Machine Learning with Brain Data

Опубликована: Янв. 1, 2025

Язык: Английский

---

Tetherless miniaturized point detector device for monitoring cortical surface hemodynamics in mice

Journal of Biomedical Optics, Год журнала: 2025, Номер 30(S2)

Опубликована: Март 19, 2025

SignificanceSeveral miniaturized optical neuroimaging devices for preclinical studies mimicking benchtop instrumentation have been proposed in the past. However, they are generally relatively large, complex, and power-hungry, limiting their usability long-term measurements freely moving animals. Further, there is limited research development of algorithms to analyze signals.AimWe aim develop a cost-effective, easy-to-use intrinsic monitoring system (TinyIOMS) that can be reliably used record spontaneous stimulus-evoked hemodynamic changes further cluster brain states based on features.ApproachWe present design fabrication TinyIOMS (8 mm×13 mm×9 mm3, 1.2 g with battery). A standard camera-based widefield (WFIOS) validate signals. Next, continuously activity 7 h chronically implanted mice. We show up 2 days intermittent recording from an animal. An unsupervised machine learning algorithm signals.ResultsWe observed data comparable WFIOS data. Stimulus-evoked recorded using was distinguishable stimulus magnitude. Using TinyIOMS, we successfully achieved continuous its home cage placed animal housing facility, i.e., outside controlled lab environment. (k-means clustering), grouping into two clusters representing asleep awake accuracy ∼91%. The same then applied 2-day-long dataset, where similar emerged.ConclusionsTinyIOMS applications Results indicate device suitable mice during behavioral synchronized video external stimuli.

Язык: Английский

---

Neuroimaging and artificial intelligence for assessment of chronic painful temporomandibular disorders—a comprehensive review

International Journal of Oral Science, Год журнала: 2023, Номер 15(1)

Опубликована: Дек. 28, 2023

Abstract Chronic Painful Temporomandibular Disorders (TMD) are challenging to diagnose and manage due their complexity lack of understanding brain mechanism. In the past few decades’ neural mechanisms pain regulation perception have been clarified by neuroimaging research. Advances in bridged gap between activity subjective experience pain. Neuroimaging has also made strides toward separating underlying chronic painful TMD. Recently, Artificial Intelligence (AI) is transforming various sectors automating tasks that previously required humans’ intelligence complete. AI started contribute recognition, assessment, The application pathophysiology diagnosis TMD still its early stages. objective present review identify contemporary approaches such as structural, functional, molecular techniques used investigate individuals. Furthermore, this guides practitioners on relevant aspects how methods can revolutionize our aid both management enhance patient outcomes.

Язык: Английский

---

Brainchop: Providing an Edge Ecosystem for Deployment of Neuroimaging Artificial Intelligence Models

Aperture Neuro, Год журнала: 2024, Номер 4

Опубликована: Сен. 5, 2024

Deep learning has proven highly effective in various medical imaging scenarios, yet the lack of an efficient distribution platform hinders developers from sharing models with end-users. Here, we describe brainchop, a fully functional web application that allows users to apply deep developed Python local neuroimaging data within their browser. While training artificial intelligence is computationally expensive, applying existing can be very fast; brainchop harnesses end user's graphics card such brain extraction, tissue segmentation, and regional parcellation require only seconds avoids privacy issues impact cloud-based solutions. The integrated visualization validate inferences, includes tools annotate edit resulting segmentations. Our pure JavaScript implementation optimized helper functions for conforming volumes filtering connected components minimal dependencies. Brainchop provides simple mechanism distributing additional image processing tasks, including registration identification abnormal tissue, tumors, lesions hyperintensities. We discuss considerations other AI model leverage this open-source resource.

Язык: Английский

---

DCA-Enhanced Alzheimer’s detection with shearlet and deep learning integration

Computers in Biology and Medicine, Год журнала: 2024, Номер 185, С. 109538 - 109538

Опубликована: Дек. 13, 2024

Язык: Английский

---

Assessment of Sports Concussion in Female Athletes: A Role for Neuroinformatics?

