Applying the UTAUT2 framework to patients' attitudes toward healthcare task shifting with artificial intelligence.

BACKGROUND: Increasing patient loads, healthcare inflation and ageing population have put pressure on the healthcare system. Artificial intelligence and machine learning innovations can aid in task shifting to help healthcare systems remain efficient and cost effective. To gain an understanding of patients' acceptance toward such task shifting with the aid of AI, this study adapted the Unified Theory of Acceptance and Use of Technology 2 (UTAUT2), looking at performance and effort expectancy, facilitating conditions, social influence, hedonic motivation and behavioural intention. METHODS: This was a cross-sectional study which took place between September 2021 to June 2022 at the National Heart Centre, Singapore. One hundred patients, aged ≥ 21 years with at least one heart failure symptom (pedal oedema, New York Heart Association II-III effort limitation, orthopnoea, breathlessness), who presented to the cardiac imaging laboratory for physician-ordered clinical echocardiogram, underwent both echocardiogram by skilled sonographers and the experience of echocardiogram by a novice guided by AI technologies. They were then given a survey which looked at the above-mentioned constructs using the UTAUT2 framework. RESULTS: Significant, direct, and positive effects of all constructs on the behavioral intention of accepting the AI-novice combination were found. Facilitating conditions, hedonic motivation and performance expectancy were the top 3 constructs. The analysis of the moderating variables, age, gender and education levels, found no impact on behavioral intention. CONCLUSIONS: These results are important for stakeholders and changemakers such as policymakers, governments, physicians, and insurance companies, as they design adoption strategies to ensure successful patient engagement by focusing on factors affecting the facilitating conditions, hedonic motivation and performance expectancy for AI technologies used in healthcare task shifting.

Explainable Risk Prediction of Post-Stroke Adverse Mental Outcomes Using Machine Learning Techniques in a Population of 1780 Patients.

Post-stroke depression and anxiety, collectively known as post-stroke adverse mental outcome (PSAMO) are common sequelae of stroke. About 30% of stroke survivors develop depression and about 20% develop anxiety. Stroke survivors with PSAMO have poorer health outcomes with higher mortality and greater functional disability. In this study, we aimed to develop a machine learning (ML) model to predict the risk of PSAMO. We retrospectively studied 1780 patients with stroke who were divided into PSAMO vs. no PSAMO groups based on results of validated depression and anxiety questionnaires. The features collected included demographic and sociological data, quality of life scores, stroke-related information, medical and medication history, and comorbidities. Recursive feature elimination was used to select features to input in parallel to eight ML algorithms to train and test the model. Bayesian optimization was used for hyperparameter tuning. Shapley additive explanations (SHAP), an explainable AI (XAI) method, was applied to interpret the model. The best performing ML algorithm was gradient-boosted tree, which attained 74.7% binary classification accuracy. Feature importance calculated by SHAP produced a list of ranked important features that contributed to the prediction, which were consistent with findings of prior clinical studies. Some of these factors were modifiable, and potentially amenable to intervention at early stages of stroke to reduce the incidence of PSAMO.

Relation Between Neck Skin Temperature Measurement and Carotid Artery Stenosis: In-Vitro Evaluation.

Carotid artery stenosis is a form of atherosclerosis, where thrombus formation restricts the passage of blood through the carotid artery leading to irreversible damage in the brain tissue. The presence of stenosis in the carotid artery results in abnormal temperature maps on the external skin surface, which can be captured and quantified using noncontact/noninvasive infrared (IR) thermal imaging/thermography. In this study, a thermally charged in vitro carotid artery flow loop, using 0% and 75% stenosis models, was designed to study the thermal effect on the external skin surface. The carotid artery flow was encapsulated with polydimethylsiloxane (PDMS) resembling neck tissue, of which the external surface temperature maps were studied using IR thermography. Using the mean temperature as a threshold value, the resultant thermal image was processed and normalized. Between the two stenosis models, disruption in the thermal features corresponding to the presence of stenosis was observed. The method described in this study paves the path to experimentally study the thermal effect of the presence of stenosis in the carotid artery.

Imaging modalities to diagnose carotid artery stenosis: progress and prospect.

