HER2GAN: Overcome the Scarcity of HER2 Breast Cancer Dataset Based on Transfer Learning and GAN Model.

INTRODUCTION: Immunohistochemistry (IHC) is crucial for breast cancer diagnosis, classification, and individualized treatment. IHC is used to measure the levels of expression of hormone receptors (estrogen and progesterone receptors), human epidermal growth factor receptor 2 (HER2), and other biomarkers, which are used to make treatment decisions and predict how well a patient will do. The evaluation of the breast cancer score on IHC slides, taking into account structural and morphological features as well as a scarcity of relevant data, is one of the most important issues in the IHC debate. Several recent studies have utilized machine learning and deep learning techniques to resolve these issues. MATERIALS AND METHODS: This paper introduces a new approach for addressing the issue based on supervised deep learning. A GAN-based model is proposed for generating high-quality HER2 images and identifying and classifying HER2 levels. Using transfer learning methodologies, the original and generated images were evaluated. RESULTS AND CONCLUSION: All of the models have been trained and evaluated using publicly accessible and private data sets, respectively. The InceptionV3 and InceptionResNetV2 models achieved a high accuracy of 93% with the combined generated and original images used for training and testing, demonstrating the exceptional quality of the details in the synthesized images.

Multimodality radiomics prediction of radiotherapy-induced the early proctitis and cystitis in rectal cancer patients: a machine learning study.

Purpose.This study aims to predict radiotherapy-induced rectal and bladder toxicity using computed tomography (CT) and magnetic resonance imaging (MRI) radiomics features in combination with clinical and dosimetric features in rectal cancer patients.Methods.A total of sixty-three patients with locally advanced rectal cancer who underwent three-dimensional conformal radiation therapy (3D-CRT) were included in this study. Radiomics features were extracted from the rectum and bladder walls in pretreatment CT and MR-T2W-weighted images. Feature selection was performed using various methods, including Least Absolute Shrinkage and Selection Operator (Lasso), Minimum Redundancy Maximum Relevance (MRMR), Chi-square (Chi2), Analysis of Variance (ANOVA), Recursive Feature Elimination (RFE), and SelectPercentile. Predictive modeling was carried out using machine learning algorithms, such as K-nearest neighbor (KNN), Support Vector Machine (SVM), Logistic Regression (LR), Decision Tree (DT), Random Forest (RF), Naive Bayes (NB), Gradient Boosting (XGB), and Linear Discriminant Analysis (LDA). The impact of the Laplacian of Gaussian (LoG) filter was investigated with sigma values ranging from 0.5 to 2. Model performance was evaluated in terms of the area under the receiver operating characteristic curve (AUC), accuracy, precision, sensitivity, and specificity.Results.A total of 479 radiomics features were extracted, and 59 features were selected. The pre-MRI T2W model exhibited the highest predictive performance with an AUC: 91.0/96.57%, accuracy: 90.38/96.92%, precision: 90.0/97.14%, sensitivity: 93.33/96.50%, and specificity: 88.09/97.14%. These results were achieved with both original image and LoG filter (sigma = 0.5-1.5) based on LDA/DT-RF classifiers for proctitis and cystitis, respectively. Furthermore, for the CT data, AUC: 90.71/96.0%, accuracy: 90.0/96.92%, precision: 88.14/97.14%, sensitivity: 93.0/96.0%, and specificity: 88.09/97.14% were acquired. The highest values were achieved using XGB/DT-XGB classifiers for proctitis and cystitis with LoG filter (sigma = 2)/LoG filter (sigma = 0.5-2), respectively. MRMR/RFE-Chi2 feature selection methods demonstrated the best performance for proctitis and cystitis in the pre-MRI T2W model. MRMR/MRMR-Lasso yielded the highest model performance for CT.Conclusion.Radiomics features extracted from pretreatment CT and MR images can effectively predict radiation-induced proctitis and cystitis. The study found that LDA, DT, RF, and XGB classifiers, combined with MRMR, RFE, Chi2, and Lasso feature selection algorithms, along with the LoG filter, offer strong predictive performance. With the inclusion of a larger training dataset, these models can be valuable tools for personalized radiotherapy decision-making.

Differentiation between Pancreatic Ductal Adenocarcinoma and Normal Pancreatic Tissue for Treatment Response Assessment using Multi-Scale Texture Analysis of CT Images.

