Quantitative Predictive Imaging Biomarkers of Lumbar Intervertebral Disc Degeneration

Rammohan Vadapalli,Raghavdutt Mulukutla,Abhinav Sriram Vadapalli,Rajanikanth Rao Vedula 대한척추외과학회 2019 Asian Spine Journal Vol.13 No.4

Study Design: Observational comparative study. Purpose: To compare fractional anisotropy (FA) maps with T2 values of the nucleus pulposus (NP) and annulus fibrosus (AF) of intervertebral discs in healthy volunteers and patients to develop a predictive disc health scale. Overview of Literature: T2-weighted magnetic resonance imaging (MRI) is not sensitive to early morphological changes and provides no quantitative biomarker profile for early degeneration. Methods: We examined 59 healthy controls and 59 patients with back pain by MRI using T2 relaxometry and diffusion tensor imaging (DTI). Each group was divided into three age subgroups: A (<30 years, n=12); B (30–50 years, n=26); and C (>50 years, n=21). We obtained FA values for AF and NP and T2 values for NP for each intervertebral disc. Furthermore, we calculated the FA (AF/NP) ratios. Results: We categorized 590 intervertebral discs from 118 participants, 566 of which were analyzed with T2 relaxometry and DTI. The T2 values were as follows: subgroup A, 55.8±4.4 ms; B, 48.5±6.9 ms; C, 45.8±8.7 ms (p <0.050). The T2 values for the healthy controls of the subgroups A, B, and C were >120 ms, 90–100 ms, and 70 ms, respectively (p <0.001). Control subgroup A had higher T2 values and AF/NP ratios than subgroups B and C; the AF values were not significantly different. Control subgroup B had higher T2 values and AF/NP ratios than subgroup C but lower FA (NP). Conclusions: FA maps of the AF/NP ratio and T2 values of NP are potential microstructure biomarkers of normal and degenerating discs and can help detect early degeneration using a predictive disc health score on a continuous scale.

Restoring one way traffic across the Roux loop: The Nakajo value

Ashwin Rammohan,Sripriya Srinivas,Yahya Al Azri,Mettu Srinivas Reddy,Mohamed Rela 한국간담췌외과학회 2018 Annals of hepato-biliary-pancreatic surgery Vol.22 No.1

Precision Livestock Feed Production Using Dynamic Approaches

( Uyeh Daniel Dooyum ),( Rammohan Mallipeddi ),( Trinadh Pamulapati ),( Tusan Park ),( Seungmin Woo ),( Junhee Kim ),( Yushin Ha ) 한국농업기계학회 2018 한국농업기계학회 학술발표논문집 Vol.23 No.2

Increasing human populations are leading to increasing and sometimes conflicting demands on land and resources. Agriculture, particularly the livestock industry will need to increase productivity using existing or fewer resources to meet demands. A resilient livestock industry will require cost-effective feed rations as feed accounts for between 60 to 80% of production costs. Techniques in computational intelligence were used in coding and fine-tuning algorithms for feed formulation. Simulations were done in MATLAB to evaluate the performance of the algorithms. Hyperspectral imaging and near infrared spectroscopy were used to acquire spectral data from mixed by-products and packaged feed. Preprocessing methods such as second derivative and normalization were applied to the data to remove noise. Data was acquired from a spectral range of 900 to 2100 nm. Prediction tools (Partial least square regression and Unscrambler) were used to build amino acids, sugars, organic acids and moisture content models. This presents rapid, non-destructive measurements with no sample preparation and high sensitivity. All the models had satisfactory R² values. Subsequently, the developed systems were combined with existing systems to propose models for precision livestock feed production.

Optimal robot utilization in protected cultivation systems

( Daniel Dooyum Uyeh ),( Rammohan Mallipeddi ),( Tusan Park ),( Hyunggyu Park ),( Yeongsu Kim ),( Seokho Kang ),( Junhee Kim ),( Yushin Ha ),( Seungmin Woo ) 한국농업기계학회 2020 한국농업기계학회 학술발표논문집 Vol.25 No.2

There is growing demand for premium and safe agricultural products. Greenhouses and plant factories which are types of protected cultivation system are being adopted to meet demand. This is because of ease in regulating the climate required for optimal plant growth which can facilitate year-round cultivation of these agriculture products and many other numerous benefits. However, poor ventilation in greenhouses can be harmful to the human workforce. This has led to the development of robots for hazardous tasks. However, mobile robots are required to navigate down every aisle to perform a task in a greenhouse, and it is difficult to predict at which point the robot will need to return to the start point, to offload or refill for transportation and spraying schedules, respectively or battery charges. This necessitates optimal scheduling of robots. In this study, we formulated the robot scheduling problem and adopted and tunned an evolutionary algorithm to solve this problem using MATLAB software. We performed simulations of different robot working scenarios integrating various parameters. Different scheduling scenarios showed varied patterns. Initial state of charge of the robot batteries impacted the working state of the robots. For examples, where the robots had fixed rates of charge and discharge of 5%, there were also more continuously working hours of robots in the two studied cases. A system for optimal scheduling of robots in a greenhouse to maximize the total work hours and sustain continuous operation was developed and proposed.