In the past few decades, imaging has been developed to a high level of sophistication. Improvements from one-dimension (1D) to 2D images, and from 2D images to 3D models, have revolutionized the field of imaging. This not only helps in diagnosing various critical and fatal diseases in the early stages but also contributes to making informed clinical decisions on the follow-up treatment profile. Carotid artery stenosis (CAS) may potentially cause debilitating stroke, and its accurate early detection is therefore important. In this paper, the technical development of various CAS diagnosis imaging modalities and its impact on the clinical efficacy is thoroughly reviewed. These imaging modalities include duplex ultrasound (DUS), computed tomography angiography (CTA) and magnetic resonance angiography (MRA). For each of the imaging modalities considered, imaging methodology (principle), critical imaging parameters, and the extent of imaging the vulnerable plaque are discussed. DUS is usually the initial recommended CAS diagnostic examination. However, for the therapeutic intervention, either MRA or CTA is recommended for confirmation, and for added information on intracranial cerebral circulation and aortic arch condition for procedural planning. Over the past few decades, the focus of CAS diagnosis has also shifted from pure stenosis quantification to plaque characterization. This has led to further advancement in the existing imaging tools and development of other potential imaging tools like Optical coherence tomography (OCT), photoacoustic tomography (PAT), and infrared (IR) thermography.

Design and validation of dual-point time-differentiated photoplethysmogram (2PPG) wearable for cuffless blood pressure estimation.

BACKGROUND & OBJECTIVES: Measurement of blood pressure (BP) in ambulatory patients is crucial for at high-risk cardiovascular patients. A non-obtrusive, non-occluding device that continuously measures BP via photoplethysmography will enable long-term ambulatory assessment of BP. The aim of this study is to validate the metasense 2PPG cuffless wearable design for continuous BP estimation without ECG. METHODS: A customized high-speed electronic optical sensor architecture with laterally spaced reflectance pulse oximetry was designed into a simple unobtrusive low-power wearable in the form of a watch. 78 volunteers with a mean age of 32.72 ± 7.4 years (21 to 64), 51% male, 49% female were recruited with ECG-2PPG signals acquired. The fiducial features of the 2PPG morphologies were then attributed to the estimator. A 9-1 K-fold cross-validation was applied in the ML. RESULTS: The correlation for PTT-SBP was 0.971 and for PTT-DBP was 0.954. The mean absolute error was 3.167±1.636 mmHg for SBP and 6.4 ± 3.9 mm Hg for DBP. The ambulatory estimate for SBP and DBP for an individual over 3 days with 8-hour recordings was 0.70-0.81 for SBP and 0.42-0.51 for DBP with a ± 2.65 mmHg for SBP and ±2.02 mmHg for DBP. For SBP, 98% of metasense measurements were within 15 mm Hg and for DBP, 91% of metasense measurements were within 10 mmHg CONCLUSIONS: The metasense device provides continuous, non-invasive BP estimations that are comparable to ambulatory BP meters. The portability and unobtrusiveness of this device, as well as the ability to continuously measure BP could one day enable long-term ambulatory BP measurement for precision cardiovascular therapeutic regimens.

Physics-informed neural network for fast prediction of temperature distributions in cancerous breasts as a potential efficient portable AI-based diagnostic tool.

This work presents the development of a novel Physics-Informed Neural Network (PINN) method for fast forward simulation of heat transfer through cancerous breast models. The proposed PINN method combines deep learning and physical principles to predict the temperature distributions in breast tissues and identify potential abnormal regions indicating the presence of tumors. The PINN model is normally trained by physics in terms of the residuals of the heat transfer equation, as well as boundary conditions with and without datasets of surface thermal imaging data concerning cancerous breast tissues, which can be used for future inverse thermal modeling to calculate tumor sizes and locations. The model is validated by comparing its predictions with those obtained by traditional Finite Element Analysis (FEA) for various cases. The comparison validates the PINN model as an accurate and fast method for thermal modeling and breast cancer diagnostic tool as the PINN simulation is found to be around 12 times faster than its FEM counterpart. The utilization of deep learning and physical principles in a diagnostic tool provides a non-invasive and safer alternative for breast self-examination compared to traditional methods such as mammography. These findings hold promise for the ongoing development of a new portable Artificial Intelligence (AI) tool for the early detection of breast cancer in breast self-examination as promoted by WHO, which is crucial for reducing mortality rates of breast cancer in the world.

Applied machine learning for blood pressure estimation using a small, real-world electrocardiogram and photoplethysmogram dataset.

Applying machine learning techniques to electrocardiography and photoplethysmography signals and their multivariate-derived waveforms is an ongoing effort to estimate non-occlusive blood pressure. Unfortunately, real ambulatory electrocardiography and photoplethysmography waveforms are inevitably affected by motion and noise artifacts, so established machine learning architectures perform poorly when trained on data of the Multiparameter Intelligent Monitoring in Intensive Care II type, a publicly available ICU database. Our study addresses this problem by applying four well-established machine learning methods, i.e., random forest regression, support vector regression, Adaboost regression and artificial neural networks, to a small, self-sampled electrocardiography-photoplethysmography dataset (n = 54) to improve the robustness of machine learning to real-world BP estimates. We evaluated the performance using a selection of optimal feature morphologies of waveforms by using pulse arrival time, morphological and frequency photoplethysmography parameters and heart rate variability as characterization data. On the basis of the root mean square error and mean absolute error, our study showed that support vector regression gave the best performance for blood pressure estimation from noisy data, achieving an mean absolute error of 6.97 mmHg, which meets the level C criteria set by the British Hypertension Society. We demonstrate that ambulatory electrocardiography- photoplethysmography signals acquired by mobile discrete devices can be used to estimate blood pressure.

