Research

---

DOI: 10.1109/ICONAT53423.2022.9726079

My friend Andrew and I began researching about how to make interpolating GEDI waveforms efficient and accurate about 8 months ago under the mentor ship of Brandon (Brown university PHD student). We have put down our findings in our research paper. Our research paper was presented at ICONAT IEEE conference in January 2022. Details of the paper are below.

Title:  Comparing Computer Resource Usage Through  Interpolating Global Ecosystem Dynamics  Investigation Light Detection and Ranging  Waveform Data

Authors: Arjun Raj, and Andrew Charles Baker

Abstract:  In our study we aim to compare the computer  resources utilized by cubic and linear splines. We will access and  use publicly available data collected by NASA’s Global Ecosystem  Dynamic Investigation’s (GEDI) Light Detection and Ranging, or  LiDAR, technology to model waveforms using cubic and linear b-spline interpolation. We will compare the viability and efficiency  of these two interpolation methods to determine which is more  applicable when working with GEDI data, and when working  with datasets similar in size, scope, and variability specifically as  it relates to artificial intelligence creation. This study will assist  researchers working with GEDI data and data like it to make  more informed decisions about creating models and the efficiency  of those models as it relates to large scale artificial intelligence  deployment. 

My friends and I created a Machine Learning algorithm, which uses CT scans to detect lung lesions caused by the coronavirus infection, and the involvement percentage.

The model's interface was developed as a web-based interface and a mobile application (under development) where test files can be uploaded.

The application of the artificial intelligence in healthcare is an actively developing area with a wide scope: from labeling pathology on medical scans to predicting diseases based on patient complaints.
At present, the problem of quick detection of lung lesions caused by the coronavirus infection on CT scans is still relevant. Resolving this issue will:

• decrease workload of radiologists
• reduce the likelihood of a medical error
• enable automated measurement of the affected lung volume when analyzing the quality of the therapy applied
• contribute to the development of the area focused on intellectual decision-making support systems.