Neuroinformatics, Год журнала: 2024, Номер 22(4), С. 607 - 618

Опубликована: Июль 30, 2024

Abstract Over the past decade, intricacies of sports-related concussions among female athletes have become readily apparent. Traditional clinical methods for diagnosing suffer limitations when applied to athletes, often failing capture subtle changes in brain structure and function. Advanced neuroinformatics techniques machine learning models invaluable assets this endeavor. While these technologies been extensively employed understanding concussion male there remains a significant gap our comprehension their effectiveness athletes. With its remarkable data analysis capacity, offers promising avenue bridge deficit. By harnessing power learning, researchers can link observed phenotypic neuroimaging sex-specific biological mechanisms, unraveling mysteries Furthermore, embedding within enable examining architecture alterations beyond conventional anatomical reference frame. In turn, allows gain deeper insights into dynamics concussions, treatment responses, recovery processes. This paper endeavors address crucial issue sex differences multimodal experimental design approaches athlete populations, ultimately ensuring that they receive tailored care require facing challenges concussions. Through better integration, feature identification, knowledge representation, validation, etc., neuroinformaticists, are ideally suited bring clarity, context, explainabilty study head injuries males females, helping define recovery.

Язык: Английский

---

Machine learning for brain tumor classification: evaluating feature extraction and algorithm efficiency

Discover Artificial Intelligence, Год журнала: 2024, Номер 4(1)

Опубликована: Дек. 19, 2024

Язык: Английский

---

Seeing more than the tip of the iceberg: Approaches to subthreshold effects in functional magnetic resonance imaging of the brain

Опубликована: Май 7, 2023

Many functional magnetic resonance imaging (fMRI) studies rely on mass-univariate inference with subsequent multiple comparison correction. Statistical results are frequently visualized as thresholded statistical maps. This approach has inherent limitations including the risk of drawing overly-selective conclusions based only selective passing such thresholds. article gives an overview both established and newly emerging approaches to supplement conventional analyses by incorporating information about subthreshold effects aim improve interpretation findings or leverage a wider array information. Topics covered include neuroimaging data visualization, p-value histogram analysis related Higher Criticism for detecting rare weak well multivariate dedicated Bayesian approaches.

Язык: Английский

---

Diagnostic Machine Learning Applications on Clinical Populations using Functional Near Infrared Spectroscopy: A Review

medRxiv (Cold Spring Harbor Laboratory), Год журнала: 2023, Номер unknown

Опубликована: Фев. 7, 2023

Abstract Functional near-infrared spectroscopy (fNIRS) and its interaction with machine learning (ML) is a popular research topic for the diagnostic classification of clinical disorders due to lack robust objective biomarkers. This review provides an overview on psychiatric diseases by using fNIRS ML. Article search was carried out 45 studies were evaluated considering their sample sizes, used features, ML methodology, reported accuracy. To our best knowledge, this first that reports applications fNIRS. We found there has been increasing trend perform fNIRS-based biomarker since 2010. The most studied populations are schizophrenia (n=12), attention deficit hyperactivity disorder (n=7), autism spectrum (n=6) populations. There significant negative correlation between size (>20) accuracy values. Support vector (SVM) deep (DL) approaches classifier (SVM = 20) (DL 10). Eight these recruited number participants more than 100 classification. Change in oxy-hemoglobin (ΔHbO) based features change deoxy-hemoglobin-based ones ΔHbO-based mean ΔHbO (n=11) functional connections (n=11). Using data might be promising approach reveal specific biomarkers

Язык: Английский

---

Predicting antipsychotic responsiveness using a machine learning classifier trained on plasma levels of inflammatory markers in schizophrenia

Research Square (Research Square), Год журнала: 2024, Номер unknown

Опубликована: Авг. 10, 2024

Background In this research study, we apply machine learning techniques to navigate the multifaceted landscape of schizophrenia. Our method entails development predictive models, emphasizing peripheral inflammatory biomarkers, which are classified into treatment response subgroups: antipsychotic-responsive, clozapine-responsive, and clozapine-resistant. Methods The study comprises 146 schizophrenia patients (49 antipsychotics-responsive, 68 29 clozapine-resistant) 49 healthy controls. Protein levels immune biomarkers were quantified using Olink Target 96 Inflammation Panel. To predict labels, a support vector classifier is trained on data matrix evaluated via leave-one-out cross-validation. Associated protein identified recursive feature elimination. Findings We constructed three separate models for binary classification: one discern controls from individuals with (AUC = 0.74), another differentiate who responsive antipsychotics 0.88), third distinguish treatment-resistant 0.78). Employing techniques, features capable distinguishing between subgroups. Interpretation demonstrates power uncover subtle signals often overlooked by traditional statistics. Unlike t-tests, it handles multiple simultaneously, capturing complex relationships. Chosen simplicity, robustness, reliance strong sets, its integration artificial intelligence like SHapely Additive exPlanations enhances model interpretability, especially biomarker screening. This highlights potential integrating in clinical practice. Not only does deepen our understanding schizophrenia's heterogeneity, but also holds promise enhancing accuracy, thereby facilitating more targeted effective interventions mental health disorder.

Язык: Английский

---