Identification of Key Breast Features Using a Neural Network: Applications of Machine Learning in the Clinical Setting of Plastic Surgery.

BACKGROUND: In plastic surgery, evaluation of breast symmetry is an important aspect of clinical practice. Computer programs have been developed for this purpose, but most of them require operator input. Artificial intelligence has been introduced into many aspects of medicine. In plastic surgery, automated neural networks for breast evaluation could improve quality of care. In this work, the authors evaluate the identification of breast features with an ad hoc trained neural network. METHODS: An ad hoc convolutional neural network was developed on the YOLOV3 platform to detect key features of the breast that are commonly used in plastic surgery for symmetry evaluation. The program was trained with 200 frontal photographs of patients who underwent breast surgery and was tested on 47 frontal images of patients who underwent breast reconstruction after breast cancer surgery. RESULTS: The program was able to detect key features in 97.74% of cases (boundaries of the breast in 94 of 94 cases, the nipple-areola complex in 94 of 94 cases, and the suprasternal notch in 41 of 47 cases). Mean time of detection was 0.52 seconds. CONCLUSIONS: The ad hoc neural network was successful in localizing key breast features, with a total detection rate of 97.74%. Neural networks and machine learning have the potential to improve the evaluation of breast symmetry in plastic surgery by automated and quick detection of features used by surgeons in practice. More studies and development are needed to further knowledge in this area.

A Novel Artificial Intelligence Model for Symmetry Evaluation in Breast Cancer Patients.

INTRODUCTION: Artificial intelligence (AI) is a milestone for human technology. In medicine, AI is set to play an important role as we progress into a new era. In plastic surgery, AI can participate in breast symmetry assessment, which until now has been mainly subjective, allowing for inconsistencies. This study aims to improve this evaluation process by integrating a novel trained neural network with the breast symmetry calculator, BAS-Calc. MATERIALS AND METHODS: We combined the BAS-Calc tool with a custom-made neural network trained to automatically detect key features of the breast. This integrated system was tested on 81 images of patients who had undergone breast reconstruction post-breast cancer treatment. Its performance was evaluated against two human observers using statistical analysis. RESULTS: Our model successfully detected 399/405 (98.51%) of landmarks. Spearman and Pearson correlation indicated a strong positive relationship while Cohen's kappa demonstrated moderate to strong agreement between human observers and AI model. Notably, the average calculation time for the AI was 0.92 seconds, 16 times faster than the 14.09 seconds for humans. CONCLUSIONS: Our AI model successfully calculated breast symmetry from images of patients who had undergone reconstructive oncological breast surgery, demonstrating high correlation with human assessments and a markedly reduced processing time. As AI continues to evolve, it is poised to become a pivotal tool in Medicine. Therefore, it is crucial for medical professionals to proactively engage in implementing AI technologies safely and effectively. Further studies are required to broaden our understanding and maximize the potential benefits in this area. Takeaway bullet points Artificial intelligence (AI) is an upcoming force to be reckoned with. AI should find its way into practical applications in plastic surgery. AI can be applied to improve patient care and evaluate aesthetic results. In this work, we present a novel AI model that automatically evaluates breast symmetry. LEVEL OF EVIDENCE IV: This journal requires that authors assign a level of evidence to each article. For a full description of these Evidence-Based Medicine ratings, please refer to the Table of Contents or the online Instructions to Authors www.springer.com/00266 .

An EEG database for the cognitive assessment of motor imagery during walking with a lower-limb exoskeleton.

One important point in the development of a brain-machine Interface (BMI) commanding an exoskeleton is the assessment of the cognitive engagement of the subject during the motor imagery tasks conducted. However, there are not many databases that provide electroencephalography (EEG) data during the use of a lower-limb exoskeleton. The current paper presents a database designed with an experimental protocol aiming to assess not only motor imagery during the control of the device, but also the attention to gait on flat and inclined surfaces. The research was conducted as an EUROBENCH subproject in the facilities sited in Hospital Los Madroños, Brunete (Madrid). The data validation reaches accuracies over 70% in the assessment of motor imagery and attention to gait, which marks the present database as a valuable resource for researches interested on developing and testing new EEG-based BMIs.

Nanoscale alterations in GABAB receptors and GIRK channel organization on the hippocampus of APP/PS1 mice.

