A5027325953- Visual perception and processing mechanisms
- Color Science and Applications
- Color perception and design
- Advanced X-ray and CT Imaging
- Innovations in Medical Education
- Acute Ischemic Stroke Management
- Ophthalmology and Visual Impairment Studies
- Artificial Intelligence in Healthcare and Education
- Gaze Tracking and Assistive Technology
- Radiation Dose and Imaging
- Radiomics and Machine Learning in Medical Imaging
- Cerebrovascular and Carotid Artery Diseases
- Tactile and Sensory Interactions
- Radiology practices and education
- Galectins and Cancer Biology
- Traumatic Brain Injury and Neurovascular Disturbances
- Visual Attention and Saliency Detection
- Cardiac Imaging and Diagnostics
- Intracerebral and Subarachnoid Hemorrhage Research
- Surface Roughness and Optical Measurements
- EEG and Brain-Computer Interfaces
- Advanced Optical Imaging Technologies
- Music Technology and Sound Studies
- Hearing Loss and Rehabilitation
- Multisensory perception and integration
Harvard University
2018-2025
Massachusetts General Hospital
2019-2025
Mass General Brigham
2021-2024
Office of Science
2022-2024
META Health
2022
Meta (United States)
2017-2021
Center for Clinical Research (United States)
2021
DASA (Brazil)
2021
Mount Auburn Hospital
2021
Capital Medical University
2021
Humans use multiple sources of sensory information to estimate environmental properties. For example, the eyes and hands both provide relevant about an object's shape. The shape using binocular disparity, perspective projection, etc. supply haptic by means tactile proprioceptive cues. Combining across cues can improve estimation object properties but may come at a cost: loss single-cue information. We report that is indeed lost when from within same modality (disparity texture gradients in...
How does the visual system combine information from different depth cues to estimate three-dimensional scene parameters? We tested a maximum-likelihood estimation (MLE) model of cue combination for perspective (texture) and binocular disparity surface slant. By factoring reliability each into process, MLE provides more reliable estimates slant than would be available either alone. measured in isolation across range slants distances using slant-discrimination task. The texture increases as...
Expanding the US Food and Drug Administration-approved indications for immune checkpoint inhibitors in patients with cancer has resulted therapeutic success immune-related adverse events (irAEs). Neurologic irAEs (irAE-Ns) have an incidence of 1%-12% a high fatality rate relative to other irAEs. Lack standardized disease definitions accurate phenotyping leads syndrome misclassification impedes development evidence-based treatments translational research. The objective this study was develop...
Importance Early detection of pneumothorax, most often via chest radiography, can help determine need for emergent clinical intervention. The ability to accurately detect and rapidly triage pneumothorax with an artificial intelligence (AI) model could assist earlier identification improve care. Objective To compare the accuracy AI vs consensus thoracic radiologist interpretations in detecting any (incorporating both nontension tension pneumothorax) pneumothorax. Design, Setting, Participants...
Non-contrast head CT (NCCT) is extremely insensitive for early (< 3-6 h) acute infarct identification. We developed a deep learning model that detects and delineates suspected infarcts on NCCT, using diffusion MRI as ground truth (3566 NCCT/MRI training patient pairs). The substantially outperformed 3 expert neuroradiologists test set of 150 scans patients who were potential candidates thrombectomy (60 stroke-negative, 90 stroke-positive middle cerebral artery territory only infarcts), with...
Hand tracking and haptics are gaining more importance as key technologies of virtual reality (VR) systems. For designing such systems, it is fundamental to understand how the appearance hands influences user experience human brain integrates vision haptics. However, currently unknown whether multi-sensory integration visual haptic feedback can be influenced by in VR. We performed a study VR gain insight into effect hand on combines signals using cue-conflict paradigm. In this paper, we show...
As head-mounted displays (HMDs) commonly present a single, fixed-focus display plane, conflict can be created between the vergence and accommodation responses of viewer. Multifocal HMDs have long been investigated as potential solution in which multiple image planes span viewer's range. Such require scene decomposition algorithm to distribute depiction objects across planes, previous work has shown that simple decompositions achieved real-time. However, recent optimal further improve...
