CVG-Net: novel transfer learning based deep features for diagnosis of brain tumors using MRI scans.

Brain tumors present a significant medical challenge, demanding accurate and timely diagnosis for effective treatment planning. These tumors disrupt normal brain functions in various ways, giving rise to a broad spectrum of physical, cognitive, and emotional challenges. The daily increase in mortality rates attributed to brain tumors underscores the urgency of this issue. In recent years, advanced medical imaging techniques, particularly magnetic resonance imaging (MRI), have emerged as indispensable tools for diagnosing brain tumors. Brain MRI scans provide high-resolution, non-invasive visualization of brain structures, facilitating the precise detection of abnormalities such as tumors. This study aims to propose an effective neural network approach for the timely diagnosis of brain tumors. Our experiments utilized a multi-class MRI image dataset comprising 21,672 images related to glioma tumors, meningioma tumors, and pituitary tumors. We introduced a novel neural network-based feature engineering approach, combining 2D convolutional neural network (2DCNN) and VGG16. The resulting 2DCNN-VGG16 network (CVG-Net) extracted spatial features from MRI images using 2DCNN and VGG16 without human intervention. The newly created hybrid feature set is then input into machine learning models to diagnose brain tumors. We have balanced the multi-class MRI image features data using the Synthetic Minority Over-sampling Technique (SMOTE) approach. Extensive research experiments demonstrate that utilizing the proposed CVG-Net, the k-neighbors classifier outperformed state-of-the-art studies with a k-fold accuracy performance score of 0.96. We also applied hyperparameter tuning to enhance performance for multi-class brain tumor diagnosis. Our novel proposed approach has the potential to revolutionize early brain tumor diagnosis, providing medical professionals with a cost-effective and timely diagnostic mechanism.

A New Scoring System of Oral Lichen Planus.

Lichen planus (LP) is a chronic, immune-mediated mucocutaneous disorder increasingly becoming common in the general population with female predominance. Clinically, there are different forms of lichen planus with the presence of the main characteristic feature of Wickham striae. Literature, to date, is abundant with various scoring systems of oral lichen planus, and among them, the most commonly followed scoring system was the one proposed by the Thongprasom system because of its simplicity and ease of application. Aim: The aim of the present study is to critically review all the disease scoring systems on oral lichen planus (OLP) that have been reported in the literature during the past decades. A systematic literature search was performed using PUBMED, MEDLINE, EMBASE, and COCHRANE Library with language restriction to English. The search was carried out incorporating the published literature from 1980 to 2020 using the MeSH (medical subject heading) terms. A literature search was done using keywords: Staging, Grading, Oral lichen planus, Diagnostic, and Therapeutic. Out of 25 publications, related to search strategy, 22 full articles, which were related to the disease scoring system for oral lichen planus, were acquired for further inspection. Out of the 22 articles, 15 articles met the inclusion criteria. The data was collected and a brief summary of the studies regarding the different disease scoring systems for oral lichen planus was explained. Taking into consideration, the parameters were not included in the previous disease scoring system. A new proposal encompassing a scoring system for oral lichen planus considering the missing parameters along with an amalgamation of histopathological criteria of dysplasia is presented. It also proposes to grade and stage the lesions and recommend appropriate therapy for each of such lesions.

Computer-aided analysis of radiological images for cancer diagnosis: performance analysis on benchmark datasets, challenges, and directions.

Radiological image analysis using machine learning has been extensively applied to enhance biopsy diagnosis accuracy and assist radiologists with precise cures. With improvements in the medical industry and its technology, computer-aided diagnosis (CAD) systems have been essential in detecting early cancer signs in patients that could not be observed physically, exclusive of introducing errors. CAD is a detection system that combines artificially intelligent techniques with image processing applications thru computer vision. Several manual procedures are reported in state of the art for cancer diagnosis. Still, they are costly, time-consuming and diagnose cancer in late stages such as CT scans, radiography, and MRI scan. In this research, numerous state-of-the-art approaches on multi-organs detection using clinical practices are evaluated, such as cancer, neurological, psychiatric, cardiovascular and abdominal imaging. Additionally, numerous sound approaches are clustered together and their results are assessed and compared on benchmark datasets. Standard metrics such as accuracy, sensitivity, specificity and false-positive rate are employed to check the validity of the current models reported in the literature. Finally, existing issues are highlighted and possible directions for future work are also suggested.