Alzheimer's disease (AD) is characterized by a reorganization of brain activity determining network hyperexcitability and loss of synaptic plasticity. Precisely, a dysfunction in metabotropic GABAB receptor signalling through G protein-gated inwardly rectifying K+ (GIRK or Kir3) channels on the hippocampus has been postulated. Thus, we determined the impact of amyloid-ß (Aß) pathology in GIRK channel density, subcellular distribution, and its association with GABAB receptors in hippocampal CA1 pyramidal neurons from the APP/PS1 mouse model using quantitative SDS-digested freeze-fracture replica labelling (SDS-FRL) and proximity ligation in situ assay (P-LISA). In wild type mice, single SDS-FRL detection revealed a similar dendritic gradient for GIRK1 and GIRK2 in CA1 pyramidal cells, with higher densities in spines, and GIRK3 showed a lower and uniform distribution. Double SDS-FRL showed a co-clustering of GIRK2 and GIRK1 in post- and presynaptic compartments, but not for GIRK2 and GIRK3. Likewise, double GABAB1 and GIRK2 SDS-FRL detection displayed a high degree of co-clustering in nanodomains (40-50 nm) mostly in spines and axon terminals. In APP/PS1 mice, the density of GIRK2 and GIRK1, but not for GIRK3, was significantly reduced along the neuronal surface of CA1 pyramidal cells and in axon terminals contacting them. Importantly, GABAB1 and GIRK2 co-clustering was not present in APP/PS1 mice. Similarly, P-LISA experiments revealed a significant reduction in GABAB1 and GIRK2 interaction on the hippocampus of this animal model. Overall, our results provide compelling evidence showing a significant reduction on the cell surface density of pre- and postsynaptic GIRK1 and GIRK2, but not GIRK3, and a decline in GABAB receptors and GIRK2 channels co-clustering in hippocampal pyramidal neurons from APP/PS1 mice, thus suggesting that a disruption in the GABAB receptor-GIRK channel membrane assembly causes dysregulation in the GABAB signalling via GIRK channels in this AD animal model.

Age-Dependent Shift of AMPA Receptors From Synapses to Intracellular Compartments in Alzheimer's Disease: Immunocytochemical Analysis of the CA1 Hippocampal Region in APP/PS1 Transgenic Mouse Model.

Synapse loss occurs early in Alzheimer's disease (AD) patients and animal models. Alterations at synaptic level are a major morphological correlate of the memory deficits and related symptoms of AD. Given the predominant roles of synaptic AMPA receptors (AMPARs) in excitatory synaptic transmission in the brain, changes in their dynamic regulation are also implicated in the pathophysiology of AD. Here, we used immunolocalization techniques to analyze the expression and subcellular distribution of AMPARs in the hippocampal region of APP/PS1 mouse model of AD. Immunoblots and histoblots revealed that the total amount of AMPARs and their regional expression pattern in the hippocampus was similar in APP/PS1 mice and in age-matched wild type mice. At the ultrastructural level, two synapse populations were examined using SDS-digested freeze-fracture replica labeling in the stratum radiatum in mice: (i) on spines of CA1 pyramidal cells; and (ii) on randomly found dendritic shafts of CA1 interneurons. While 1- and 6-months-old APP/PS1 mice exhibited no change, we observed a significant reduction at 12 months in AMPAR density at synapses in both pyramidal cells and interneurons, compared to wild-type. This reduction of AMPARs in dendritic spines was accompanied by a significant increase in AMPAR subunit proteins identified in intracellular compartments. Our data demonstrate an age-dependent reduction of synaptic AMPARs in APP/PS1 mice, which may contribute to impaired learning and memory at later stages of AD.

Density of GABAB Receptors Is Reduced in Granule Cells of the Hippocampus in a Mouse Model of Alzheimer's Disease.

Metabotropic γ-aminobutyric acid (GABAB) receptors contribute to the control of network activity and information processing in hippocampal circuits by regulating neuronal excitability and synaptic transmission. The dysfunction in the dentate gyrus (DG) has been implicated in Alzheimer´s disease (AD). Given the involvement of GABAB receptors in AD, to determine their subcellular localisation and possible alteration in granule cells of the DG in a mouse model of AD at 12 months of age, we used high-resolution immunoelectron microscopic analysis. Immunohistochemistry at the light microscopic level showed that the regional and cellular expression pattern of GABAB1 was similar in an AD model mouse expressing mutated human amyloid precursor protein and presenilin1 (APP/PS1) and in age-matched wild type mice. High-resolution immunoelectron microscopy revealed a distance-dependent gradient of immunolabelling for GABAB receptors, increasing from proximal to distal dendrites in both wild type and APP/PS1 mice. However, the overall density of GABAB receptors at the neuronal surface of these postsynaptic compartments of granule cells was significantly reduced in APP/PS1 mice. Parallel to this reduction in surface receptors, we found a significant increase in GABAB1 at cytoplasmic sites. GABAB receptors were also detected at presynaptic sites in the molecular layer of the DG. We also found a decrease in plasma membrane GABAB receptors in axon terminals contacting dendritic spines of granule cells, which was more pronounced in the outer than in the inner molecular layer. Altogether, our data showing post- and presynaptic reduction in surface GABAB receptors in the DG suggest the alteration of the GABAB-mediated modulation of excitability and synaptic transmission in granule cells, which may contribute to the cognitive dysfunctions in the APP/PS1 model of AD.

Reduction in the neuronal surface of post and presynaptic GABAB receptors in the hippocampus in a mouse model of Alzheimer's disease.