An Integrated Intelligent System for Breast Cancer Detection at Early Stages Using IR Images and Machine Learning Methods with Explainability.

Breast cancer is the second most common cause of death among women. An early diagnosis is vital for reducing the fatality rate in the fight against breast cancer. Thermography could be suggested as a safe, non-invasive, non-contact supplementary method to diagnose breast cancer and can be the most promising method for breast self-examination as envisioned by the World Health Organization (WHO). Moreover, thermography could be combined with artificial intelligence and automated diagnostic methods towards a diagnosis with a negligible number of false positive or false negative results. In the current study, a novel intelligent integrated diagnosis system is proposed using IR thermal images with Convolutional Neural Networks and Bayesian Networks to achieve good diagnostic accuracy from a relatively small dataset of images and data. We demonstrate the juxtaposition of transfer learning models such as ResNet50 with the proposed combination of BNs with artificial neural network methods such as CNNs which provides a state-of-the-art expert system with explainability. The novelties of our methodology include: (i) the construction of a diagnostic tool with high accuracy from a small number of images for training; (ii) the features extracted from the images are found to be the appropriate ones leading to very good diagnosis; (iii) our expert model exhibits interpretability, i.e., one physician can understand which factors/features play critical roles for the diagnosis. The results of the study showed an accuracy that varies for the most successful models amongst four implemented approaches from approximately 91% to 93%, with a precision value of 91% to 95%, sensitivity from 91% to 92 %, and with specificity from 91% to 97%. In conclusion, we have achieved accurate diagnosis with understandability with the novel integrated approach.

Dwarf Kingfisher-Inspired Bionic Flapping Wing and Its Aerodynamic Performance at Lowest Flight Speed.

This paper aims to understand the aerodynamic performance of a bio-inspired flapping-wing model using the dwarf Kingfisher wing as the bionic reference. The paper demonstrates the numerical investigation of the Kingfisher-inspired flapping-wing followed by experimental validation to comprehend the results fully and examine the aerodynamic characteristics at a flight velocity of 4.4 m/s, with wingbeat frequencies of 11 Hz, 16 Hz, and 21 Hz, at various angles of rotation ranging from 0° to 20° for each stroke cycle. The motivation to study the performance at low speed is based on lift generation as a challenge at low speed as per quasi-steady theory. The temporal evolution of the mean force coefficients has been plotted for various angles of rotation. The results show amplification of the maximum value for the cycle average lift and drag coefficient as the rotation angle increases. The history of vertical force and the flow patterns around the wing is captured in a full cycle with asymmetric lift development in a single stroke cycle. It is observed from the results that the downstroke generates more lift force in magnitude compared to the upstroke. In addition to the rotation angle, lift asymmetry is also affected by wing-wake interaction. Experimental results reveal that there is a stable leading-edge vortex developed in the downstroke, which sheds during the upstroke. An optimum lift and thrust flapping flight can be achieved, with a lift coefficient of 3.45 at 12°. The experimental and parametric study results also reveal the importance of passive rotation in wings for aerodynamic performance and wing flexibility as an important factor for lift generation.

A method to produce high contrast vein visualization in active dynamic thermography (ADT).

In this study, a method that will aid in the visualization of vein topology on a target area on the body of a human subject is demonstrated. An external cooling means is configured to cool the left forearm of fourteen study participants, effecting an active thermal change or recovery in the target area upon removal of cooling. An infrared (IR) thermal camera was used to capture a series of transient thermal images. These images were then processed to extract Dynamic synthetic images (SI) throughout the active thermal change or recovery process. Dynamic SI was calculated using a quantitative parameter called tissue activity ratio (TAR), which is defined by the rate of rewarming to the rate of cooling at each pixel of interest. A fixed step size of rewarming temperature (0.5 °C) was used to progressively extract multiple synthetic images throughout the whole recovery process. Compared to a Static SI extraction method, where only a single SI results from the whole active dynamic thermography (ADT) sequence, this study demonstrates a live feed of high contrast vein visualizations by using the Dynamic SI method. Furthermore, the dependency of Dynamic SI contrast on the temperature of the external cooling stimulation was investigated. Three cooling stimulation temperatures (5 °C, 8 °C, and 11 °C) were tested, where no statistically significant difference in the resulting SI contrast was found. Lastly, a discussion is put forth on assisting venipuncture or cannulation-based clinical applications, through the incorporation of the proposed method with a projection system.