Identification of Severe Grading in Knee OsteoArthritis from MRI using Ensemble Deep Learning.

BACKGROUND: Accurate finding of Knee Osteoarthritis (KOA) from structural Magnetic Resonance Imaging (MRI) is a difficult task and is greatly subject to user variation. Furthermore, the identification of knee osteoarthritis (KOA) from MRI scans presents a challenge due to the limited information available. A novel methodology using an ensemble Deep Learning algorithm, combining EfficientNet-B3 and ResNext-101 architectures, aims to forecast KOA advancement, bridging the identified gap in clinical trials. OBJECTIVES: The study aims to develop a precise predictive model for knee osteoarthritis using advanced deep-learning architectures and structural MRI scan data. By utilizing an ensemble technique, the model's accuracy in predicting disease development is enhanced, surpassing the limitations of traditional biomarkers. METHODS: The study used the Osteoarthritis Initiative dataset to develop an ensemble Deep Learning model that combined EfficientNet-B3 and ResNext-101 architectures. Techniques like cropping, gamma correction, and in-slice rotation were used to expand the dataset and improve the model's generalization capacity. RESULTS: The Deep Learning model demonstrated 93% validation accuracy on the OAI dataset, accurately capturing subtle patterns of knee osteoarthritis progression. Augmentation approaches enhanced its resilience. CONCLUSION: Our ensemble Deep Learning approach, using ResNext-101 and EfficientNet-B3 architectures, accurately predicts knee osteoarthritis courses using structural MRI data, demonstrating the importance of data augmentation for improved predictive tools.

Feasibility of Point Shear Wave Elastography for Evaluating Renal Cortical Thickness: A Prospective Study.

BACKGROUND: Chronic Kidney Disease [CKD] affects individuals of different age groups worldwide. Moreover, CKD is associated with several risk factors, including obesity, lifestyle, and hypertension, which are common in the Middle East. Ultrasonography is the examination of choice for CKD. In recent years, Shear Wave Elastography [SWE] has developed through the continued development of ultrasound and received substantial attention ;therefore, it can be used to measure tissue stiffness. The study aimed to use point Shear Wave Elastography [p-SWE] to determine the correlation between diabetes and cortical renal thickness in detecting pathologies. METHODS: This study was performed at the King Abdul-Aziz University Hospital. We examined 61 patients who underwent SWE. The patients were classified into two groups based on the presence or absence of type 2 Diabetes Mellitus [DM]. RESULTS: The results showed that there was a significant correlation between cortical stiffness and DM duration [p<0.005]. In addition, there was a negative correlation between cortical stiffness and cortical thickness [p=0.147] in patients with DM. Moreover, the eGFR decreased with an increase in cortical stiffness [p=0.499]. The cortical thickness in patients with and without DM was 0.750 ± 0.2 kPa and 0.788 ± 0.4 kPa, respectively. The kidney stiffness in patients with DM and control patients was 8.5 ± 8.6 cm and 14.0 ± 25.16 cm, respectively. CONCLUSION: This study showed that kidney p-SWE measurements were reliable. Therefore, further studies assessing kidney stiffness in patients with and without people with diabetes are recommended.

Effectiveness of radiology modalities in diagnosing and characterizing brain disorders.

OBJECTIVES: To observe the accuracy of Computed Tomography (CT) and Magnetic Resonance Imaging (MRI) scans in evaluating neurological disorders. METHODS: This retrospective research used CT or MRI to diagnose and characterize brain disorders. Patients' records suffering from neurological disorders were considered eligible for inclusion, regardless of the time of appearance of symptoms, the severity of their symptoms, or their final clinical diagnosis. The exclusion criteria for this study involved patients who did not undergo either a CT or MRI scan. A chi-square test was performed to observe the association between the study variables. A total of 3155 cases were analyzed. RESULTS: The most prevalent comorbid was dyslipidemia 670 (21.6%) followed by hypertension 548 (17.6%). Overall brain disorders were confirmed in 2426 (77%) patients. It was observed that half of the patients 1543 (48.9%) were diagnosed with stroke. It was found that the accuracy of CT and MRI was 78% and 74% respectively. The association of modalities, patient type, and gender with the confirmation of diseases was not found significant (p=>0.05). CONCLUSION: Our study revealed that CT and MRI were accurate by more than 75% and no difference was between both techniques to detect neurological disorders.