Background: Pancreatic ductal adenocarcinoma (PDAC) is the most prevalent type of pancreas cancer with a high mortality rate and its staging is highly dependent on the extent of involvement between the tumor and surrounding vessels, facilitating treatment response assessment in PDAC. Objective: This study aims at detecting and visualizing the tumor region and the surrounding vessels in PDAC CT scan since, despite the tumors in other abdominal organs, clear detection of PDAC is highly difficult. Material and Methods: This retrospective study consists of three stages: 1) a patch-based algorithm for differentiation between tumor region and healthy tissue using multi-scale texture analysis along with L1-SVM (Support Vector Machine) classifier, 2) a voting-based approach, developed on a standard logistic function, to mitigate false detections, and 3) 3D visualization of the tumor and the surrounding vessels using ITK-SNAP software. Results: The results demonstrate that multi-scale texture analysis strikes a balance between recall and precision in tumor and healthy tissue differentiation with an overall accuracy of 0.78±0.12 and a sensitivity of 0.90±0.09 in PDAC. Conclusion: Multi-scale texture analysis using statistical and wavelet-based features along with L1-SVM can be employed to differentiate between healthy and pancreatic tissues. Besides, 3D visualization of the tumor region and surrounding vessels can facilitate the assessment of treatment response in PDAC. However, the 3D visualization software must be further developed for integrating with clinical applications.

Endorectal ultrasound radiomics in locally advanced rectal cancer patients: despeckling and radiotherapy response prediction using machine learning.

PURPOSE: The current study aimed to evaluate the association of endorectal ultrasound (EUS) radiomics features at different denoising filters based on machine learning algorithms and to predict radiotherapy response in locally advanced rectal cancer (LARC) patients. METHODS: The EUS images of forty-three LARC patients, as a predictive biomarker for predicting the treatment response of neoadjuvant chemoradiotherapy (NCRT), were investigated. For despeckling, the EUS images were preprocessed by traditional filters (bilateral, wiener, lee, frost, median, and wavelet filters). The rectal tumors were delineated by two readers separately, and radiomics features were extracted. The least absolute shrinkage and selection operator were used for feature selection. Classifiers including logistic regression (LR), K-nearest neighbor (KNN), support vector machine (SVM), random forest, naive Bayes, and decision tree were trained using stratified fivefold cross-validation for model development. The area under the curve (AUC) of the receiver operating characteristic curve followed by accuracy, precision, sensitivity, and specificity were obtained for model performance assessment. RESULTS: The wavelet filter had the best results with means of AUC: 0.83, accuracy: 77.41%, precision: 82.15%, and sensitivity: 79.41%. LR and SVM by having AUC: 0.71 and 0.76; accuracy: 70.0% and 71.5%; precision: 75.0% and 73.0%; sensitivity: 69.8% and 80.2%; and specificity: 70.0% and 60.9% had the highest model's performance, respectively. CONCLUSION: This study demonstrated that the EUS-based radiomics model could serve as pretreatment biomarkers in predicting pathologic features of rectal cancer. The wavelet filter and machine learning methods (LR and SVM) had good results on the EUS images of rectal cancer.

Development and validation of an individual job performance questionnaire (IJPQ).

BACKGROUND: Developing reliable tools to tap into all the behavioral dimensions of individual job performance and identifying the right sub-dimensions is necessary for both research and practice. OBJECTIVE: This study aimed at developing and validating an IJPQ that addresses shortcomings of existing questionnaires. METHODS: After a comprehensive systematic literature review, a framework consisting of four dimensions, including task performance (TP), contextual performance (CP), counterproductive work behavior (CWB), and adaptive performance (AP) was structured for measuring IJP. As well, 45 sub-dimensions were identified for measuring IJP's dimensions. Content and face validity were evaluated, and item impact score (IS), content validity index (CVI), Kappa, and content validity ratio (CVR) were calculated. For reliability and confirmatory factor analysis (CFA), 525 workers completed the validated questionnaire and Cronbach alpha and goodness of fit indexes were determined, respectively. RESULTS: Of the 62 items generated to measure dimensions, 53 were approved. Based on item-level CVI, of the 53 items, only 45 items were accepted. Finally, the results of item level CVR led to the extraction of 27 questions to evaluate IJP. The obtained scale level CVI and scale level CVR were 0.91 and 0.68, respectively. Based on the results obtained from 525 Iranian workers, values of Cronbach's Alpha, X2/df, RMSEA, and P-value were in the acceptable range. CONCLUSIONS: Conclusively, a questionnaire containing 20 items was developed and validated for measuring IJP of Iranian worker's culture. The four dimensions of TP, CO, CWB, and AP consisted of 6, 5, 5, and 4 items each, respectively. Overall, IJPQ is a theory-based, reliable, and valid instrument for assessing job performance.

