Current Medical Imaging - Volume 17, Issue 1, 2021

Background: Colon cancer generally begins as a neoplastic growth of tissue, called polyps, originating from the inner lining of the colon wall. Most colon polyps are considered harmless but over the time, they can evolve into colon cancer, which, when diagnosed in later stages, is often fatal. Hence, time is of the essence in the early detection of polyps and the prevention of colon cancer. Methods: To aid this endeavor, many computer-aided methods have been developed, which use a wide array of techniques to detect, localize and segment polyps from CT Colonography images. In this paper, a comprehensive state-of-the-art method is proposed and categorize this work broadly using the available classification techniques using Machine Learning and Deep Learning. Conclusion: The performance of each of the proposed approach is analyzed with existing methods and also how they can be used to tackle the timely and accurate detection of colon polyps.

Due to the serious consequences of lung cancer, medical associations use computer-aided diagnostic procedures to diagnose this disease more accurately. Despite the damaging effects of lung cancer on the body, the lifetime of cancer patients can be extended by early diagnosis. Data mining techniques are practical in diagnosing lung cancer in its first stages. This paper surveys a number of leading data mining-based cancer diagnosis approaches. Moreover, this review draws a comparison between data mining approaches in terms of selection criteria and presents the advantages and disadvantages of each method.

Background: The videos produced during wireless capsule endoscopy have larger data size causing difficulty in transmission with limited bandwidth. The constraint on wireless capsule endoscopy hinders the performance of the compression module. Objectives: The objectives of this paper are as follows: (i) to conduct an extensive review of the lossless compression techniques and (ii) to find out the limitations of the existing system and the possibilities for improvement. Methods: The literature review was conducted with a focus on the compression schemes satisfying minimum computational complexity, less power dissipation and low memory requirements for hardware implementation. A thorough study of various lossless compression techniques was conducted under two perspectives, i.e., techniques applied to Bayer CFA and RGB images. The detail of the various stages of wireless capsule endoscopy compression was investigated to have a better understanding. The suitable performance metrics for evaluating the compression techniques were listed from various literature studies. Results: In addition to the Gastrolab database, WEO clinical endoscopy atlas and Gastrointestinal atlas were found to be better alternatives for experimentation. Pre-processing operations, especially new subsampling patterns need to be given more focus to exploit the redundancies in the images. Investigations showed that encoder module can be modified to bring more improvement towards compression. The real-time endoscopy still exists as a promising area for exploration. Conclusion: This review presents a research update on the details of wireless capsule endoscopy compression together with the findings as an eye-opener and guidance for further research.

Background: Immunization is a significant public health intervention to reduce child mortality and morbidity. However, its coverage, in spite of free accessibility, is still very low in developing countries. One of the primary reasons for this low coverage is the lack of analysis and proper utilization of immunization data at various healthcare facilities. Purpose: In this paper, the existing machine learning-based data analytics techniques have been reviewed critically to highlight the gaps where this high potential data could be exploited in a meaningful manner. Results: It has been revealed from our review that the existing approaches use data analytics techniques without considering the complete complexity of Expanded Program on Immunization which includes the maintenance of cold chain systems, proper distribution of vaccine and quality of data captured at various healthcare facilities. Moreover, in developing countries, there is no centralized data repository where all data related to immunization is being gathered to perform analytics at various levels of granularities. Conclusion: We believe that the existing non-centralized immunization data with the right set of machine learning and Artificial Intelligence-based techniques will not only improve the vaccination coverage but will also help in predicting the future trends and patterns of its coverage in different geographical locations.

Background: Detection of brain tumor is a complicated task, which requires specialized skills and interpretation techniques. Accurate brain tumor classification and segmentation from MR images provide an essential choice for medical treatments. Different objects within an MR image have similar size, shape, and density, which makes the tumor classification and segmentation even more complex. Objective: Classification of the brain MR images into tumorous and non-tumorous using deep features and different classifiers to get higher accuracy. Methods: In this study, a novel four-step process is proposed; pre-processing for image enhancement and compression, feature extraction using convolutional neural networks (CNN), classification using the multilayer perceptron and finally, tumor segmentation using enhanced fuzzy cmeans method. Results: The system is tested on 65 cases in four modalities consisting of 40,300 MR Images obtained from the BRATS-2015 dataset. These include images of 26 Low-Grade Glioma (LGG) tumor cases and 39 High-Grade Glioma (HGG) tumor cases. The proposed CNN feature-based classification technique outperforms the existing methods by achieving an average accuracy of 98.77% and a noticeable improvement in the segmentation results are measured. Conclusion: The proposed method for brain MR image classification to detect Glioma Tumor detection can be adopted as it gives better results with high accuracies.

