Wyniki wyszukiwania - BazTech

1

Efficient and robust medical image watermarking based on optimal subband tree structuring and discrete fractional fourier transform

Chacko Anusha, Chacko Shanty

Computer Assisted Methods in Engineering and Science

|

2023

|

Vol. 30, no. 4

557--581

EN

In order to solve the security problems associated with medical information and improve the robustness of watermarking algorithms for medical images, a unique approach to watermarking based on block operations is presented. This study considers the medical images as the cover image, with the watermark logo considered secret information that needs to be protected over the wireless transmission in telemedicine. In the embedding phase, input with the discrete fractional Fourier transform is first applied to the input, and then level 2 wavelet decomposition is carried out to determine the optimal sub-band tree. For each tree node on level 2, the approximated and detailed coefficient is determined through the feature analysis perspective. The novelty of the adopted methodology is its simplified transformation and embedding process. Upon receiving a complex matrix, it separates the real part from imaginary part where block transformation is carried out for embedding the watermark pixels. In the extraction phase, just a reverse operation is performed. The watermarking evaluation is performed by simulating various image processing attacks on watermarked medical images. The simulation outcome demonstrates the effectiveness of that proposed watermarking scheme against various attacks. The proposed watermarking technique is robust under various attacks based on image statistics such as PSNR, BER, and the correlation coefficient.

2

Transfer learning techniques for medical image analysis: A review

Kora Padmavathi, Ooi Chui Ping, Faust Oliver, Raghavendra U., Gudigar Anjan, Chan Wai Yee, Meenakshi K., Swaraja K., Plawiak Pawel, Acharya U. Rajendra

Biocybernetics and Biomedical Engineering

|

2022

|

Vol. 42, no. 1

79--107

EN

Medical imaging is a useful tool for disease detection and diagnostic imaging technology has enabled early diagnosis of medical conditions. Manual image analysis methods are labor-intense and they are susceptible to intra as well as inter-observer variability. Automated medical image analysis techniques can overcome these limitations. In this review, we investigated Transfer Learning (TL) architectures for automated medical image analysis. We discovered that TL has been applied to a wide range of medical imaging tasks, such as segmentation, object identification, disease categorization, severity grading, to name a few. We could establish that TL provides high quality decision support and requires less training data when compared to traditional deep learning methods. These advantageous properties arise from the fact that TL models have already been trained on large generic datasets and a task specific dataset is only used to customize the model. This eliminates the need to train the models from scratch. Our review shows that AlexNet, ResNet, VGGNet, and GoogleNet are the most widely used TL models for medical image analysis. We found that these models can understand medical images, and the customization refines the ability, making these TL models useful tools for medical image analysis.

3

SVIA dataset: A new dataset of microscopic videos and images for computer-aided sperm analysis

Chen Ao, Li Chen, Zou Shuojia, Rahaman Md Mamunur, Yao Yudong, Chen Haoyuan, Yang Hechen, Zhao Peng, Hu Weiming, Liu Wanli, Grzegorzek Marcin

Biocybernetics and Biomedical Engineering

|

2022

|

Vol. 42, no. 1

204--214

EN

Computer-Aided Sperm Analysis (CASA) is a widely studied topic in the diagnosis and treatment of male reproductive health. Although CASA has been evolving, there is still a lack of publicly available large-scale image datasets for CASA. To fill this gap, we provide the Sperm Videos and Images Analysis (SVIA) dataset, including three different subsets, subset-A, subset-B and subset-C, to test and evaluate different computer vision techniques in CASA. For subset-A, in order to test and evaluate the effectiveness of SVIA dataset for object detection, we use five representative object detection models and four commonly used evaluation metrics. For subset-B, in order to test and evaluate the effectiveness of SVIA dataset for image segmentation, we used eight representative methods and three standard evaluation metrics. Moreover, to test and evaluate the effectiveness of SVIA dataset for object tracking, we have employed the traditional kNN with progressive sperm (PR) as an evaluation metric and two deep learning models with three standard evaluation metrics. For subset-C, to prove the effectiveness of SVIA dataset for image denoising, nine denoising filters are used to denoise thirteen kinds of noise, and the mean structural similarity is calculated for evaluation. At the same time, to test and evaluate the effectiveness of SVIA dataset for image classification, we evaluate the results of twelve convolutional neural network models and six visual transformer models using four commonly used evaluation metrics. Through a series of experimental analyses and comparisons in this paper, it can be concluded that this proposed dataset can evaluate not only the functions of object detection, image segmentation, object tracking, image denoising, and image classification but also the robustness of object detection and image classification models. Therefore, SVIA dataset can fill the gap of the lack of large-scale public datasets in CASA and promote the development of CASA. Dataset is available at: .https://github.com/Demozsj/Detection-Sperm.