Segmentation of pancreatic ductal adenocarcinoma (PDAC) and surrounding vessels in CT images using deep convolutional neural networks and texture descriptors.

Fully automated and volumetric segmentation of critical tumors may play a crucial role in diagnosis and surgical planning. One of the most challenging tumor segmentation tasks is localization of pancreatic ductal adenocarcinoma (PDAC). Exclusive application of conventional methods does not appear promising. Deep learning approaches has achieved great success in the computer aided diagnosis, especially in biomedical image segmentation. This paper introduces a framework based on convolutional neural network (CNN) for segmentation of PDAC mass and surrounding vessels in CT images by incorporating powerful classic features, as well. First, a 3D-CNN architecture is used to localize the pancreas region from the whole CT volume using 3D Local Binary Pattern (LBP) map of the original image. Segmentation of PDAC mass is subsequently performed using 2D attention U-Net and Texture Attention U-Net (TAU-Net). TAU-Net is introduced by fusion of dense Scale-Invariant Feature Transform (SIFT) and LBP descriptors into the attention U-Net. An ensemble model is then used to cumulate the advantages of both networks using a 3D-CNN. In addition, to reduce the effects of imbalanced data, a multi-objective loss function is proposed as a weighted combination of three classic losses including Generalized Dice Loss (GDL), Weighted Pixel-Wise Cross Entropy loss (WPCE) and boundary loss. Due to insufficient sample size for vessel segmentation, we used the above-mentioned pre-trained networks and fine-tuned them. Experimental results show that the proposed method improves the Dice score for PDAC mass segmentation in portal-venous phase by 7.52% compared to state-of-the-art methods in term of DSC. Besides, three dimensional visualization of the tumor and surrounding vessels can facilitate the evaluation of PDAC treatment response.

Assessment of Motor Cortex in Active, Passive and Imagery Wrist Movement Using Functional MRI.

BACKGROUND: Functional Magnetic resonance imaging (fMRI) measures the small fluctuation of blood flow happening during task-fMRI in brain regions. OBJECTIVE: This research investigated these active, imagery and passive movements in volunteers design to permit a comparison of their capabilities in activating the brain areas. MATERIAL AND METHODS: In this applied research, the activity of the motor cortex during the right-wrist movement was evaluated in 10 normal volunteers under active, passive, and imagery conditions. T2* weighted, three-dimensional functional images were acquired using a BOLD sensitive gradient-echo EPI (echo planar imaging) sequence with echo time (TE) of 30 ms and repetition time (TR) of 2000 ms. The functional data, which included 248 volumes per subject and condition, were acquired using the blocked design paradigm. The images were analyzed by the SPM12 toolbox, MATLAB software. RESULTS: The findings determined a significant increase in signal intensity of the motor cortex while performing the test compared to the rest time (p< 0.05). It was also observed that the active areas in hand representation of the motor cortex are different in terms of locations and the number of voxels in different wrist directions. Moreover, the findings showed that the position of active centers in the brain is different in active, passive, and imagery conditions. CONCLUSION: Results confirm that primary motor cortex neurons play an essential role in the processing of complex information and are designed to control the direction of movement. It seems that the findings of this study can be applied for rehabilitation studies.

Automated measurement of iris surface smoothness using anterior segment optical coherence tomography.

Fuchs uveitis (FU) is a chronic and often unilateral ocular inflammation and characteristic iris atrophic changes, other than heterochromia, are common in FU and are key to the correct diagnosis in many cases. With the advent of anterior segment optical coherence tomography (AS-OCT), some investigators attempted to quantitatively study these atrophic changes; mostly by introducing various methods to measure iris thickness in AS-OCT images. We aimed to present an automated method in an observational case series to measure the smoothness index (SI) of the iris surface in AS-OCT images. The ratio of the length of the straight line connecting the most peripheral and central points of the anterior iris border (in nasal and temporal sides) to the actual length of this border on AS-OCT images, was defined as SI. In a uveitis referral center, twenty-two eyes of 11 patients with unilateral Fuchs uveitis (FU) (7 female) and 22 eyes of 11 healthy control subjects underwent AS-OCT imaging. Image J and a newly developed MATLAB algorithm were used for manual and automated SI measurements, respectively. Agreement between manual and automated measurements was evaluated with Bland-Altman analysis and interclass correlation coefficient. The inter-eye difference of SI was compared between the FU group and the control group. Automated mean overall SI was 0.868 ± 0.037 and 0.840 ± 0.039 in FU and healthy fellow eyes, respectively (estimated mean difference = - 0.028, 95% CI [- 0.038, - 0.018], p < 0.001). Bland- Altman plots showed good agreement between two methods in both healthy and FU eyes. The interclass correlation coefficient between the manual and automated measurements in the FU and healthy fellow eyes was 0.958 and 0.964, respectively. The inter-eye difference of overall SI was 0.029 ± 0.015 and 0.012 ± 0.008 in FU group and control group, respectively (p = 0.01). We concluded that the automated algorithm can rapidly and conveniently measure SI with results comparable to the manual method.