Cuprizone leads to demyelination of the corpus callosum (CC) and activates progenitor cells in adjacent subventricular zone (SVZ), a stem cell niche which contributes remyelination. The healthy SVZ contains semi-activated microglia constitutively expresses pro-inflammatory molecule galectin-3 (Gal-3) suggesting uniquely regulates inflammation. We studied inflammatory response cuprizone CC Gal-3 knockout mice using immunohistochemistry with vitro neurosphere assay. caused loss myelin basic...
An ideal Mixed Reality (MR) system would only present virtual information (e.g., a label) when it is useful to the person. However, deciding label challenging: depends on variety of factors, including current task, previous knowledge, context, etc. In this paper, we propose Reinforcement Learning (RL) method learn show or hide an object's given eye movement data. We demonstrate capabilities approach by showing that intelligent agent can cooperative policies better support users in visual...
Background Intracranial hemorrhage is a critical finding on computed tomography (CT) of the head. This study compared accuracy an artificial intelligence (AI) model (Annalise Enterprise CTB Triage Trauma) to consensus neuroradiologist interpretations in detecting 4 subtypes: acute subdural/epidural hematoma, subarachnoid hemorrhage, intra‐axial and intraventricular hemorrhage. Methods A retrospective stand‐alone performance assessment was conducted data sets cases noncontrast CT head...
The distribution of empirical corresponding points in the two retinas has been well studied along horizontal and vertical meridians, but not other parts visual field. Using an apparent-motion paradigm, we measured positions those across central portion We found that Hering-Hillebrand deviation (a from Vieth-Müller circle) Helmholtz shear disparity (backward slant horopter) exist throughout also no evidence for non-zero disparities points. used data to find combination space binocular eye...
Multiple sclerosis (MS) frequently starts near the lateral ventricles, which are lined by subventricular zone (SVZ) progenitor cells that can migrate to lesions and contribute repair. Because MS‐induced inflammation may decrease SVZ proliferation thus limit repair, we studied role of galectin‐3 (Gal‐3), a proinflammatory protein. Gal‐3 expression was increased in periventricular regions human MS post‐mortem brain samples also upregulated murine model, Theiler's encephalomyelitis virus (TMEV)...
BACKGROUND: Hypertrophic cardiomyopathy (HCM) is associated with significant morbidity and mortality, including sudden cardiac death in the young. Its prevalence estimated to be 1 500, although many people are undiagnosed. The ability screen electrocardiograms for its presence could improve detection enable earlier diagnosis. This study evaluated accuracy of an artificial intelligence device (viz HCM) detecting HCM based on a 12-lead electrocardiogram. METHODS: was previously trained using...
Color vision supports two distinct visual functions: discrimination and constancy. Discrimination requires that the response to objects within a scene be different. Constancy any object same across scenes. Across changes in scene, adaptation can improve by optimizing use of available range. Similarly, constancy stabilizing fixed illumination. Can common mechanisms achieve these goals simultaneously? We develop theoretical framework for answering this question present several example...
Objective: To explore medical students’ views on support services, stigma, and teaching of wellbeing in light their experiences stress distress. Design, participants setting: Quantitative survey students at five universities Australia New Zealand November 2007. Main outcome measures: Medical stigma attached to undergoing distress, wellbeing. Results: 1328 completed the (26% response rate). Seventy-one per cent were aware services university. Of these, 46% believed adequately promoted, 49%...
Abstract Stroke is a leading cause of death and disability. The ability to quickly identify the presence acute infarct quantify volume on magnetic resonance imaging (MRI) has important treatment implications. We developed machine learning model that used apparent diffusion coefficient weighted series. It was trained 6,657 MRI studies from Massachusetts General Hospital (MGH; Boston, USA). All were labelled positive or negative for (classification annotation) with 377 having region interest...
Color vision is useful for detecting surface boundaries and identifying objects. Are the signals used to perform these two functions processed by common mechanisms, or has visual system optimized its processing separately each task? We measured effect of mean chromaticity luminance on color discriminability appearance under well-matched stimulus conditions. In discrimination experiments, a pedestal spot was presented in one interval + test second. Observers indicated which contained test....