• Journal of Biomedical Engineering Research
• /
• v.42 no.3
• /
• pp.71-79
• /
• 2021

Currently, the naked eyes-based diagnosis of bone metastases on CT images relies on qualitative assessment. For this reason, there is a great need for a state-of-the-art approach that can assess and follow-up the bone metastases with quantitative biomarker. Radiomics can be used as a biomarker for objective lesion assessment by extracting quantitative numerical values from digital medical images. In this study, therefore, we evaluated the clinical applicability of non-invasive and objective bone metastases computer-aided diagnosis using radiomics-based biomarkers in CT. We employed a total of 21 approaches consist of three-classifiers and seven-feature selection methods to predict bone metastases and select biomarkers. We extracted three-dimensional features from the CT that three groups consisted of osteoblastic, osteolytic, and normal-healthy vertebral bodies. For evaluation, we compared the prediction results of the classifiers with the medical staff's diagnosis results. As a result of the three-class-classification performance evaluation, we demonstrated that the combination of the random forest classifier and the sequential backward selection feature selection approach reached AUC of 0.74 on average. Moreover, we confirmed that 90-percentile, kurtosis, and energy were the features that contributed high in the classification of bone metastases in this approach. We expect that selected quantitative features will be helpful as biomarkers in improving the patient's survival and quality of life.

• Journal of the Korea Society of Computer and Information
• /
• v.26 no.11
• /
• pp.21-31
• /
• 2021

As the COVID-19 pandemic rapidly changes healthcare around the globe, the need for smart healthcare that allows for remote diagnosis is increasing. The current classification of respiratory diseases cost high and requires a face-to-face visit with a skilled medical professional, thus the pandemic significantly hinders monitoring and early diagnosis. Therefore, the ability to accurately classify and diagnose respiratory sound using deep learning-based AI models is essential to modern medicine as a remote alternative to the current stethoscope. In this study, we propose a deep learning-based respiratory sound classification model using data collected from medical experts. The sound data were preprocessed with BandPassFilter, and the relevant respiratory audio features were extracted with Log-Mel Spectrogram and Mel Frequency Cepstral Coefficient (MFCC). Subsequently, a Parallel CNN network model was trained on these two inputs using stacking ensemble techniques combined with various machine learning classifiers to efficiently classify and detect abnormal respiratory sounds with high accuracy. The model proposed in this paper classified abnormal respiratory sounds with an accuracy of 96.9%, which is approximately 6.1% higher than the classification accuracy of baseline model.

• Phonetics and Speech Sciences
• /
• v.14 no.4
• /
• pp.45-56
• /
• 2022

This paper proposes an optimal methodology for automatically diagnosing and predicting the severity of the chronic kidney disease (CKD) using patients' utterances. In patients with CKD, the voice changes due to the weakening of respiratory and laryngeal muscles and vocal fold edema. Previous studies have phonetically analyzed the voices of patients with CKD, but no studies have been conducted to classify the voices of patients. In this paper, the utterances of patients with CKD were classified using the variety of utterance types (sustained vowel, sentence, general sentence), the feature sets [handcrafted features, extended Geneva Minimalistic Acoustic Parameter Set (eGeMAPS), CNN extracted features], and the classifiers (SVM, XGBoost). Total of 1,523 utterances which are 3 hours, 26 minutes, and 25 seconds long, are used. F1-score of 0.93 for automatically diagnosing a disease, 0.89 for a 3-classes problem, and 0.84 for a 5-classes problem were achieved. The highest performance was obtained when the combination of general sentence utterances, handcrafted feature set, and XGBoost was used. The result suggests that a general sentence utterance that can reflect all speakers' speech characteristics and an appropriate feature set extracted from there are adequate for the automatic classification of CKD patients' utterances.

• Korean Journal of Radiology
• /
• v.25 no.1
• /
• pp.43-54
• /
• 2024

Objective: To evaluate the added value of diffusion-weighted imaging (DWI)-based quantitative parameters to distinguish uterine sarcomas from atypical leiomyomas on preoperative magnetic resonance imaging (MRI). Materials and Methods: A total of 138 patients (age, 43.7 ± 10.3 years) with uterine sarcoma (n = 44) and atypical leiomyoma (n = 94) were retrospectively collected from four institutions. The cohort was randomly divided into training (84/138, 60.0%) and validation (54/138, 40.0%) sets. Two independent readers evaluated six qualitative MRI features and two DWI-based quantitative parameters for each index tumor. Multivariable logistic regression was used to identify the relevant qualitative MRI features. Diagnostic classifiers based on qualitative MRI features alone and in combination with DWI-based quantitative parameters were developed using a logistic regression algorithm. The diagnostic performance of the classifiers was evaluated using a cross-table analysis and calculation of the area under the receiver operating characteristic curve (AUC). Results: Mean apparent diffusion coefficient value of uterine sarcoma was lower than that of atypical leiomyoma (mean ± standard deviation, 0.94 ± 0.30 10^-3 mm²/s vs. 1.23 ± 0.25 10^-3 mm²/s; P < 0.001), and the relative contrast ratio was higher in the uterine sarcoma (8.16 ± 2.94 vs. 4.19 ± 2.66; P < 0.001). Selected qualitative MRI features included ill-defined margin (adjusted odds ratio [aOR], 17.9; 95% confidence interval [CI], 1.41-503, P = 0.040), intratumoral hemorrhage (aOR, 27.3; 95% CI, 3.74-596, P = 0.006), and absence of T2 dark area (aOR, 83.5; 95% CI, 12.4-1916, P < 0.001). The classifier that combined qualitative MRI features and DWI-based quantitative parameters showed significantly better performance than without DWI-based parameters in the validation set (AUC, 0.92 vs. 0.78; P < 0.001). Conclusion: The addition of DWI-based quantitative parameters to qualitative MRI features improved the diagnostic performance of the logistic regression classifier in differentiating uterine sarcomas from atypical leiomyomas on preoperative MRI.