The Role of MRI in Diagnosing Agenesis of the Corpus Callosum (ACC): A Rare Disorder Present at Birth.

BACKGROUND: Agenesis of the corpus callosum (ACC) is a rare hereditary nervous system defect present at birth. ACC is an uncommon condition that is unrepresentative in the general population because some cases do not present with any identifiable symptoms in the early stage. CASE REPORT: We present a case of ACC in a two-month-old male patient who was diagnosed after birth. Although the initial brain ultrasound (US) showed dilation of the lateral ventricles and the absence of the corpus callosum, these findings were not fully confirmed. Therefore, magnetic resonance imaging (MRI) of the brain was conducted to confirm the complex diagnosis, and the examination revealed complete ACC. Diagnosing ACC in a neonate demonstrates the complexity of diagnosis through the clinical presentation, especially at an early age. CONCLUSION: The clinical utility of neonatal US and MRI highlights the importance of an early diagnosis of ACC. MRI is more effective than the US in detecting this condition, and these imaging modalities provide the patient with an early diagnosis, which helps in treatment management.

Anthropomorphic Carotid Artery Phantom for Ultrasound Flow and Vessel Wall Elastography Imaging.

OBJECTIVE: Shear wave elastography imaging (SWE) is a non-invasive US technique that has been developed to provide quantitative information about tissue elasticity. This technique might be useful in the identification of vascular risk factors. Arterial wall thickness and inner diameter vary with age and disease, which may impact shear wave propagation. The effect of arterial geometry on SWE has not yet been thoroughly investigated. Therefore, this study aimed to investigate the impact of different wall thickness, pulsation and imaging planes on YM estimates, to gain more information about the source of variability associated with SWE. METHODS: Poly(vinyl alcohol) cryogel (PVA-c) fabrication has been used for phantom design and construction. The agar-based TMM was used to surround the tubes. The inlet and outlet of the phantom were connected to a programmable gear pump using c-flex tubing to form a closed loop. Image J profiling was used to clarify the anomalies further detected using SWE. RESULTS: The 4 F/T cycle vessel phantom has shown less YM variability than in the 6 F/T cycle. YM ranged from 8 kPa for a 1 mm thickness tube to 53 kPa for the thickest 6 mm wall thickness for the softer 4 F/T cycle tube. Vessel phantoms embedded in TMM show higher variability than vessel phantoms submerged in water. YM ranged from 32 kPa for a 1 mm thickness tube to 117 kPa for the thickest 6 mm wall thickness for the softer 4 F/T cycle tube. CONCLUSION: SWE variability in measurements was higher in phantoms embedded in TMM compared to those submerged in water. It is recommended that combine the transverse and longitudinal imaging planes to provide a better understanding of disease over the full vessel circumference.

Interobserver variability in ultrasound assessment of thyroid nodules.

The first diagnostic tool for thyroid disease management is ultrasound. Despite its importance, ultrasound is an extremely subjective procedure that requires a high level of performance skill. Few studies have assessed thyroid ultrasound performance and its effectiveness, particularly the variability between observers in the assessment of ultrasound images. This study evaluated the variability in ultrasound assessments and diagnoses of thyroid nodules between 2 radiologists. In this retrospective study, 75 thyroid nodules in 39 patients were reviewed by 2 experienced radiologists. The nodule composition, margin, shape, calcification, and vasculitis were determined using echogenicity. The study evaluation included these 5 assessments and the final diagnosis. Interobserver variation was determined using Cohen kappa statistics. The interobserver agreements in the interpretation of echogenicity, shape, and margin were fair (κ = 0.21-0.40), whereas there were substantial agreements for vascularity and calcification (κ = 0.62-0.78). The agreements between the observers for individual ultrasound features in this study were the highest for vascularity and the presence/absence of calcification. The interobserver reproducibility for thyroid nodule ultrasound reporting was adequate, but the diagnostic evaluation ability of the observers was inconsistent. The variability in the interpretation of sonographic features could influence the level of suspicion of thyroid malignancy. This study emphasizes the need for consistency in the training of sonographic interpretation of thyroid nodules, particularly for echogenicity, shape, and margin.