Ultrasound elastography using shear wave interference patterns: a finite element study of affecting factors.

Elastography as one of the non-invasive medical imaging techniques which can help determine the stiffness of organs and other structures is currently attracting more attention. An interesting imaging rate-independent technique which has been discussed in literature uses shear wave interference patterns (SWIP). In this method, two external continuous harmonic vibration sources were used to induced SWIP and the resulting tissue displacements are mapped using ultrasonic imaging called sonoelastography. In this paper, a finite element model (FEM) of viscoelastic soft tissue with circular stiffer lesion inside, is simulated for testing the effect of stimulation characteristics on the propagation of SWIPs and shear speed map reconstruction. Also, we proposed an elastography probe, including miniature vibration sources and ultrasound transducer, which can be appropriate for experimental tests. The elastographic average speed ratio (ASR) and some scores like Dice coefficient, related to the binary image of shear speed map, are calculated for quantitatively measuring the effect of different contributing harmonic vibration parameters. Results show that the potential of providing useful diagnostic information can be improved if the preferable parameters are considered for implementation. According to these results the ASR, Dice and Jaccard scores would diverge from the ground truth of FEA if the parameter level is not selected correctly. Particularly, the Dice and Jaccard coefficients are obtained about 0.9 and 0.8, respectively, for the best vibration parameters level choice.

Quantifying cortical bone free water using short echo time (STE-MRI) at 1.5 T.

PURPOSE: The purpose of our study was to use Dual-TR STE-MR protocol as a clinical tool for cortical bone free water quantification at 1.5 T and validate it by comparing the obtained results (MR-derived results) with dehydration results. METHODS: Human studies were compliant with HIPPA and were approved by the institutional review board. Short Echo Time (STE) MR imaging with different Repetition Times (TRs) was used for quantification of cortical bone free water T1 (T1free) and concentration (ρfree). The proposed strategy was compared with the dehydration technique in seven bovine cortical bone samples. The agreement between the two methods was quantified by using Bland and Altman analysis. Then we applied the technique on a cross-sectional population of thirty healthy volunteers (18F/12M) and examined the association of the biomarkers with age. RESULTS: The mean values of ρfree for bovine cortical bone specimens were quantified as 4.37% and 5.34% by using STE-MR and dehydration techniques, respectively. The Bland and Altman analysis showed good agreement between the two methods along with the suggestion of 0.99% bias between them. Strong correlations were also reported between ρfree (r2 = 0.62) and T1free and age (r2 = 0.8). The reproducibility of the method, evaluated in eight subjects, yielded an intra-class correlation of 0.95. CONCLUSION: STE-MR imaging with dual-TR strategy is a clinical solution for quantifying cortical bone ρfree and T1free.

Segmentation of COVID-19 pneumonia lesions: A deep learning approach.

Background: Lung CT scan has a pivotal role in diagnosis and monitoring of COVID-19 patients, and with growing number of affected individuals, the need for artificial intelligence (AI)-based systems for interpretation of CT images is emerging. In current investigation we introduce a new deep learning-based automatic segmentation model for localization of COVID-19 pulmonary lesions. Methods: A total of 2469 CT scan slices, containing 1402 manually segmented abnormal and 1067 normal slices form 55 COVID-19 patients and 41 healthy individuals, were used to train a deep convolutional neural network (CNN) model based on Detectron2, an open-source modular object detection library. A dataset, including 1224 CT slices of 18 COVID-19 patients and 9 healthy individuals, was used to test the model. Results: The accuracy, sensitivity, and specificity of the trained model in marking a single image slice with COVID-19 lesion were 0.954, 0.928, and 0.961, respectively. Considering a threshold of 0.4% for percentage of lung involvement, the model was capable of diagnosing the patients with COVID-19 pneumonia, with a sensitivity of 0.982% and a specificity of 88.5%. Furthermore, the mean Intersection over Union (IoU) index for the test dataset was 0.865. Conclusion: The deep learning-based automatic segmentation method provides an acceptable accuracy in delineation and localization of COVID-19 lesions, assisting the clinicians and researchers for quantification of abnormal findings in chest CT scans. Moreover, instance segmentation is capable of monitoring longitudinal changes of the lesions, which could be beneficial to patients' follow-up.