• Proceedings of the Korea Institute of Convergence Signal Processing
• /
• 2005.11a
• /
• pp.178-181
• /
• 2005

Support vector machines is a powerful learning algorithm and attempt to separate belonging to two given sets in N-dimensional real space by a nonlinear surface, often only implicitly dened by a kernel function. We described breast tissue images analyses using texture features from Haar wavelet transformed images to classify breast lesion of ductal organ Benign, DCIS and CA. The approach for creating a classifier is composed of 2 steps: feature extraction and classification. Therefore, in the feature extraction step, we extracted texture features from wavelet transformed images with $10{\times}$ magnification. In the classification step, we created four classifiers from each image of extracted features using SVM(Support Vector Machines). In this study, we conclude that the best classifier in histological sections of breast tissue in the texture features from second-level wavelet transformed images used in Polynomial function.

• Journal of Korea Multimedia Society
• /
• v.15 no.2
• /
• pp.195-203
• /
• 2012

This paper proposes ground target classification algorithm based on decision fusion and feature extraction method using multi-sensor images. The decisions obtained from the individual classifiers are fused by applying a weighted voting method to improve target recognition rate. For classifying the targets belong to the individual sensors images, features robust to scale and rotation are extracted using the difference of brightness of CM images obtained from CCD image and the boundary similarity and the width ratio between the vehicle body and turret of target in FLIR image. Finally, we verity the performance of proposed ground target classification algorithm and feature extraction method by the experimentation.

• Journal of Internet Computing and Services
• /
• v.13 no.2
• /
• pp.99-108
• /
• 2012

Many methods had been developed to improve the accuracy of extracting information from a vast amount of data. This paper combined a number of natural language processing methods such as NER (named entity recognition), sentence extraction, and part of speech tagging to carry out text analysis. The data source is comprised of texts obtained from the web using a domain-specific data extraction agent. A framework for the extraction of information from unstructured data was developed using the aforementioned natural language processing methods. We simulated the performance of our work in the extraction and analysis of texts for the detection of organizational structures. Simulation shows that our study outperformed other NER classifiers such as MUC and CoNLL on information extraction.

• 제어로봇시스템학회:학술대회논문집
• /
• 2004.08a
• /
• pp.2037-2040
• /
• 2004

In this paper we introduce a new prototype based classifiers for overlapping data, where training pattern can be overlap on the feature space. The proposed classifier is based on the prototype from neural network classifier (NNC)[1] for overlap data. The method automatically chooses the initial center and two radiuses for each class. The center is used as a mean representative of training data for each class. The unclassified pattern is classified by measure distance from the class center. If the distance is in the lower (shorter radius) the unknown pattern has the high percentage of being in this class. If the distance is between the lower and upper (further radius), the pattern has the probability of being in this class or others. But if the distance is outside the upper, the pattern is not in this class. We borrow the words upper and lower from the rough set to represent the region of certainty [3]. The training algorithm to find number of cluster and their parameters (center, lower, upper) is presented. The clustering result is tested using patterns from Thai handwritten letter and the clustering result is very similar to human eyes clustering.

• Journal of Institute of Control, Robotics and Systems
• /
• v.21 no.12
• /
• pp.1109-1114
• /
• 2015

This paper presents an improved learning-based visual inspection method for auto parts inspection in severe lighting changes. Automobile sunroof frames are produced automatically by robots in most production lines. In the sunroof frame manufacturing process, there is a quality problem with some parts such as volts are missed. Instead of manual sampling inspection using some mechanical jig instruments, a learning-based machine vision system was proposed in the previous research[1]. But, in applying the actual sunroof frame production process, the inspection accuracy of the proposed vision system is much lowered because of severe illumination changes. In order to overcome this capricious environment, some selective feature vectors and cascade classifiers are used for each auto parts. And we are able to improve the inspection accuracy through the re-learning concept for the misclassified data. The effectiveness of the proposed visual inspection method is verified through sufficient experiments in a real sunroof production line.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.14 no.9
• /
• pp.3762-3781
• /
• 2020

Infrared pedestrian target recognition is a vital problem of significant interest in computer vision. In this work, a novel infrared pedestrian target recognition method that uses heterogeneous features with multi-kernel transfer learning is proposed. Firstly, to exploit the characteristics of infrared pedestrian targets fully, a novel multi-scale monogenic filtering-based completed local binary pattern descriptor, referred to as MSMF-CLBP, is designed to extract the texture information, and then an improved histogram of oriented gradient-fisher vector descriptor, referred to as HOG-FV, is proposed to extract the shape information. Second, to enrich the semantic content of feature expression, these two heterogeneous features are integrated to get more complete representation for infrared pedestrian targets. Third, to overcome the defects, such as poor generalization, scarcity of tagged infrared samples, distributional and semantic deviations between the training and testing samples, of the state-of-the-art classifiers, an effective multi-kernel transfer learning classifier called MK-TrAdaBoost is designed. Experimental results show that the proposed method outperforms many state-of-the-art recognition approaches for infrared pedestrian targets.