Radiographers' awareness level of MRI-induced vertigo and their perspectives on the post-examination care provided to patients in Saudi Arabia.

INTRODUCTION: Vertigo has been reported by operators and patients during magnetic resonance imaging (MRI) examinations and found to increase in severity as the strength of the scanner magnet increases. This study examined a cohort of MRI radiographers' awareness of MRI-induced vertigo and their perspectives on post-MRI care. METHODS: This cross-sectional study used a web-based survey distributed to 110 radiographers. The 18-item survey included questions to elicit demographic information, MRI radiographers' awareness of MRI-induced vertigo, and their perspectives on the post-MRI care that should be provided to patients. Responses were collected between June 2021 and January 2022. The collected data were analyzed using SPSS, version 27. RESULTS: A total of 110 MRI radiographers completed the survey. Participants were predominantly male (64.5 %) and working in public practice (91.8 %). Almost all the radiographers were aware of MRI-induced vertigo. About two-thirds of participants knew patients needed assistance off the couch. Nearly all participants knew patients should be asked about their experience with MRI-induced vertigo after their procedures. There were statistically significant associations between the size of magnetic field strength used by the participants and their appreciation of the needed support for patients post-MRI examinations (p= 0.012). CONCLUSION: This study provides the first insight into Saudi Arabian MRI radiographers' awareness and perceptions of MRI-induced vertigo. Radiographers were largely aware of MRI-induced vertigo and the supportive care they were supposed to provide their patients. IMPLICATIONS FOR PRACTICE: The current study points to a need for training to expand awareness levels of MRI-induced vertigo among a few Saudi MRI radiographers.

Cloud Computing-Based Framework for Breast Tumor Image Classification Using Fusion of AlexNet and GLCM Texture Features with Ensemble Multi-Kernel Support Vector Machine (MK-SVM).

Breast cancer is common among women all over the world. Early identification of breast cancer lowers death rates. However, it is difficult to determine whether these are cancerous or noncancerous lesions due to their inconsistencies in image appearance. Machine learning techniques are widely employed in imaging analysis as a diagnostic method for breast cancer classification. However, patients cannot take advantage of remote areas as these systems are unavailable on clouds. Thus, breast cancer detection for remote patients is indispensable, which can only be possible through cloud computing. The user is allowed to feed images into the cloud system, which is further investigated through the computer aided diagnosis (CAD) system. Such systems could also be used to track patients, older adults, especially with disabilities, particularly in remote areas of developing countries that do not have medical facilities and paramedic staff. In the proposed CAD system, a fusion of AlexNet architecture and GLCM (gray-level cooccurrence matrix) features are used to extract distinguishable texture features from breast tissues. Finally, to attain higher precision, an ensemble of MK-SVM is used. For testing purposes, the proposed model is applied to the MIAS dataset, a commonly used breast image database, and achieved 96.26% accuracy.

Investigating Confidence Levels of Saudi Clinical Supervisors in Radiology Departments.

Background: Effective teaching and supervision within hospitals play an essential role in training radiography students. However, inadequate preparation of teaching roles has been highlighted over the last three decades as a problem for many radiographers. This can lead to inadequate preparation and a lack of confidence in the supervisory role, which may affect the students' learning experience. Few studies in Saudi Arabia have investigated the skills and resources needed by radiographers to become effective and confident teachers. Therefore, this study aimed to explore the experiences and confidence of clinical radiographers in teaching radiography students and establish the areas of support they require to be more effective in their clinical teaching role. Methods: An online questionnaire and semi-structured interviews were used to collect data from radiographers working in Saudi Arabia's radiology departments. Radiographers who were involved in the supervision of students are included in the study. A total of 159 radiographers participated in the study. Results: The findings showed that radiographers were reasonably confident in four domains: introducing students and familiarizing them within the practice environment, supervision, facilitating students' learning, and assisting students to integrate into the practice environment while some areas required further development. The finding also indicated high number of students believed that providing an accurate perspective on the philosophy of the environment is not applicable to them. Conclusion: The article concludes with a recommendation for further support and guidance for radiographers in teaching roles from institutions. The study provided insights into the world of clinical supervisors in radiology departments. Informative feedback to students during their clinical training by clinical supervisors is a key strategy to fill the gap between theory and practice experienced by students. Additionally, the importance for implementation of ongoing professional development for radiographers is advised to ensure the quality of clinical placement for radiography students.