4

Review on approaches to concept detection in medical images

Miranda Diana, Thenkanidiyoor Veena, Dinesh Dileep Aroor

Biocybernetics and Biomedical Engineering

|

2022

|

Vol. 42, no. 2

453--462

EN

Concept detection in medical images involves the identification of various biomedical semantic entities in the images. This is a non-trivial task due to the highly heterogeneous nature of medical images. This heterogeneity is caused due to the different body parts, presence of abnormalities, and imaging techniques used to capture the image. In this paper, a thorough survey on important approaches to concept detection in medical images is presented. Methods such as multi-label classification, sequence-to-sequence learning, detecting concepts from captions, and similarity search-based approaches to concept detection in medical images are reviewed. This paper also highlights the challenges associated with the medical image concept detection task. Possible avenues for further research are also discussed.

5

Ortogonalna realizacja potokowego filtru Laplace'a do przetwarzania obrazów medycznych 2D

Poczekajło Paweł, Suszyński Robert

Przegląd Elektrotechniczny

|

2020

|

R. 96, nr 9

80--84

PL

W artykule przedstawiono dwie realizację filtru Gaussa 1D. Pierwsza oparta jest na bezpośredniej implementacji splotu, druga oparta została na filtrze ortogonalnym realizowanym za pomocą rotatorów Givensa. Obie realizacje został przeanalizowane pod kątem wrażliwości na kwantowanie współczynników dla 8-, 16- i 24-bitowych rejestrów. Wyznaczono i porównano błędy średniokwadratowe charakterystyki amplitudowej oraz błędy dla odpowiedzi systemu na pobudzenie losowym szumem i deltą Kroneckera.

EN

In the paper, two realizations of 1D Gauss filter are presented. The first realization is based on direct structure with convolution, in the second orthogonal filter with use Givens rotations is realized. Both systems are analyzed of sensitivity on coefficient quantization for 8-, 16- and 24- bits length of register. Also determined mean squared errors for amplitude characteristices, impulse responses and responses on noise excitation.

6

A new computer-based approach for fully automated segmentation of knee meniscus from magnetic resonance images

Saygılı A., Albayrak S.

Biocybernetics and Biomedical Engineering

|

2017

|

Vol. 37, no. 3

432--442

EN

Menisci are tissues that enable mobility and absorb excess loads on the knee. Problems in meniscus can trigger the disorder of osteoarthritis (OA). OA is one of the most common causes of disability, especially among young athlethes and elderly people. Therefore, the early diagnosis and treatment of abnormalities that occur in the meniscus are of significant importance. This study proposes a new computer-based and fully automated approach to support radiologists by: (i) the segmentation of medial menisci, (ii) enabling early diagnosis and treatment, and (iii) reducing the errors caused by MR intra-reader variability. In this study, 88 different MR images provided by the Osteoarthritis Initiative (OAI) are used. The histogram of oriented gradients (HOG) and local binary patterns (LBP) methods are used for feature extraction from these MR images along with the extreme learning machine (ELM) and random forests (RF) methods which are used for model learning (regression). As the first step of the pipeline, the most compact rectangular patches bounding the menisci are located. After this, meniscus boundaries are revealed by the morphological processes. Then, the similarities between these boundaries and the ground truth images are measured and compared with each other. The highest score is acquired with Dice similarity measurement with a success rate of 82%. A successful segmentation is performed on the diseased knee MR images. The proposed approach can be implemented as a decision support system for radiologists, while the segmented menisci can be used in classification of meniscal tear in future studies.