Semi-automatic measurement of rigid gas-permeable contact lens movement in keratoconus patients using blinking images.

PURPOSE: To introduce a method for estimation of the rigid gas-permeable contact lens (RGP) movement. MATERIALS AND METHODS: Videos captured from normal blinking of keratoconus patients while wearing RGP lenses were used for this study. The videos are recorded using the CCD camera of a smart phone attached to the eyepiece of the slit lamp. The algorithm starts with extracting two frames of the video related to the highest and lowest positions of the lens during blinking, followed by an appropriate edge detection method. In the next step circular Hough transform is used to find the center of lens and to segment it in each image. Finally the lens movement is estimated by measuring vertical displacement of the lens center between these two frames. RESULTS: Mean and standard deviation of the difference between real movement and results of the algorithm for 20 cases are -8.66% and 10.71% respectively. The results are highly correlated with Pearson coefficient 0.986 P < 0.001. Bland-Altman plot with 95% levels of agreement (LoA) shows an agreement between exact manual measurement method and the proposed algorithm. CONCLUSION: The proposed algorithm shows a relatively high accuracy as the first attempt and compared to the routine qualitative visual estimation. Considering the importance of the lens movement, although this system was not tested on a series of RGP fitting patients yet, semi-automatic measurement may potentially help practitioners decide the appropriate RGP lens fit and reduce the fitting time.

Weighting Criteria and Prioritizing of Heat stress indices in surface mining using a Delphi Technique and Fuzzy AHP-TOPSIS Method.

BACKGROUND: Heat stress as a physical harmful agent can increase the risk of health and safety problems in different workplaces such as mining. Although there are different indices to assess the heat stress imposed on workers, choosing the best index for a specific workplace is so important. Since various criteria affect an index applicability, extracting the most effective ones and determining their weights help to prioritize the existing indices and select the optimal index. METHODS: In order to achieve this aim, present study compared some heat stress indices using effective methods. The viewpoints of occupational health experts and the qualitative Delphi methods were used to extract the most important criteria. Then, the weights of 11 selected criteria were determined by Fuzzy Analytic Hierarchy Process. Finally, fuzzy TOPSIS technique was applied for choosing the most suitable heat stress index. RESULTS: According to result, simplicity, reliability, being low cost, and comprehensiveness were the most determinative criteria for a heat stress index. Based on these criteria and their weights, the existing indices were prioritized. Eventually, wet bulb glob temperature appropriated the first priority and it was proposed as an applicable index for evaluating the heat stress at outdoor hot environments such as surface mines. CONCLUSIONS: The use of these strong methods allows introducing the most simple, precise, and applicable tool for evaluation the heat stress in hot environments. It seems that WBGT acts as an appropriate index for assessing the heat stress in mining activities at outdoors.

Modeling of diffuse reflectance of light in heterogeneous biological tissue to analysis of the effects of multiple scattering on reflectance pulse oximetry.

Fingertip-type pulse oximeters are popular, but their inconvenience for long-term monitoring in daily life means that other types of wearable pulse oximeters, such as reflectance pulse oximeters, need to be developed. For the purpose of developing reflection pulse oximetry, we have analyzed the light propagation in tissue to calculate and estimate the measured intensities of reflected light using the analytical and numerical solutions of the diffusion approximation equation. The reflectance of light from the biological tissue is investigated from theoretical and experimental perspectives, for light in the visible and near-infrared wavelengths. To establish the model, the calculated curves were compared with the analytical solution (AS) of the diffusion approximation equation in biological tissue. The results validated that the diffusion approximation equation could resolve the heterogeneous advanced tissue and the finite element method (FEM) could offer the simulation with higher efficiency and accuracy. Our aim has been to demonstrate the power of the FEM and AS in modeling of the steady-state diffusion approximation in a heterogeneous medium. Also, experimental data and the Monte Carlo model as a gold standard were used to verify the effectiveness of these methods.