Ensemble of Support Vector Machine and Meta classification

Samira ghorbanpourbarforoushi,Vikas Palakonda,Rammohan Mallipeddi 한국통신학회 2018 한국통신학회 학술대회논문집 Vol.2018 No.6

Optimization of Total Mixed Ration (TMR) feed formulation for Cattle Using Evolutionary Systems

( Uyeh Daniel Dooyum ),( Rammohan Mallipeddi ),( Trinadh Pamulapati ),( Tusan Park ),( Junhee Kim ),( Seungmin Woo ),( Yushin Ha ) 한국농업기계학회 2017 한국농업기계학회 학술발표논문집 Vol.22 No.2

Different food products obtained from cattle are important sources of nutrients to humans worldwide. About 60~80% of production cost is incurred by animal feed which is dependent on the class and phase of the animal. This is a non-deterministic polynomial-time hardness (NP-hard) recurrent problem due to the existence of both linear and non-linear constraints. For most programming tools, satisfying constraints at specific stages of breeding are often not possible. This entails the advancement of a program that has an intrinsic mechanism to allow easing of constraints based on the necessities of it's users. Consequently, the formulation of the problem and system of optimization should facilitate this. We modified the conventional formulations using parameters that allows tolerance that accommodate constraints slackening and proposed a method exploring evolutionary systems. Results of our experimental trials indicated that easing requirements has direct connections with the cost and in some instances quantity of the feed. We propose a system which possesses the capability to explore for a feasible and least cost result if available or the best solution if otherwise and find the suitable components to be used in the formulation.

A Genetic Algorithm-Based Moving Object Detection for Real-time Traffic Surveillance

Giyoung Lee,Mallipeddi, Rammohan,Gil-Jin Jang,Minho Lee IEEE 2015 IEEE signal processing letters Vol.22 No.10

<P>Recent developments in vision systems such as distributed smart cameras have encouraged researchers to develop advanced computer vision applications suitable to embedded platforms. In the embedded surveillance system, where memory and computing resources are limited, simple and efficient computer vision algorithms are required. In this letter, we present a moving object detection method for real-time traffic surveillance applications. The proposed method is a combination of a genetic dynamic saliency map (GDSM), which is an improved version of dynamic saliency map (DSM) and background subtraction. The experimental results show the effectiveness of the proposed method in detecting moving objects.</P>

Trajectory-based vehicle tracking at low frame rates

Lee, Giyoung,Mallipeddi, Rammohan,Lee, Minho Elsevier 2017 expert systems with applications Vol.80 No.-

<P><B>Abstract</B></P> <P>In smart cities, an intelligent traffic surveillance system plays a crucial role in reducing traffic jams and air pollution, thus improving the quality of life. An intelligent traffic surveillance should be able to detect and track multiple vehicles in real-time using only limited resources. Conventional tracking methods usually run at a high video-sampling rate, assuming that the same vehicles in successive frames are similar and move only slightly. However, in cost effective embedded surveillance systems (e.g., a distributed wireless network of smart cameras), video frame rates are typically low because of limited system resources. Therefore, conventional tracking methods perform poorly in embedded surveillance systems because of discontinuity of the moving vehicles in the captured recordings. In this study, we present a fast and light algorithm that is suitable for an embedded real-time visual surveillance system to detect effectively and track multiple moving vehicles whose appearance and/or position changes abruptly at a low frame rate. For effective tracking at low frame rates, we propose a new matching criterion based on greedy data association using appearance and position similarities between detections and trackers. To manage abrupt appearance changes, manifold learning is used to calculate appearance similarity. To manage abrupt changes in motion, the next probable centroid area of the tracker is predicted using trajectory information. The position similarity is then calculated based on the predicted next position and progress direction of the tracker. The proposed method demonstrates efficient tracking performance during rapid feature changes and is tested on an embedded platform (ARM with DSP-based system).</P> <P><B>Highlights</B></P> <P> <UL> <LI> A new vehicle tracking method is proposed for an embedded traffic surveillance system. </LI> <LI> The proposed method demonstrates efficient tracking performance at a low frame rate. </LI> <LI> The proposed method employs greedy data association based on appearance and position similarities. </LI> <LI> To manage abrupt appearance changes, manifold learning is used. </LI> <LI> To manage abrupt motion changes, trajectory information is used to predict the next probable position. </LI> </UL> </P>

