KMITL

Permanent URI for this communityhttps://dspace.kmitl.ac.th/handle/123456789/1

Browse

Search Results

Now showing 1 - 10 of 10

Some of the metrics are blocked by your
consent settings
Impact of the Internet of Medical Things on Artificial Intelligence-enhanced medical imaging systems from 2019 to 2023
(2025-01-01)
Locharoenrat, Kitsakorn
This review addresses the disease diagnosis from brain, eye, and lung scan images based on non-invasive imaging technologies using the Internet of Medical Things (IoMT) and Artificial Intelligence (AI) systems, a topic that has been neglected in the recent literature. Combining imaging modalities with IoMT and AI is expected to enhance both medical diagnoses and personalized treatment plans. We searched various scientific databases for details on IoMT and AI in medical imaging technologies from 2019 to 2023, focusing on different imaging modalities. We investigated the performance of AI-based algorithms in imaging modalities such as X-ray, Computed Tomography, Magnetic Resonance Imaging, Positron Emission Tomography, and Optical Coherence Tomography using the following metrics: accuracy, precision, recall, sensitivity, specificity, and F-1 score, and then analyzed their balanced performance in six issues: enhancement of medical image quality, improvement of clinical diagnoses, support for clinical decision-making, consideration of input data, time efficiency, and data management. Advanced understanding of the IoMT and AI applications in medical imaging technologies would help identify unexplored opportunities and provide directions for future research to enhance the clinical applicability.
Some of the metrics are blocked by your
consent settings
Development of a screening method for adulteration detection in coconut milk via mid-infrared spectroscopy: A study of linear and nonlinear regression method
(2022-12-01)
Sitorus, Agustami
;
Bulan, Ramayanty
In the present study, we developed a screening method for detection of adulteration in coconut milk via mid-infrared spectroscopy. Linear and nonlinear regression methods (principal component regression (PCR), partial least squares regression (PLSR), and support vector machine regression (SVMR)) were employed and compared to achieve an optimal screening method. Spectral data were scanned using the FTIR benchtop with a wavelength range of 4000–16702 nm. The calibration models of the linear and nonlinear regression methods were developed using the leave-one-out cross-validation method before testing using predictive data that had been prepared. Furthermore, five spectral data treatment techniques were employed to improve the accuracy of the proposed calibration model. The results obtained show that the SVMR method is better than PCR and PLSR for the detection of adulteration in coconut milk by mid-infrared data spectroscopy. The coefficient of determination for calibration (R<sup>2</sup><inf>c</inf>) and prediction (R<sup>2</sup><inf>p</inf>), the root mean square error of calibration (RMSEC) and prediction (RMSEP) and the ratio of prediction to deviation (RPD) using the SVMR method were 100%, 0.81, 98.40%, 0.87 and 7.86, respectively. Furthermore, based on RPD analysis, it is known that the SVMR model can be used to perform excellent quality control of water-adulterated coconut milk.
Some of the metrics are blocked by your
consent settings
Wi-Fi Received Signal Strength-based Indoor Localization System Using K-Nearest Neighbors fingerprint integrated D∗algorithm
(2022-01-01)
Jarawan, Tanatthep
;
Kamsing, Patcharin
;
Torteeka, Peerapong
;
Manuthasna, Shariff
;
Hematulin, Warunyu
The indoor localization system is essential since the Global Positioning System cannot give an accurate position indoors, especially when several floor plans are considered. WiFi received signal strength becomes an alternative indicator for indoor localization systems. The experiment proposed a localization system created by integrating and working between the K-Nearest Neighbors algorithm and the D∗algorithm. The result illustrates the optimal path from the start point to the target point by avoiding the obstacle performing exceptionally well. The K-Nearest Neighbors algorithm provide the result for localization with Root Mean Square Errors of displacement at 1.190 meters, 2.491 meters, and 1.363 meters of X-Axis Y-Axis, respectively. The proposed indoor localization system can have various applications considering different environmental factors in different applications, such as the size of unmanned aerial vehicles when applying indoor unmanned aerial vehicles.
Some of the metrics are blocked by your
consent settings
Wi-Fi Received Signal Strength-based Indoor Localization System Using K-Nearest Neighbors fingerprint integrated D*algorithm
(2021-02-07)
Jarawan, Tanatthep
;
Kamsing, Patcharin
;
Tortceka, Peerapong
;
Manuthasna, Shariff
;
Hematulin, Warunyu
The indoor localization system is essential since the Global Positioning System cannot give an accurate position indoors, especially when several floor plans are considered. Wi-Fi received signal strength becomes an alternative indicator for indoor localization systems. The experiment proposed a localization system created by integrating and working between the K-Nearest Neighbors algorithm and the D*algorithm. The result illustrates the optimal path from the start point to the target point by avoiding the obstacle performing exceptionally well. The K-Nearest Neighbors algorithm provide the result for localization with Root Mean Square Errors of displacement at 1.190 meters, 2.491 meters, and 1.363 meters of X-Axis Y-Axis, respectively. The proposed indoor localization system can have various applications considering different environmental factors in different applications, such as the size of unmanned aerial vehicles when applying indoor unmanned aerial vehicles.
Some of the metrics are blocked by your
consent settings
Hard deterministic particle swarm optimisation for certain result solution
(2021-01-31)
Thongkrairat, Somsin
;
Chutchavong, Vanvisa
