A systematic review and meta-analysis of artificial intelligence versus clinicians for skin cancer diagnosis

Scientific research of artificial intelligence (AI) in dermatology has increased exponentially. The objective of this study was to perform a systematic review and meta-analysis to evaluate the performance of AI algorithms for skin cancer classification in comparison to clinicians with different levels of expertise. Based on PRISMA guidelines, 3 electronic databases (PubMed, Embase, and Cochrane Library) were screened for relevant articles up to August 2022. The quality of the studies was assessed using QUADAS-2. A meta-analysis of sensitivity and specificity was performed for the accuracy of AI and clinicians. Fifty-three studies were included in the systematic review, and 19 met the inclusion criteria for the meta-analysis. Considering all studies and all subgroups of clinicians, we found a sensitivity (Sn) and specificity (Sp) of 87.0% and 77.1% for AI algorithms, respectively, and a Sn of 79.78% and Sp of 73.6% for all clinicians (overall); differences were statistically significant for both Sn and Sp. The difference between AI performance (Sn 92.5%, Sp 66.5%) vs. generalists (Sn 64.6%, Sp 72.8%), was greater, when compared with expert clinicians. Performance between AI algorithms (Sn 86.3%, Sp 78.4%) vs expert dermatologists (Sn 84.2%, Sp 74.4%) was clinically comparable. Limitations of AI algorithms in clinical practice should be considered, and future studies should focus on real-world settings, and towards AI-assistance.

Identification of hemodynamic biomarkers for bicuspid aortic valve induced aortic dilation using machine learning

Recent advances in medical imaging have confirmed the presence of altered hemodynamics in bicuspid aortic valve (BAV) patients. Therefore, there is a need for new hemodynamic biomarkers to refine disease monitoring and improve patient risk stratification. This research aims to analyze and extract multiple correlation patterns of hemodynamic parameters from 4D Flow MRI data and find which parameters allow an accurate classification between healthy volunteers (HV) and BAV patients with dilated and non-dilated ascending aorta using machine learning. Sixteen hemodynamic parameters were calculated in the ascending aorta (AAo) and aortic arch (AArch) at peak systole from 4D Flow MRI. We used sequential forward selection (SFS) and principal component analysis (PCA) as feature selection algorithms. Then, eleven machine-learning classifiers were implemented to separate HV and BAV patients (non- and dilated ascending aorta). Multiple correlation patterns from hemodynamic parameters were extracted using hierarchical clustering. The linear discriminant analysis and random forest are the best performing classifiers, using five hemodynamic parameters selected with SFS (velocity angle, forward velocity, vorticity, and backward velocity in AAo; and helicity density in AArch) a 96.31 ± 1.76% and 96.00 ± 0.83% accuracy, respectively. Hierarchical clustering revealed three groups of correlated features. According to this analysis, we observed that features selected by SFS have a better performance than those selected by PCA because the five selected parameters were distributed according to 3 different clusters. Based on the proposed method, we concluded that the feature selection method found five potentially hemodynamic biomarkers related to this disease.

Characterization of spinal cord damage based on automatic video analysis of froglet swimming

Xenopus laevis frogs are a widely used organism to study aspects of modern biology (Harland and Grainger, 2011). Its central nervous system is particularly interesting, because in certain stages of metamorphosis the spinal cord can regenerate after injury and recover swimming. With this in mind, automatic gait analysis could help evaluate the regenerative performance by means of a method that automatically and quantitatively establishes the degree in froglets' limb movement. Here, we present an algorithm that characterizes spinal cord damage in froglets. The proposed method tracks the position of the limbs throughout videos and extracts kinematic features, which posteriorly serve to differentiate froglets with different levels of damage to the spinal cord. The detection algorithm and kinematic features chosen were validated in a pattern recognition experiment in which 90 videos (divided equally in three classes: uninjured, hemisected and transected) were classified. We conclude that our system is effective in the characterization of damage to the spinal cord through video analysis of a swimming froglet with a 97% accuracy. These results potentially validate this methodology to automatically compare the recovery of spinal cord function after different treatments without the need to manually process videos. In addition, the procedure could be used to measure the kinematics and behavioral response of froglets to different experimental conditions such as nutritional state, stress, genetic background and age.