The Motivational Factors of Choosing Diagnostic Radiology as a Profession Among Saudi Radiography Students.

Purpose: To identify factors influencing Saudi students to choose radiography as their academic major and future career field. Material and Method: This study involved quantitative (online questionnaire) and qualitative (semi-structured interview) approaches. An online questionnaire was distributed among (n = 308) students. The questionnaire contained 30 questions covering the following three domains: economy, vocational and personal. A total of 25 individual semi-structured interviews were conducted with purposive sampling of radiography students in seven universities (public and private) across the western region of Saudi Arabia. Interview responses were coded, and main themes were extracted based on Miles and Huberman's framework. Results: The findings demonstrated that radiography was the first option as profession for 44% of the study participants. Several factors that study participants considered important were in the realm of patient care, helping patients, radiographer-patient relations, science-based profession, and the desire to work in the healthcare system. A few participants (14%) reported that they are planning to change their profession to another medical speciality. Four themes were identified from the interviews: 1) Profession decision-making, 2) changing career", 3) difficulties and challenges, and 4) recommended radiography as a profession. Conclusion: The results of the study support the need to bridge the gap between high school, universities, and employment through a collaborative network to assist students in exploring their career path by providing sufficient information and experience.

Automatic skin lesions detection from images through microscopic hybrid features set and machine learning classifiers.

Skin cancer occurrences increase exponentially worldwide due to the lack of awareness of significant populations and skin specialists. Medical imaging can help with early detection and more accurate diagnosis of skin cancer. The physicians usually follow the manual diagnosis method in their clinics but nonprofessional dermatologists sometimes affect the accuracy of the results. Thus, the automated system is required to assist physicians in diagnosing skin cancer at early stage precisely to decrease the mortality rate. This article presents an automatic skin lesions detection through a microscopic hybrid feature set and machine learning-based classification. The employment of deep features through AlexNet architecture with local optimal-oriented pattern can accurately predict skin lesions. The proposed model is tested on two open-access datasets PAD-UFES-20 and MED-NODE comprising melanoma and nevus images. Experimental results on both datasets exhibit the efficacy of hybrid features with the help of machine learning. Finally, the proposed model achieved 94.7% accuracy using an ensemble classifier. RESEARCH HIGHLIGHTS: The deep features accurately predicted skin lesions through AlexNet architecture with local optimal-oriented pattern. Proposed model is tested on two datasets PAD-UFES-20, MED-NODE comprising melanoma, nevus images and exhibited high accuracy.

Sensitivity and specificity of nuclear medicines (DTPA and DMSA) with magnetic resonance imaging in diagnosing bone metastasis.

BACKGROUND: The frequency of bone metastases in individuals increases at advanced stages of cancer, mostly in patients suffering from lung, breast, or prostate cancer. The study aims to evaluate the effectiveness of bone metastases diagnosis of nuclear medicine, CT scan, and MRI in detecting bone metastases among patients with lung, breast, and prostate carcinoma. MATERIAL AND METHODS: Retrospective study design was adopted for the analysis of 120 recruited patients (with the presence of bone metastasis) following a series of examinations and tests. RESULTS: Better sensitivity (73.33%) and specificity (94.66%) for MRI as compared to SPECT. MRI also proved to be more sensitive (68%) and specific (95.74%), as compared to the findings of the CT scan. CONCLUSIONS: The results conclude that MRI provided favorable diagnostic performance for bone metastasis. It emphasizes that diagnosis using MRI may enable practitioners to devise optimal carcinoma treatment strategies. The healthcare practitioners need to assess the MRI findings to determine improved treatment plans.