Stochastic optimisation's performance is globally acknowledged for its flexibility, robustness and performance - the result of optimisation is usually outstanding and acceptable. However, stochastic methods benefit from random variance to produce feasible results. However, there is one significant disadvantage when it comes to the certainty of results. The stochastic method may give an unsatisfied result. Even though it is an unlikely possibility, it can happen. Therefore, we propose an algorithm that eliminates this problem by changing the calculation core to a deterministic base. In this paper, we select Particle Swarm Optimisation (PSO) as the prototype algorithm and modify the particle moving method to generate an inevitable result. The results show that our process can produce a certain result solution, and the output from our deterministic algorithm is also acceptable, like the original algorithm.
Some of the metrics are blocked by your
consent settings
Optimum Generated Power with the Minimum Cost of a Radial Network using Firefly and Genetic Algorithms
(2021-01-01)
Abdelwareth, Mohamed Els S.
;
Riawan, Dedet Candra
;
Chompoo-Inwai, Chow
This paper presented two artificial intelligence methods to find the optimum output power from Diesel generator (DG), Photovoltaic system (PV) and batteries to satisfy the load with the minimum cost considering the minimum losses. Our case study was a micro-gird 20 kV radial network consists of 21 busses located in Tomia island, south-east Sulawesi Island, Indonesia. Firefly algorithm (FA) and Genetic algorithm (GA) used in this study to do the optimization and chose the optimum operation. Forward-Backward sweep method used for the power flow calculations.
Some of the metrics are blocked by your
consent settings
The development of Artificial Neural Networks (ANN) for falls detection
(2017-06-07)
Yodpijit, Nantakrit
;
Sittiwanchai, Teppakorn
;
Jongprasithporn, Manutchanok
This paper presents the new design and development of a wearable-based fall detection system using an Accelerometer and Gyroscope as motion sensors for detecting body orientation and movement. The Threshold Based and Artificial Neural Networks (ANN) algorithm were developed to differentiate between Activities of Daily Living (ADL). Results indicated the possibility of using the new threshold-based method with ANN algorithm to reduce the number of false positive (false alarm) outcomes and improve the accuracy of fall detection system.
Some of the metrics are blocked by your
consent settings
Simplified genetic algorithm: Simplify and improve RGA for parameter optimizations
(2014-01-01)
Ngamtawee, Ruksaphol
;
Wardkein, Paramote
The structural complexity and complicated generic operators of Genetic Algorithm (GA) contribute to its slow computational speed. Furthermore, GA and other similar algorithms with a small population size are vulnerable to the problem of premature convergence. Premature convergence causes the algorithms to stagnate and stop searching, giving rise to wasteful computation. Even though the problem can be addressed with a larger population size, computational time is inevitably increased. This research paper has thus proposed Simplified Genetic Algorithm (SimpGA). This algorithm utilizes a one-pair-builtall structure in which only two parent chromosomes are required to produce the entire population (offspring). Rather than relying on the conventional operators, simplified operators, i.e. timer mutation, diform crossover and topmost selection, are used in the proposed SimpGA. In addition, tests are carried out with SimpGA on four test functions and four applications. The experimental results show that SimpGA is simpler to implement and performs well, especially in a small population environment.
Some of the metrics are blocked by your
consent settings
An efficient dynamic multicast routing algorithm with advance resource reservation awareness
(2004-06-28)
Pavarangkoon, Praphan
;
Gunabhibal, Apipol
;
Pornavalai, Chotipat
;
Varakulsiripunth, Ruttikorn
Reserving the resources for the requested applications is one of the most effective schemes that were proposed to offer the time-critical multimedia applications recently. Because the resource is limited, an ability to provide resource reservation in advance is essential in multi-party applications with dynamic accessing and leaving of user such as modern integrated (voice, video and data) collaboration system. Modified Greedy (MG) algorithm was proposed for multicast routing in advance reservation environment (time of joining in and leaving from multicast group need to be informed to source node). However, MG algorithm has an assumption that member of multicast group join in and leave from multicast session punctually. In this work, we propose the late version of MG algorithm that yields a good performance in the situation that the inaccuracy of information given by user might be happened, as shown in our simulation results.
Some of the metrics are blocked by your
consent settings
A congestion-control policy on the Internetwork gateway
(1986-01-01)
Varakulsiripunth, Ruttikorn
;
Shiratori, Norio
;
Noguchi, Shoichi
A new approach to congestion-control policy for the interconnected networks system, called "FCP", is proposed. The main purpose of the FCP policy is to avoid the congestion of a network as well as to guarantee the performance of each network up to the desired value. In FCP policy, the traffic that arrives at aateway is restricted by that gateway in compliance with the acceptable rate, called "PAR", corresponding to its destination network. And the PAR corresponding to a network is determined based on the performance requirement of that network. The determination of PAR in FCP policy is presented and analyzed quantitatively in terms of network throughput, whereby the queuing models for a gateway and a network are developed. The result of the analytical evaluation of FCP policy is reduced to the problem of solving simultaneous non-linear equations with several variables and of a higher degree. Therefore, the systematic algorithm for computing PARs in FCP policy is also given. The numerical applications demonstrate the good efficiency of FCP policy to achieve the desired performance for each network. © 1986.

KMITL

Browse

Filters

Settings

Sort By

Results per page

Search Results