Background: The brain is the most complex organ of the human body with millions of connections and activations. The electromagnetic signals are generated inside the brain due to a mental or physical task performed. These signals excite a bunch of neurons within a particular lobe depending upon the nature of the task performed. To localize this activity, certain machine learning (ML) techniques in conjunction with a neuroimaging technique (M/EEG, fMRI, PET) are developed. Different ML techniques are provided in the literature for brain source localization. Among them, the most common are: minimum norm estimation (MNE), low resolution brain electromagnetic tomography (LORETA) and Bayesian framework based multiple sparse priors (MSP). Aims: In this research work, EEG is used as a neuroimaging technique. Methods: EEG data is synthetically generated at SNR=5dB. Afterwards, ML techniques are applied to estimate the active sources. Each dataset is run for multiple trials (>40). The performance is analyzed using free energy and localization error as performance indicators. Furthermore, MSP is applied with a variant number of patches to observe the impact of patches on source localization. Results: It is observed that with an increased number of patches, the sources are localized with more precision and accuracy as expressed in terms of free energy and localization error, respectively. Conclusion: The patches optimization within the Bayesian Framework produces improved results in terms of free energy and localization error.

Background: Medical imaging encloses different imaging techniques and processes to image the human body for medical diagnostic and treatment purposes. Hence it plays an important role to improve public health. The technological development in biomedical imaging specifically in X-ray, Computed Tomography (CT), nuclear ultrasound including Positron Emission Tomography (PET), optical and Magnetic Resonance Imaging (MRI) can provide valuable information unique to a person. Objective: In health care applications, the images are needed to be exchanged mostly over a wireless medium. The diagnostic images with confidential information of a patient need to be protected from unauthorized access during transmission. In this paper, a novel encryption method is proposed to improve the security and integrity of medical images. Methods: Chaotic map along with DNA cryptography is used for encryption. The proposed method describes a two-phase encryption of medical images. Results: The performance of the proposed method is also tested by various analysis metrics. The robustness of the method against different noises and attacks is analyzed. Conclusion: The results show that the method is efficient and well suitable for medical images.

Background: BSI calculated from bone scintigraphy using ^99mtechnetium-methylene diphosphonate (^99mTc-MDP) is used as a quantitative indicator of metastatic bone involvement in bone metastasis diagnosis, therapeutic effect assessment, and prognosis prediction. However, the BONE NAVI, which calculates BSI, only supports bone scintigraphy using ^99mTc-MDP. Aims: We developed a method in collaboration with the Tokyo University of Agriculture and Technology to calculate bone scan index (BSI) employing deep learning algorithms with bone scintigraphy images using ^99mtechnetium-hydroxymethylene diphosphonate (^99mTc-HMDP). We used a convolutional neural network (CNN), enabling the simultaneous processing of anterior and posterior bone scintigraphy images named CNNapis. Objectives: The purpose of this study is to investigate the usefulness of the BSI calculated by CNNapis as bone imaging and bone metabolic biomarkers in patients with bone metastases from prostate cancer. Methods: At our hospital, 121 bone scintigraphy scans using ^99mTc-HMDP were performed and analyzed to examine bone metastases from prostate cancer, revealing the abnormal accumulation of radioisotope (RI) at bone metastasis sites. Blood tests for serum prostate-specific antigen (PSA) and alkaline phosphatase (ALP) were performed concurrently. BSI values calculated by CNNapis were used to quantify the metastatic bone tumor involvement. Correlations between BSI and PSA and between BSI and ALP were calculated. Subjects were divided into four groups by BSI values (Group 1, 0 to <1; Group 2, 1 to <3; Group 3, 3 to <10; Group 4, >10), and the PSA and ALP values in each group were statistically compared. Results: Patients diagnosed with bone metastases after bone scintigraphy were also diagnosed with bone metastases using CNNapis. BSI corresponding to the range of abnormal RI accumulation was calculated. PSA and BSI (r = 0.2791) and ALP and BSI (r = 0.6814) correlated positively. Significant intergroup differences in PSA between Groups 1 and 2, Groups 1 and 4, Groups 2 and 3, and Groups 3 and 4 and in ALP between Groups 1 and 4, Groups 2 and 4, and Groups 3 and 4 were found. Conclusion: BSI calculated using CNNapis correlated with ALP and PSA values and is useful as bone imaging and bone metabolic biomarkers, indicative of the activity and spread of bone metastases from prostate cancer.

