Ajin Joy

Dr. Ajin Joy

Radiological Sciences, David Geffen School of Medicine at UCLA, Los Angeles, California, USA
orcid

Research Focus & Metrics

Academic Career Length

13Years (2014-Present)

(Calculated once Scilit publication data is synced: current year − year of first publication + 1)

Representative Publications

View All Publications
7 June 2017 in Magnetic Resonance in Medicine
Verified
Journal Article
Closed access
PDF file
Ajin JoyAjin Joy
Ajin Joy
JPJoseph Suresh Paul
Joseph Suresh Paul

7 March 2018 in Magnetic Resonance in Medicine
Verified
Journal Article
Closed access
PDF file
Ajin JoyAjin Joy
Ajin Joy
JPJoseph Suresh Paul
Joseph Suresh Paul

1 December 2014
Verified
Conference Paper
Closed access
Ajin JoyAjin Joy
Ajin Joy
VJV. S. Jayanthi
V. S. Jayanthi
DBD. Baskar
D. Baskar
The increase in the number of vehicles has seen a rise in the accident rate. Pedestrians and vehicles on the move affect the behavior of the driver extensively. As an effort to reduce the number of accidents by assisting the driver, we propose a software system based on neural network and optical flow analysis to detect objects as observed by a driver. The proposed system tracks vehicles by detecting its number plate and the optical flow associated with it. It has four stages. First, the system is trained by using a number of images. In the second stage, object detection is performed for each image of the sequence using the trained system. Following this, the optical flow is calculated for each image as an estimate of the motion in the scene. In the fourth and final stage, optical flow is combined with the detected objects and both their direction and magnitude are identified. The results obtained detects cautions and alerts the driver about the impending dangerous situations that might befall. This software is also suitable for systems to study the behavior of the driver and if the software is used in all vehicles on the road, creating a database of the detected number plates, it can give the location of the stream of vehicles on the road.

30 July 2019 in Magnetic Resonance in Medicine
Verified
Journal Article
Closed access
PDF file
Ajin JoyAjin Joy
Ajin Joy
MJMathews Jacob
Mathews Jacob
JPJoseph Suresh Paul
Joseph Suresh Paul

5 October 2023 in Scientific Reports
Verified
Journal Article
Open access
PDF file
Ajin JoyAjin Joy
Ajin Joy
michael albert thomasmichael albert thomas
michael albert thomas
Acquisition duration of correlated spectroscopy in vivo can be longer due to a large number of t1 increments along the indirect (F1) dimension. Limited number of t1 increments on the other hand leads to poor spectral resolution along F1. Covariance transformation (CT) instead of Fourier transform along t1 is an alternative way of increasing the resolution of the 2D COSY spectrum. Prospectively undersampled five-dimensional echo-planar correlated spectroscopic imaging (EP-COSI) data from ten malignant patients and ten healthy women were acquired and reconstructed using compressed sensing. The COSY spectrum at each voxel location was then generated using FFT, CT and a variant of CT called Inner Product (IP). Metabolite and lipid ratios were computed with respect to water from unsuppressed one-dimensional spectrum. The effects of t1-ridging artifacts commonly seen with FFT were not observed with CT/IP. Statistically significant differences were observed in the fat cross peaks measured with CT/IP/FFT. Spectral resolution was increased ~ 8.5 times (~ 19.53 Hz in FFT, ~ 2.32 Hz in CT/IP) without affecting the spectral width along F1 was possible with CT/IP. CT and IP enabled substantially increased F1 resolution effectively with significant gain in scan time and reliable measure of unsaturation index as a biomarker for malignant breast cancer.

11 July 2023 in Metabolites
Verified
Journal Article
Open access
PDF file
Ajin JoyAjin Joy
Ajin Joy
Marlene LinMarlene Lin
Marlene Lin
MJMelissa Joines
Melissa Joines
ASAndres Saucedo
Andres Saucedo
SLStephanie Lee-Felker
Stephanie Lee-Felker
michael albert thomasmichael albert thomas
michael albert thomas
The main objective of this work was to evaluate the application of individual and ensemble machine learning models to classify malignant and benign breast masses using features from two-dimensional (2D) correlated spectroscopy spectra extracted from five-dimensional echo-planar correlated spectroscopic imaging (5D EP-COSI) and diffusion-weighted imaging (DWI). Twenty-four different metabolite and lipid ratios with respect to diagonal fat peaks (1.4 ppm, 5.4 ppm) from 2D spectra, and water and fat peaks (4.7 ppm, 1.4 ppm) from one-dimensional non-water-suppressed (NWS) spectra were used as the features. Additionally, water fraction, fat fraction and water-to-fat ratios from NWS spectra and apparent diffusion coefficients (ADC) from DWI were included. The nine most important features were identified using recursive feature elimination, sequential forward selection and correlation analysis. XGBoost (AUC: 93.0%, Accuracy: 85.7%, F1-score: 88.9%, Precision: 88.2%, Sensitivity: 90.4%, Specificity: 84.6%) and GradientBoost (AUC: 94.3%, Accuracy: 89.3%, F1-score: 90.7%, Precision: 87.9%, Sensitivity: 94.2%, Specificity: 83.4%) were the best-performing models. Conventional biomarkers like choline, myo-Inositol, and glycine were statistically significant predictors. Key features contributing to the classification were ADC, 2D diagonal peaks at 0.9 ppm, 2.1 ppm, 3.5 ppm, and 5.4 ppm, cross peaks between 1.4 and 0.9 ppm, 4.3 and 4.1 ppm, 2.3 and 1.6 ppm, and the triglyceryl–fat cross peak. The results highlight the contribution of the 2D spectral peaks to the model, and they demonstrate the potential of 5D EP-COSI for early breast cancer detection.