The hippocampus plays key roles in learning and memory and is a main target of Alzheimer's disease (AD), which causes progressive memory impairments. Despite numerous investigations about the processes required for the normal hippocampal functions, the neurotransmitter receptors involved in the synaptic deficits by which AD disables the hippocampus are not yet characterized. By combining histoblots, western blots, immunohistochemistry and high-resolution immunoelectron microscopic methods for GABAB receptors, this study provides a quantitative description of the expression and the subcellular localization of GABAB1 in the hippocampus in a mouse model of AD at 1, 6 and 12 months of age. Western blots and histoblots showed that the total amount of protein and the laminar expression pattern of GABAB1 were similar in APP/PS1 mice and in age-matched wild-type mice. In contrast, immunoelectron microscopic techniques showed that the subcellular localization of GABAB1 subunit did not change significantly in APP/PS1 mice at 1 month of age, was significantly reduced in the stratum lacunosum-moleculare of CA1 pyramidal cells at 6 months of age and significantly reduced at the membrane surface of CA1 pyramidal cells at 12 months of age. This reduction of plasma membrane GABAB1 was paralleled by a significant increase of the subunit at the intracellular sites. We further observed a decrease of membrane-targeted GABAB receptors in axon terminals contacting CA1 pyramidal cells. Our data demonstrate compartment- and age-dependent reduction of plasma membrane-targeted GABAB receptors in the CA1 region of the hippocampus, suggesting that this decrease might be enough to alter the GABAB -mediated synaptic transmission taking place in AD.

SK2 Channels Associate With mGlu1α Receptors and CaV2.1 Channels in Purkinje Cells.

The small-conductance, Ca2+-activated K+ (SK) channel subtype SK2 regulates the spike rate and firing frequency, as well as Ca2+ transients in Purkinje cells (PCs). To understand the molecular basis by which SK2 channels mediate these functions, we analyzed the exact location and densities of SK2 channels along the neuronal surface of the mouse cerebellar PCs using SDS-digested freeze-fracture replica labeling (SDS-FRL) of high sensitivity combined with quantitative analyses. Immunogold particles for SK2 were observed on post- and pre-synaptic compartments showing both scattered and clustered distribution patterns. We found an axo-somato-dendritic gradient of the SK2 particle density increasing 12-fold from soma to dendritic spines. Using two different immunogold approaches, we also found that SK2 immunoparticles were frequently adjacent to, but never overlap with, the postsynaptic density of excitatory synapses in PC spines. Co-immunoprecipitation analysis demonstrated that SK2 channels form macromolecular complexes with two types of proteins that mobilize Ca2+: CaV2.1 channels and mGlu1α receptors in the cerebellum. Freeze-fracture replica double-labeling showed significant co-clustering of particles for SK2 with those for CaV2.1 channels and mGlu1α receptors. SK2 channels were also detected at presynaptic sites, mostly at the presynaptic active zone (AZ), where they are close to CaV2.1 channels, though they are not significantly co-clustered. These data demonstrate that SK2 channels located in different neuronal compartments can associate with distinct proteins mobilizing Ca2+, and suggest that the ultrastructural association of SK2 with CaV2.1 and mGlu1α provides the mechanism that ensures voltage (excitability) regulation by distinct intracellular Ca2+ transients in PCs.

Differential association of GABAB receptors with their effector ion channels in Purkinje cells.

Metabotropic GABAB receptors mediate slow inhibitory effects presynaptically and postsynaptically through the modulation of different effector signalling pathways. Here, we analysed the distribution of GABAB receptors using highly sensitive SDS-digested freeze-fracture replica labelling in mouse cerebellar Purkinje cells. Immunoreactivity for GABAB1 was observed on presynaptic and, more abundantly, on postsynaptic compartments, showing both scattered and clustered distribution patterns. Quantitative analysis of immunoparticles revealed a somato-dendritic gradient, with the density of immunoparticles increasing 26-fold from somata to dendritic spines. To understand the spatial relationship of GABAB receptors with two key effector ion channels, the G protein-gated inwardly rectifying K+ (GIRK/Kir3) channel and the voltage-dependent Ca2+ channel, biochemical and immunohistochemical approaches were performed. Co-immunoprecipitation analysis demonstrated that GABAB receptors co-assembled with GIRK and CaV2.1 channels in the cerebellum. Using double-labelling immunoelectron microscopic techniques, co-clustering between GABAB1 and GIRK2 was detected in dendritic spines, whereas they were mainly segregated in the dendritic shafts. In contrast, co-clustering of GABAB1 and CaV2.1 was detected in dendritic shafts but not spines. Presynaptically, although no significant co-clustering of GABAB1 and GIRK2 or CaV2.1 channels was detected, inter-cluster distance for GABAB1 and GIRK2 was significantly smaller in the active zone than in the dendritic shafts, and that for GABAB1 and CaV2.1 was significantly smaller in the active zone than in the dendritic shafts and spines. Thus, GABAB receptors are associated with GIRK and CaV2.1 channels in different subcellular compartments. These data provide a better framework for understanding the different roles played by GABAB receptors and their effector ion channels in the cerebellar network.