Sishir Surjit

Global Challenge: Combining Artificial Intelligence with healthcare to create systems that help improve medical diagnostics

Abstract:

This paper investigates strategies to improve the performance of current medical diagnostic AI models, specifically focusing on the diagnosis of prostate cancer. Given the significant discrepancies in diagnostic outcomes due to radiologist bias, overwork, and a shortage of expertise, particularly in low-income countries. This study proposes a novel diagnostic Machine/Deep Learning model that leverages incremental learning methods to interpret and evaluate the multicenter radiomics data from various imaging modalities (MRI, CT, etc.). The incremental learning methods were leveraged due to their current ability and potential to adapt to newer incoming datasets while retaining the utility of prior datasets, allowing for the creation of more robust and accurate models. Using large datasets from the Prostate Imaging: Cancer AI (PICAI) competition, our research evaluates the model’s performance, adaptability, and potential to reduce diagnostic inconsistencies and bias. The proposed solution primarily aims to alleviate radiologists’ workload while providing more accurate and consistent diagnoses, thereby addressing both local and global challenges in cancer diagnostics.

This research possesses a strong connection with the author’s GLASS Global Challenge and research areas. The research correlates greatly with the Health and Data Science focus in the Tandon Area of Excellence, as it utilises large-scale radiomics data, and utilizes specific parameters and features to determine health information of patients. Additionally, the use of AI and machine learning in healthcare highlights the increasing importance of technology in addressing global health challenges. This research and its goals of creating a successful ML/Deep Learning algorithm aligns greatly with the UN sustainable goals of Good Health and Well-Being, as it aims to provide a low-cost diagnostic system that can be implemented in areas with minimal medical help. This work and research correlates greatly with the experiences the author has had within GLASS and the impact it has had. Due to his experience in NYU Abu Dhabi he was able to become a part of the LANTRN lab in NYU which solidified his interest in progressing medical devices and eventually medical systems to improve their accessibility. Furthermore, the experiences of the author within his enrollment in the data analytics stream of the Global accelerator program and his time as the Head of Purchasing in the General Engineering department introduced him to large datasets and how to utilize them to gain useful insights, which greatly assisted him in the pursuit of his global challenge.

Bio:

Sishir is an undergraduate senior at NYU Tandon pursuing a degree in Chemical and Biomolecular Engineering, with a focus on biomedical systems and devices. He is passionate about advancing healthcare through technology and aspires to build a career in the biomedical and bioinformatics sectors, ideally within R&D or product development, where he aims to contribute towards creating innovative products at a leading biomedical engineering company.

Within NYU, he served as a Teaching Assistant for the General Engineering Department at Tandon, where he taught first-year students introductory engineering skills such as 3D modelling and Arduino while guiding them in their semester-long projects. Within the department he also holds the Head of Purchasing role, ensuring the various facets of budgeting, funding and item inventory are maintained for the smooth workflow of the course. Currently, he operates as an Resident Assistant, ensuring the safety of the residents while organizing various events that introduce them to cultures and promote their professional development.

During his study abroad experience at NYU Abu Dhabi, Sishir became a research assistant for the LANTRN lab where he performed immunohistological and cryostat procedures on mice brain samples. During this time he also led a research project aiming to create an electrode that could deliver direct stimulation to gut tissue.

During his free time, Sishir enjoy swimming, badminton and occasionally shooting pool. Whenever he is able to, he enjoys trekking back in his hometown in India."