Evaluation of RGP Contact Lens Fitting in Keratoconus Patients Using Hierarchical Fuzzy Model and Genetic Algorithms.

Keratoconus is a progressive non-inflammatory disease of the cornea. Rigid gas permeable contact lenses (RGPs) are prescribed when the disease progresses. Contact lens fitting and assessment is very difficult in these patients and is a concern of ophthalmologists and optometrists. In this study, a hierarchical fuzzy system is used to capture the expertise of experienced ophthalmologists during the lens evaluation phase of prescription. The system is fine-tuned using genetic algorithms. Sensitivity, specificity and accuracy of the final system are 88.9%, 94.4% and 92.6% respectively.

Reduced Imaging Rate in Liver Elastometery Using Shear Wave Interference Patterns.

Inducing interference patterns of shear wave is one of the proposed methods for reducing the frame rate in measuring wave speed during tissue elastography. Previously, the Nyquist rate must be met in order to provide an appropriate image for extracting the patterns with a reasonable accuracy. In this article we propose a technique based on image registration, and apply it to ultrasound images acquired before and after inducing the shear waves to estimate the amplitude of displacement. The displacement of the tissue is then used to form the interference pattern of shear waves. The method does not induce any restrictions on the time interval between images, so the tissue elasticity can be calculated independent of the imaging rate. The average error in measuring the elasticity of the simulated phantom is 13.7%.

A method for semi-automatic nuchal translucency thickness measurement.

Increased nuchal translucency (NT) thickness is associated with major abnormalities of the heart, great arteries and a wide range of genetic syndromes. NT thickness is one of the most important ultrasound markers in fetal diseases screening. To find an applicable method that accurately detects NT, we use an algorithm based on modified level set for semi-automated segmentation of the region. A regression model is used to find the appropriate direction of NT thickness measurement. The results show 4.5% error.

A novel approach in water quality assessment based on fuzzy logic.

The present work aimed at developing a novel water quality index based on fuzzy logic, that is, a comprehensive artificial intelligence (AI) approach to the development of environmental indices for routine assessment of surface water quality, particularly for human drinking purposes. Twenty parameters were included based on their critical importance for the overall water quality and their potential impact on human health. To assess the performance of the proposed index under actual conditions, a case study was conducted at Mamloo dam, Iran, employing water quality data of four sampling stations in the water basin of the dam from 2006 to 2009. Results of this study indicated that the general quality of water in all the sampling stations over all the years of the study period is fairly low (yearly averages are usually in the range of 45-55). According to the results of ANOVA test, water quality did not significantly change over time in any of the sampling stations (P > 0.05). In addition, comparison of the outputs of the fuzzy-based proposed index proposed with those of the NSF water quality index (the WQI) and Canadian Water Quality Index (CWQI) showed similar results and were sensitive to changes in the level of water quality parameters. However, the index proposed by the present study produced a more stringent outputs compared to the WQI and CWQI. Results of the sensitivity analysis suggested that the index is robust against the changes in the rules. In conclusion, the proposed index seems to produce accurate and reliable results and can therefore be used as a comprehensive tool for water quality assessment, especially for the analysis of human drinking water.

Abnormal red blood cells detection using adaptive neuro-fuzzy system.

Features like size, shape, and volume of red blood cells are important factors in diagnosing related blood disorders such as iron deficiency and anemia. This paper proposes a method to detect abnormality in red blood cells using cell microscopic images. Adaptive local thresholding and bounding box methods are used to extract inner and outer diameters of red cells. An adaptive network-based fuzzy inference system (ANFIS) is used to classify blood samples to normal and abnormal. Accuracy of the proposed method and area under ROC curve are 96.6% and 0.9950 respectively.

Classification of wheeze sounds using cepstral analysis and neural networks.

Wheezes are abnormal, continuous sounds heard over large airways and chest. They are divided to two groups based on relative intensity of airway obstruction. They are usually heard in asthma, pneumonia, emphysema and chronic obstructive pulmonary diseases (COPD). We present a classification method to discriminate between polyphonic and monophonic wheeze sounds using multilayer perceptron (MLP) neural network and mel-frequency cepstral coefficients (MFCC). Wheeze signals are divided to segments with 50% overlap. MFCC features are then extracted. Groups with different numbers of MFCC powerful features are compared by receiver operating characteristic (ROC) curves. The test results show an accuracy of 92.8%.