Welcome to My Research World

About Me

I am a dedicated Ph.D. student specializing in the intersection of machine learning, deep learning, and healthcare. My passion lies in exploring how these advanced technologies can revolutionize healthcare systems, diagnostics, and patient care.

My Research Interests

• Machine Learning in Healthcare: Leveraging machine learning algorithms to improve diagnostic accuracy and treatment effectiveness.
• Deep Learning Applications: Exploring the capabilities of deep neural networks in processing complex medical data.
• Health Analytics: Utilizing data analytics to uncover insights in healthcare data and improve patient outcomes.

Resume

• View Resume -Download my resume

• View CV -Download my CV

Publications

• Recognition of atypical behavior in autism diagnosis from video using pose estimation over time: A study on implementing a computer vision-based approach for automatic Autism Spectrum Disorder prediction, utilizing video data of children to distinguish between typical and atypical behaviors through pose estimation, motion pattern analysis, and binary classification, achieving a 72.4% accuracy. Link to paper
• Additional value of augmenting current subscales in braden scale with advanced machine learning technique for pressure injury risk assessment: S Incidence of PI Link to paper
• An efficient 3d synthetic model generation pipeline for human pose data augmentation: Insights into this particular publication. Link to paper
• Detection of glycemic excursions using morphological and time-domain ECG features: Insights into this particular publication. Link to paper
• Masters’ Thesis : Data Efficient Analysis of Human Behavior by Spatio-Temporal Pose Generation and Inference: Insights into this particular publication. Link to paper
• More publications and Patents can be found here

Projects

• SNOOZE: An intelligent personal organizer that assists you with ddaily emails - summaries, sending emails, scheduling meetings ,and generating to-do lists. (Link to Github)
• Echoes of Choices: Imagine a story navigation game where a user plays as a character in a story which is derived from either a real life story or movie. Using the structure of decisions made by the user, behavior patterns can be studied and compared to baseline decisions. Uses LLM , user - system prompting, Psychology and intelligent visaulisation. (Link to Github)
• Hypoglycemia Classification from ECG: Hypoglycemia (Low blood glucose) is a symptom of Type 2 Diabetes and detection is done using 1 lead ECG which is captured using Zephyr Bioharrness.Link to Github
• 3D Pose Estimation: Using point clouds captured using IR camera like Intell real sense, the algorithm takes into account the point clouds and then uses a novel 3D registration technique to generate a 3D pose that could be used in modeling software like UNITY. Link to paper

Contact

Feel free to reach out to me for collaborations or discussions related to my research fields.

• Email: kathan@usa.com
• LinkedIn: Kathan Vyas
• Twitter: @kathanvyas

---

Thank you for visiting my page. I am always eager to engage in conversations about technology and its impact on healthcare.