Background: Breast cancer is a curable disease if diagnosed at an early stage. The chances of having breast cancer are the lowest in married women after the breast-feeding phase because the cancer is formed from the blocked milk ducts. Introduction: Nowadays, cancer is considered the leading cause of death globally. Breast cancer is the most common cancer among females. It is possible to develop breast cancer while breast-feeding a baby, but it is rare. Mammography is one of the most effective methods used in hospitals and clinics for early detection of breast cancer. Various researchers are used in artificial intelligence- based mammogram techniques. This process of mammography will reduce the death rate of the patients affected by breast cancer. This process is improved by the image analysing, detection, screening, diagnosing, and other performance measures. Methods: The radial basis neural network will be used for classification purposes. The radial basis neural network is designed with the help of the optimization algorithm. The optimization is to tune the classifier to reduce the error rate with the minimum time for the training process. The cuckoo search algorithm will be used for this purpose. Results: Thus, the proposed optimum RBNN is determined to classify breast cancer images. In this, the three sets of properties were classified by performing the feature extraction and feature reduction. In this breast cancer MRI image, the normal, benign, and malignant is taken to perform the classification. The minimum fitness value is determined to evaluate the optimum value of possible locations. The radial basis function is evaluated with the cuckoo search algorithm to optimize the feature reduction process. The proposed methodology is compared with the traditional radial basis neural network using the evaluation parameter like accuracy, precision, recall and f1-score. The whole system model is done by using Matrix Laboratory (MATLAB) with the adaptation of 2018a since the proposed system is most efficient than most recent related literature. Conclusion: Thus, it concluded with the efficient classification process of RBNN using a cuckoo search algorithm for breast cancer images. The mammogram images are taken into recent research because breast cancer is a major issue for women. This process is carried to classify the various features for three sets of properties. The optimized classifier improves performance and provides a better result. In this proposed research work, the input image is filtered using a wiener filter, and the classifier extracts the feature based on the breast image.

Background: Scanning a patient’s lungs to detect Coronavirus 2019 (COVID-19) may lead to similar imaging of other chest diseases. Thus, a multidisciplinary approach is strongly required to confirm the diagnosis. There are only a few works targeted at pathological x-ray images. Most of the works only target single disease detection which is not good enough. Some works have been provided for all classes. However, the results suffer due to lack of data for rare classes and data unbalancing problem. Methods: Due to the rise in COVID-19 cases, medical facilities in many countries are overwhelmed and there is a need for an intelligent system to detect it. Few works have been done regarding the detection of the coronavirus but there are many cases where it can be misclassified as some techniques are not efficient and can only identify specific diseases. This work is a deep learning- based model to distinguish COVID-19 cases from other chest diseases. Results: A Deep Neural Network model provides a significant contribution in terms of detecting COVID-19 and provides an effective analysis of chest-related diseases taking into account both age and gender. Our model achieves 87% accuracy in terms of GAN-based synthetic data and presents four different types of deep learning-based models that provide comparable results to other state-of-the-art techniques. Conclusion: The healthcare industry may face unfavorable consequences if the gap in the identification of all types of pneumonia is not filled with effective automation.

Background and Purpose: Recent studies have shown that diffusion tensor imaging (DTI) parameters are used to follow the patients with breast cancer and correlate well as a prognostic parameter of breast cancer. However, as far as we know, there is no data to compare the DTI features of breast cancer brain metastases according to molecular subtypes in the literature. Our aim is to evaluate whether there are any differences in DTI parameters of brain metastases in patients with breast cancer according to molecular subtypes. Methods: Twenty-seven patients with breast cancer and 82 metastatic brain lesions were included. We classified subjects into three subgroups according to their hormone expression; Group 0, triple- negative (n; 6, 19 lesions), group 1, HER2-positive (n;16, 54 lesions) and group 2, hormone-- positive group (n; 5, 9 lesions). The apparent diffusion coefficient (ADC), fractional anisotropy (FA), axial diffusivity (AD), and radial diffusivity (RD) values in DTI were measured and compared between three groups. Results: ADC, AD and RD values of group 2 were significantly lower compared to group 0. No significant differences were found in FA, ADC, AD and RD values between the group 0 and 1 and the group 1 and 2. Conclusion: Metastasis of aggressive triple-negative breast cancer showed higher ADC values compared to the less aggressive hormone-positive group. Higher ADC values in brain metastases of breast cancer may indicate a poor prognosis, so DTI findings could play a role in planning appropriate treatment.

