Suggestions
Visar Berisha
Professor and Scientist-Entrepreneur
Visar Berisha is an accomplished academic and researcher currently serving as the Associate Dean of Research Commercialization at the Fulton Schools of Engineering at Arizona State University (ASU). He has been a faculty member at ASU since 2013, holding joint appointments in the College of Health Solutions and the School of Electrical, Computer and Energy Engineering. His academic background includes a Ph.D. in Electrical Engineering from ASU, which he completed in 2007.134
Research and Academic Contributions
Berisha's research primarily focuses on machine learning, statistical signal processing, and clinical speech analytics. He aims to develop innovative tools to enhance the understanding of speech perception, particularly in the context of neurological conditions. His work has led to the development of data-driven models that mimic human cognition, with applications in healthcare and technology.135
He has co-founded Aural Analytics, a company that specializes in speech analytics, and has held various research positions, including at the MIT Lincoln Laboratory and Raytheon. His contributions to the field have been recognized through numerous awards, including the 2021 College of Health Solutions Research Award and the 2017 Scrip Award for Best Technological Development for Clinical Trials.145
Publications and Impact
Berisha has an extensive publication record, with numerous articles in peer-reviewed journals focusing on topics such as speech analysis and the application of machine learning in healthcare. His research has been influential in advancing the understanding of speech disorders and the development of technologies for their assessment and treatment.256
Professional Engagement
In addition to his research, Berisha is actively involved in various academic committees and initiatives at ASU, contributing to faculty recruitment and interdisciplinary collaborations. He is recognized for his efforts in mentoring faculty interested in entrepreneurship and for organizing workshops related to speech analytics.1345
Overall, Visar Berisha is a prominent figure in the intersection of engineering and health sciences, dedicated to leveraging technology to improve human health outcomes.
Highlights
Happy to share our latest paper in Nature Portfolio npj Digital Medicine: Responsible development of clinical speech AI: Bridging the gap between clinical research and technology
Julie Liss and I have been collaborating for over 10 years on clinical speech analytics. Last year, we combined our insights into a perspectives paper that integrates the lessons we’ve learned during this time. We argue that the co-design of new clinical speech AI pipelines, involving collaboration between the AI, speech science, and clinical research communities, will result in models that are easier to train (due to lower sample complexity), are more likely to generalize, and are inherently interpretable. In the paper, we also provide guidelines on how to achieve this collaborative approach.
The paper presents: • A new framework for organizing health conditions by their impact on speech. This framework facilitates a shared understanding of how clinical conditions affect speech, allowing model developers to focus their efforts on feasible problems. • A discussion of relevant contexts of use for clinical speech AI that goes beyond cross-sectional classification on high-dimensional features. We urge model developers to consider the clinical use case prior to data collection or model development. • An approach to the safe development and validation of models. This provides a rough roadmap for developing and validating new models aligned with the context of use. • A discussion of ethical, privacy, and security considerations. We emphasize the importance of developing AI models that prioritize patient confidentiality and consent, ensuring that these technologies are implemented responsibly and transparently.
https://lnkd.in/gwASMcZy
#ClinicalSpeechAI #HealthcareInnovation #AI #DigitalHealth #VoiceBiomarkers
Register for our tutorial at INTERSPEECH 2024: Responsible Development of Clinical Speech AI
Now that INTERSPEECH 2024 decisions are out and you are registering for the conference, please consider registering for our full day tutorial. The topics to be covered include:
- designing speech elicitation tasks for different clinical conditions,
- speech data collection and hardware verification,
- developing and validating speech representations for clinical measures,
- developing reliable clinical prediction models, and
- ethical considerations.
These topics are going to be covered by experts in this field: Visar Berisha Julie Liss Si-ioi Ng Ingo Siegert Nina R. B. Nicholas Cummins
More information on the tutorial in the link below
https://lnkd.in/gYRnK6bA
