This results in swelling, mild redness and scaling of the skin, all of which can culminate in ulceration. METHODS: We present a method to model a wide variety of geometries of limbs affected by oedema and venous ulcers. The shape modelling is based on the PDE method where a set of boundary curves are extracted from 3D scan data and are utilised as boundary conditions to solve a PDE, which provides the geometry of an affected limb. For this work we utilise a mixture of fourth order and sixth order PDEs, the solutions of which enable us to obtain a good representative shape of the limb and associated ulcers in question. RESULTS: A series of examples are discussed demonstrating the capability of the method to produce good representative shapes of limbs by utilising a series of curves extracted from the scan data. In particular we show how the method could be used to model the shape of an arm and a leg with an associated ulcer. CONCLUSION: We show how PDE based shape modelling techniques can be utilised to generate a variety of limb shapes and associated ulcers by means of a series of curves extracted from scan data. We also discuss how the method could be used to manipulate a generic shape of a limb and an associated wound so that the model could be fine-tuned for a particular patient.