Clinical, Health, Medical Informatics
Machine Learning for Applied Clinical Informatics
Application of natural language processing, signal processing, image processing, computer vision, artificial intelligence and machine learning techniques to clinical, heath, and medical informatics problems. AI for NICU/PICU/ICU/CICU data, Critical Care, Emergency Medicine, Psychiatry, Sepsis, Bronchopulmonary Dysplasia. AI/ML for pharmacokinetic/pharmacodynamic (PK/PD) data. Healthcare NLP for EHR/EMR/ePHI/PHR/PRO Data - Unstructured text data analysis with ML/DL - Clinical data - MLDevOps + FHIR. AI for Healthcare and Clinical Implementation. AI for NICU/PICU/ICU/CICU. mHealth, IoToys in pediatrics.
Project pages:
AI/NLP/Imaging for NICU: NLP-PICC, PICCLineNet, BPD, Sepsis, NPD
EHR : RareDiseases, ASD
Others: CPIID, mHealth
Selected publications:
M. M. S. Missen, A. Javed, H. Asmat, M. Nosheen, M. Coustaty, N. Salamat, V. B. S. Prasath. Systematic review and usability evaluation of writing mobile apps for children. New Review of Hypermedia and Multimedia, 25(3), 137-160, December 2019. Special issue on Advances in Multimedia and Educational Technology. doi:10.1080/13614568.2019.1677787
M. Shah, D. Shu, V. B. S. Prasath, Y. Ni, A. Schapiro, K. Dufendach. Machine learning for detection of correct peripherally inserted central catheter tip position from radiology reports in infants. Applied Clinical Informatics, 12(04), 856-863, August 2021. doi:10.1055/s-0041-1735178 [NLP-PICC]
M. Shah, D. Jain, V. B. S. Prasath, K. Dufendach. Artificial intelligence in bronchopulmonary dysplasia - Current research and unexplored frontiers. Pediatric Research, 93(2), 287-290, January 2023. doi:10.1038/s41390-022-02387-z
Deoghare et al., PICCLineNet: Deep learning for detecting peripherally inserted central catheter (PICC) lines and tips from radiological images in infants. 2024. Preliminary version at arXiv:24xx.abcde. [PICCLineNet]
Prasath et al., Challenges of heterogenous data standardization and organization - An informatics perspective. 2024. In preparation.
AI Text Analytics
Natural language processing (NLP) is becoming commonplace and in the last few years large language models are increasingly used in processing various types of data (EHR, radiology reports,...). Hence personal genomes will be increasingly utilized for precision medicine. It is therefore very important to develop new approaches using latest data science tools for solving problems in bioinformatics. Prasath Lab is interested in leveraging AI/ML/DL to solve challenges in large-scale text processing in the clinical informatics domain. Given the expertise and experience with the multidisciplinary projects and the proven track-record in bringing quantitative approaches from mathematics, computer science, and statistics we are well-poised to be a connector among different domains. We are also part of multiple collaborative efforts and our research interests span the full spectrum of informatics:
Large Language Models (LLMs)
GPT and variants
xAI applications in text data
Generative AI for NLP
Interpretable AI Tools
Multimodal Fusion (text, imaging, genomics,...)
Foundation Models
Selected publications:
S. Rawat, V. Vijayarajan, V. B. S. Prasath. Modified K-means with cluster assignment - Application to COVID-19 data. arXiv, February 2024. doi:10.48550/arXiv.2402.03380
M. Imam, S. Hamzah, S. Aggarwal, S. Dev, V. B. S. Prasath. On utilizing transformer-based models for preventing harmful response in LLMs. Submitted, 2024.
V. B. S. Prasath. Leveraging large language models (LLMs) for conversational nutrition guidance – Oppurtunities and challenges. In preparation, 2024.