7

Zastosowanie sieci komórkowych do wykrywania konturów w obrazach medycznych

Chen Z.

Przegląd Elektrotechniczny

|

2009

|

R. 85, nr 4

56-58

PL

Artykuł przedstawia teoretyczne i praktyczne aspekty sieci komórkowych w zastosowaniu do wykrywania konturów w obrazach medycznych. Na podstawie obrazu wycinka skóry zwierzęcej przedstawiono i przedyskutowano różne aspekty metodologii obliczeń naturalnych przy użyciu sieci komórkowych.

EN

There is a strong need for integrated study of both theoretical and practical aspects of cellular automata (CA). In this article we report our ongoing work of exploring the nature of CA through an experimental study, focusing on an application of CA for edge detection in medical images. Through an example we describe the algorithm developed, followed by a discussion. The main theme of this paper is to advocate a thorough analysis of the methodology in the context of natural computing. We also point out the indication of such integrated study to education.

8

Decomposition of medical image based on grade multivariate methodology

Grzegorek M.

Biocybernetics and Biomedical Engineering

|

2007

|

Vol. 27, no. 4

29-48

EN

The paper presents disjoint decomposition of the set of pixels of a NMR image and also of its fragment suggested as interesting by a medical consultant. Each pixel is described by the value of gray level gl, gradient module gm and items constructed on the basis of gm and gradient modules of adjacent pixels. The obtained dataset with rows corresponding to pixels and columns corresponding to variables is then processed by the algorithm called GCCA (Grade Correspondence Cluster Analysis). This rearranges the initial ordering of pixels (and also of variables) and then divides the set of rows into a chosen number of clusters. Pixels in each cluster are visualized as a separate subimage. The resulting decomposition (an ordered sequence of sub images) depends on the choice of a threshold parameter b which strongly influences the comparison of gm with gradient modules in pixel's neighborhood. It is shown how b should be selected to specify the edge of the lateral ventricle and to investigate homogeneity of gm's neighborhoods in the area indicated for the consultant.

9

Fuzzy algorithm for madical image processing

Śrutek M., Śrutek E.

Image Processing & Communications

|

2006

|

Vol. 11, no 1

115-122

EN

Fuzzy rules were used to enhance images in the pre-processing step. The contrast enhancement and filtration methods are proposed to eliminate noise and artifacts from images. Implementation of fuzzy rules resulted in considerable image quality improvement. Next, the segmentation of regions corresponding to abnormalities is performed and geometrical and statistical parameters of the detected structures are calculated. Final classification algorithm is based on fuzzy rules.

10

Comparative investigations of similarity measure exploited in medical images preselection

Wróbel K., Porwik P.

Journal of Medical Informatics & Technologies

|

2004

|

Vol. 8

II25--31

EN

Nowadays, the most significant impact of digital image processing in the area of applications are real–world problems. Many new technological trends in medicine and digital processing have been implemented. Several factors indicate such development. A major one is the perpetually declining cost of the computer equipment required. Both processing unit and capacity of storage devices continue to become less expensive year by year. Another factor is the increasing availability of equipment for digitising and displaying images. In modern image processing, images have to be compared each other because such approach allows us to automate of retrieval process. Computer image retrieving is today especially important in medical diagnostics [1,7] or in preliminary images selection [8,9]. Today, in the digital image processing are used techniques and methods which have well known mathematical backgrounds. It can be observed, that in the area of digital signal processing, the Hough and well known the Fourier transform are exploited very often. These transforms are frequently use in image retrieving and can be implemented as computer applications. In many cases the mentioned methods give promising results in images classification or preselection [1,2,4,5,11]. Special properties of such transforms can be used in statistical or comparative goals, especially when searched information has graphic form. Taking into account the mentioned applications, transforms as methods of preliminary medical images selection have been investigated. From this reason pictures, analysing in the paper, to medical images have been limited.