안구운동 기반의 사용자 묵시적 의도 판별 분석 모델

장영민(Young-Min Jang),Rammohan Mallipeddi,김철수(Cheol-Su Kim),이민호(Minho Lee) 대한전자공학회 2013 전자공학회논문지 Vol.50 No.6

최근 사용자의 생체 신호 정보를 기반으로 사용자 인지향상을 위하여, 상황에 적합한 서비스를 제공하기 위한 인간-컴퓨터/기계 상호작용 (Human computer/machine interaction: HCI/HMI) 시스템이 급격하게 증가하고 있는 추세이다. 이와 같이 인간-컴퓨터/기계 상호작용 기반의 효과적인 사용자 인지향상 시스템을 개발하기 위해서는 사용자의 명시적 의도 파악과 더불어 사용자의 묵시적 의도 파악이 중요하다. 사람의 시각 운동 이론에 따르면, 사람의 안구운동 정보와 동공 반응은 사람의 의도와 행동에 대하여 많은 량의 정보를 제공한다. 이에 본 논문에서는 사용자의 묵시적 의도를 판별하기 위하여, 피험자에게 제공되는 자극영상의 관심(흥미) 영역 (area of interest: AOI) 내에서의 안구운동 패턴인 응시 시간/횟수, 동공 응답 패턴의 동공크기와 동공의 크기변화인 기울기 정보를 분석하는 새로운 접근 방법을 제안한다. 제안하는 모델은 항행적 의도 발생, 정보적 의도발생, 정보적 의도 소멸과 같은 세 가지 유형으로 인간의 묵시적 의도를 식별한다. 여기서 항행적 의도란 주어진 자극영상 내에서 무언가 흥미로운 것을 찾는 행위를 말하며, 이에 반해 정보적 의도는 특정 위치에서 특정 객체는 찾는 행위를 의미한다. 본 연구에서는 사용자 안구운동 패턴과 동공분석 정보 기반으로 서로 다른 묵시적 의도인 항행적 의도, 정보적 의도 발생, 그리고 정보적 의도 소멸 사이에서 그 천이를 감지할 수 있는 계층적 SVM (hierarchical support vector machine: H-SVM)을 이용하였다. Recently, there has been tremendous increase in human-computer/machine interaction system, where the goal is to provide with an appropriate service to the user at the right time with minimal human inputs for human augmented cognition system. To develop an efficient human augmented cognition system based on human computer/machine interaction, it is important to interpret the user’s implicit intention, which is vague, in addition to the explicit intention. According to cognitive visual-motor theory, human eye movements and pupillary responses are rich sources of information about human intention and behavior. In this paper, we propose a novel approach for the identification of human implicit visual search intention based on eye movement pattern and pupillary analysis such as pupil size, gradient of pupil size variation, fixation length/count for the area of interest. The proposed model identifies the human’s implicit intention into three types such as navigational intent generation, informational intent generation, and informational intent disappearance. Navigational intent refers to the search to find something interesting in an input scene with no specific instructions, while informational intent refers to the search to find a particular target object at a specific location in the input scene. In the present study, based on the human eye movement pattern and pupillary analysis, we used a hierarchical support vector machine which can detect the transitions between the different implicit intents - navigational intent generation to informational intent generation and informational intent disappearance.

Interactive livestock feed ration optimization using evolutionary algorithms

Dooyum, Uyeh Daniel,Mallipeddi, Rammohan,Pamulapati, Trinadh,Park, Tusan,Kim, Junhee,Woo, Seungmin,Ha, Yushin Elsevier 2018 Computers and electronics in agriculture Vol.155 No.-

<P> The profitability of the livestock industry largely depends on cost-effective feed ration formulation as feed accounts for between 60 and 80% of production costs. Therefore, feed formulation is a recurring problem for breeders. In addition, the presence of linear and non-linear constraints, and multiple possible combinations that are subject to upsurge makes the formulation of feed a Non-deterministic Polynomial-time hard (NP-hard) problem. Generally, feed formulation is done by specifying the nutritional requirements as rigid constraints and an algorithm attempts to find a feasible cost-effective formulation. However, relaxing the constraints can sometimes provide a huge reduction in the cost of feed while not seriously affecting the economic performance of the livestock. This entails the development of a feed formulation software that has an inbuilt mechanism to enable relaxation to the constraints based on the users’ necessities. Accordingly, the problem formulation and the optimization algorithm should facilitate this. We modified the conventional problem formulation with a tolerance parameter (as a percentage of the actual value) to accommodate the relaxation of constraints. We solved this problem with differential evolution, a variant of evolutionary algorithms, which are good for handling NPhard problems. In addition, the relaxation of the constraints was done in an interactive way using the proposed method without penalties. In other words, the proposed method is flexible and possesses the ability to search for a feasible and least-cost solution if available or otherwise, the best solution and finds the suitable feed components to be used in ration formulation at an optimal cost depending on the nutrient requirements and growth stage of the animal. </P>