Researchers at the University at Buffalo have developed SYNAPSE (Synchronized Neural and Auditory Pupillometry for Sensory Evaluation), a multimodal assessment platform that objectively evaluates hearing-related disorders by combining synchronized EEG brain-activity recordings with pupil-response monitoring. The system uses around-the-ear electrodes, infrared eye tracking, and standardized auditory tests to measure responses linked to tinnitus, hyperacusis, misophonia, and related sound-tolerance disorders. By integrating physiological biomarkers rather than relying solely on patient self-reports, SYNAPSE may improve diagnostic accuracy and treatment monitoring. The patent-pending technology has been implemented in human participant studies, reached TRL 6 development, and is available for licensing or collaboration. SYNAPSE could expand access to more objective and reliable assessment of auditory disorders, particularly for individuals who have difficulty describing symptoms. Improved diagnosis and treatment monitoring may lead to better clinical care, enhanced patient outcomes, and new insights into the biological mechanisms underlying sound-sensitivity conditions.

This technology introduces a novel artificial intelligence model that improves the detection of previously unseen electromagnetic waveforms by combining time-domain and frequency-domain signal representations. Using a cosine similarity loss function, the system enhances extraction of class-specific features, enabling more accurate identification of unknown signals than conventional approaches. Testing demonstrated approximately a 10% improvement in detection accuracy compared to models lacking this capability. Potential applications include electronic warfare, spectrum management, intelligence and surveillance operations, radio astronomy, and communication security. By providing more reliable waveform classification and interference detection, the technology supports improved situational awareness and spectrum utilization. The patent-pending technology is at TRL 3 and available for licensing.

Candidemia is a major cause of hospital-acquired bloodstream infections, with high mortality and substantial healthcare costs. However, delayed diagnosis, limited susceptibility testing, antifungal resistance, and the complexity of treatment continue to challenge effective patient management. This review examines recent advances in candidemia epidemiology, diagnosis, and therapy. It also discusses emerging approaches, including artificial intelligence-assisted diagnosis, vaccine-based prevention, and synergistic therapies, that could reshape candidemia management and improve outcomes.

The Research Highlighter-MatchMaker Project, developed at the University at Albany, uses artificial intelligence and natural language processing to help researchers identify funding opportunities, find collaborators, and connect with research communities. By integrating publication data from Scopus and funding information from SPIN, the platform delivers personalized recommendations through searchable tools and automated email alerts. Users can search by researcher, topic, or proposal text, enabling more accurate matching than traditional keyword-based systems. The technology has the potential to accelerate interdisciplinary collaboration, improve grant competitiveness, and expand access to research opportunities, particularly for early-career researchers. The patent-pending technology is at TRL 3 and available for licensing.

This technology improves neural network security by introducing controlled weight perturbations during training, creating models that are more resistant to unauthorized access, snooping, cryptanalysis, and other attacks. Unlike traditional approaches that rely on external security measures, the method embeds protection directly into the training process, securing the network from the first layer onward and extending safeguards throughout the model. The approach is particularly well suited for mobile, edge, and IoT devices with limited computing resources, while also strengthening machine learning models delivered through APIs. By enhancing security without significant performance penalties, the technology supports safer deployment of AI systems in connected environments. The patent-pending technology is at TRL 3 and available for licensing.

Researchers at the University at Buffalo have developed AudioSight, a smartphone-based screening platform that objectively assesses hearing-related disorders using pupil responses to sound. The system combines a low-cost 3D-printed smartphone attachment, near-infrared illumination, and artificial intelligence to analyze audio-evoked pupillary responses. Advanced machine learning models process features such as response latency, dilation speed, and recovery time to generate risk scores for conditions including hidden hearing loss, tinnitus, and hyperacusis. Designed for self-administration and remote use, the technology could expand access to hearing screening for children, older adults, cognitively impaired individuals, and underserved populations. The TRL 6 technology is available for licensing or collaboration. AudioSight could make hearing-disorder screening more accessible, affordable, and objective by enabling at-home assessment and teleaudiology services. Earlier identification of hearing disorders may improve treatment outcomes, reduce barriers to care, and support populations that have difficulty completing conventional hearing tests.

Using artificial intelligence, climate models, and weather observations, researchers find strong evidence that greenhouse gases emitted since the 2015 Paris Climate Agreement have intensified Europe’s recent heat waves.