Summary: Automatic video tracking system to measure froglet swimming and characterize the degree of spinal cord damage after spinal cord injury. It can potentially be used to compare the recovery of spinal cord function after different treatments.

Palaeopermeability anisotropy and geometrical properties of sealed-microfractures from micro-CT analyses: An open-source implementation

Fault zone permeability and the real 3D-spatial distribution of the fault-related fracture networks are critical in the assessment of fault zones behavior for fluids. The study of the real 3D-spatial distribution of the microfracture network, using X-ray micro-computed tomography, is a crucial factor to unravel the real structural permeability conditions of a fault-zone. Despite the availability of several commercial software for rock properties estimation from X-ray micro-computed tomography scanning, their high cost and lack of programmability encourage the use of open-source data treatment. This work presents the implementation of a methodology flow for the quantification of both structural and geometrical parameters (fractures density, fractures aperture, fractures porosity, and fractures surface area), and the modeling of palaeopermeability of fault-related fractured samples, with focus in the proper spatial orientation of both the sample and the results. This is performed with an easy to follow step-by-step implementation, by a combination of open-source software, newly implemented codes, and numerical methods. This approach keeps track of the sample's spatial orientation from the physical to the virtual world, thus assessing any fault-related palaeopermeability anisotropy.

Modeling Search Behaviors during the Acquisition of Expertise in a Sequential Decision-Making Task

Our daily interaction with the world is plagued of situations in which we develop expertise through self-motivated repetition of the same task. In many of these interactions, and especially when dealing with computer and machine interfaces, we must deal with sequences of decisions and actions. For instance, when drawing cash from an ATM machine, choices are presented in a step-by-step fashion and a specific sequence of choices must be performed in order to produce the expected outcome. But, as we become experts in the use of such interfaces, is it possible to identify specific search and learning strategies? And if so, can we use this information to predict future actions? In addition to better understanding the cognitive processes underlying sequential decision making, this could allow building adaptive interfaces that can facilitate interaction at different moments of the learning curve. Here we tackle the question of modeling sequential decision-making behavior in a simple human-computer interface that instantiates a 4-level binary decision tree (BDT) task. We record behavioral data from voluntary participants while they attempt to solve the task. Using a Hidden Markov Model-based approach that capitalizes on the hierarchical structure of behavior, we then model their performance during the interaction. Our results show that partitioning the problem space into a small set of hierarchically related stereotyped strategies can potentially capture a host of individual decision making policies. This allows us to follow how participants learn and develop expertise in the use of the interface. Moreover, using a Mixture of Experts based on these stereotyped strategies, the model is able to predict the behavior of participants that master the task.

The impact of MEG source reconstruction method on source-space connectivity estimation: A comparison between minimum-norm solution and beamforming

Despite numerous important contributions, the investigation of brain connectivity with magnetoencephalography (MEG) still faces multiple challenges. One critical aspect of source-level connectivity, largely overlooked in the literature, is the putative effect of the choice of the inverse method on the subsequent cortico-cortical coupling analysis. We set out to investigate the impact of three inverse methods on source coherence detection using simulated MEG data. To this end, thousands of randomly located pairs of sources were created. Several parameters were manipulated, including inter- and intra-source correlation strength, source size and spatial configuration. The simulated pairs of sources were then used to generate sensor-level MEG measurements at varying signal-to-noise ratios (SNR). Next, the source level power and coherence maps were calculated using three methods (a) L2-Minimum-Norm Estimate (MNE), (b) Linearly Constrained Minimum Variance (LCMV) beamforming, and (c) Dynamic Imaging of Coherent Sources (DICS) beamforming. The performances of the methods were evaluated using Receiver Operating Characteristic (ROC) curves. The results indicate that beamformers perform better than MNE for coherence reconstructions if the interacting cortical sources consist of point-like sources. On the other hand, MNE provides better connectivity estimation than beamformers, if the interacting sources are simulated as extended cortical patches, where each patch consists of dipoles with identical time series (high intra-patch coherence). However, the performance of the beamformers for interacting patches improves substantially if each patch of active cortex is simulated with only partly coherent time series (partial intra-patch coherence). These results demonstrate that the choice of the inverse method impacts the results of MEG source-space coherence analysis, and that the optimal choice of the inverse solution depends on the spatial and synchronization profile of the interacting cortical sources. The insights revealed here can guide method selection and help improve data interpretation regarding MEG connectivity estimation.