Objectives: The aim of this study was to investigate the value of hepatic T2 imaging for the evaluation of chronic hepatitis-B-related acute-on-chronic liver failure (HBV-ACLF). Methods: Three groups of patients underwent liver MRI utilising m-GRASE sequence (multi-echo gradient and spin echo): HBV-ACLF patients (n = 28), chronic hepatitis B patients (n = 11), and healthy control patients (n = 14). A T2 image was produced using post-processing software, and the mean T2 (relaxation time) value was calculated. Blood biochemical indices for the HBV-ACLF and Chronic Hepatitis B were obtained within 2 days pre- or post-MR scanning. The patients’ T2 values, and the correlation between their biochemical indices and T2 values were analysed. A receiver operating characteristic curve was employed to evaluate the efficiency of utilising T2 values in the diagnosis of HBV-ACLF. Results: There were significant variations in the T2 values (χ² = 19.074, P < 0.001) among the 3 groups. The AUC of T2 values for diagnosing HBV-ACLF was 0.86 (P < 0.001), with a cut-off value of 57.73 ms. A moderately positive correlation was observed between the T2 value and the international normalised ratio, prothrombin time, and hyaluronic acid values (rs = 0.65, P < 0.001; rs = 0.67, P < 0.001; rs = 0.39, P = 0.025). A moderately negative correlation was observed between the T2 value and the prothrombin activity, albumin, and prealbumin values (rs = -0.67, P < 0.001; rs = -0.48, P = 0.004; rs = -0.37, P = 0.030). Conclusion: T2 values could accurately reflect liver function state, as they correlated well with certain biochemical indices, illustrating good diagnostic efficiency for diagnosing HBV-ACLF.

Background: Breast cancer is considered as one of the most perilous sickness among females worldwide and the ratio of new cases is increasing yearly. Many researchers have proposed efficient algorithms to diagnose breast cancer at early stages, which have increased the efficiency and performance by utilizing the learned features of gold standard histopathological images. Objective: Most of these systems have either used traditional handcrafted or deep features, which had a lot of noise and redundancy, and ultimately decrease the performance of the system. Methods: A hybrid approach is proposed by fusing and optimizing the properties of handcrafted and deep features to classify the breast cancer images. HOG and LBP features are serially fused with pre-trained models VGG19 and InceptionV3. PCR and ICR are used to evaluate the classification performance of the proposed method. Results: The method concentrates on histopathological images to classify the breast cancer. The performance is compared with the state-of-the-art techniques, where an overall patient-level accuracy of 97.2% and image-level accuracy of 96.7% is recorded. Conclusion: The proposed hybrid method achieves the best performance as compared to previous methods and it can be used for the intelligent healthcare systems and early breast cancer detection.

Objective: Seizure is the most common neurological disorders in children, where 4-10% of the cases experience at least one seizure before the age of 16. The most frequent causes of seizures in children are fever, epilepsy, infection and brain damage. The aim of this study was to investigate the frequency of organic lesions in MRI of children with seizures unrelated to fever. Materials and Methods: This cross-sectional study included children presented with fever-unrelated seizures. The MRI was examined by a radiologist to identify abnormal findings in each patient. A researcher-made questionnaire including general information, history of head trauma, obstructed labor and the history of seizure was completed for the patients. Results: Of 287 children with fever-related seizure, 127 (45.7%) were male and 151 (54.3%) were female. History of seizure, history of obstructed labor, abnormal MRI, complete delay, use of antiepileptic drug and history of trauma were 22(9.9%), 1 (0.4%), 11(4%), 5(1.8%), 259(93.2%) and 12 (4.3%), respectively. Of 11 patients with abnormal MRI, 4 had MTS lesions, 2 had tumor lesions, 2 had scarring trauma, 1 had an epidural abscess and 1 had meningitis. The frequency of organic lesions had no significant differences based on gender, use of antiepileptic drug and traumatic history, but it had a significant relation with obstructed labor andthehistory of seizure. Conclusion: The results showed that organic brain lesions in children with fever-unrelated seizure had a significant relationship with the history of seizure and obstructed maternal labor.