Arterial Stiffness Assessment in Healthy Participants Using Shear Wave Elastography.

BACKGROUND: Arterial stiffness is an important biomarker for cardiovascular disease. Shear wave elastography (SWE) provides quantitative estimates of tissue stiffness. OBJECTIVE: This study aimed to provide reference values for arterial wall, assessing the suitability of SWE to quantify elasticity of the common carotid artery (CCA) and evaluating inter-and intra-observer reproducibility. METHODS: A Supersonic Aixplorer ultrasound system with L15-4 probe was used to scan longitudinal sections of the CCA. Young's modulus (YM) was measured within 2-mm regions of interest. Reproducibility was assessed within a subgroup of 16 participants by two operators (one novice and one experienced) during two sessions >one week apart. RESULTS: This study involves seventy-three participants with a mean age of 40±10 years and a body mass index of 26 ±6 kg/m2. YM estimates were 59 kPa ±19 in men and 56 kPa ±12 in women. The average YM of the CCA walls was 58 kPa ±15 (57 ±15 kPa for the anterior wall and 58 ±20 kPa for the posterior wall, p=0.75). There was no significant difference in the mean of YM estimates of the CCA between the observers (observer: one 51 ±14 kPa and observer two: 55 ±17 kPa[p=0.46]). Interand intra-observer reproducibility was fair to good (Intra-class correlations, ranging from 0.46 to 0.71). Inter-frame variability was 28%. CONCLUSION: In healthy individuals, SWE provided an estimate of YM of the CCA (58 kPa) with fair to good reproducibility. This study demonstrated the potential of using SWE for assessing biomechanical properties of blood vessels.

Machine Learning with Quantum Seagull Optimization Model for COVID-19 Chest X-Ray Image Classification.

Early and accurate detection of COVID-19 is an essential process to curb the spread of this deadly disease and its mortality rate. Chest radiology scan is a significant tool for early management and diagnosis of COVID-19 since the virus targets the respiratory system. Chest X-ray (CXR) images are highly useful in the effective detection of COVID-19, thanks to its availability, cost-effective means, and rapid outcomes. In addition, Artificial Intelligence (AI) techniques such as deep learning (DL) models play a significant role in designing automated diagnostic processes using CXR images. With this motivation, the current study presents a new Quantum Seagull Optimization Algorithm with DL-based COVID-19 diagnosis model, named QSGOA-DL technique. The proposed QSGOA-DL technique intends to detect and classify COVID-19 with the help of CXR images. In this regard, the QSGOA-DL technique involves the design of EfficientNet-B4 as a feature extractor, whereas hyperparameter optimization is carried out with the help of QSGOA technique. Moreover, the classification process is performed by a multilayer extreme learning machine (MELM) model. The novelty of the study lies in the designing of QSGOA for hyperparameter optimization of the EfficientNet-B4 model. An extensive series of simulations was carried out on the benchmark test CXR dataset, and the results were assessed under different aspects. The simulation results demonstrate the promising performance of the proposed QSGOA-DL technique compared to recent approaches.

COVID-19 Identification System Using Transfer Learning Technique With Mobile-NetV2 and Chest X-Ray Images.

Diagnosis is a crucial precautionary step in research studies of the coronavirus disease, which shows indications similar to those of various pneumonia types. The COVID-19 pandemic has caused a significant outbreak in more than 150 nations and has significantly affected the wellness and lives of many individuals globally. Particularly, discovering the patients infected with COVID-19 early and providing them with treatment is an important way of fighting the pandemic. Radiography and radiology could be the fastest techniques for recognizing infected individuals. Artificial intelligence strategies have the potential to overcome this difficulty. Particularly, transfer learning MobileNetV2 is a convolutional neural network architecture that can perform well on mobile devices. In this study, we used MobileNetV2 with transfer learning and augmentation data techniques as a classifier to recognize the coronavirus disease. Two datasets were used: the first consisted of 309 chest X-ray images (102 with COVID-19 and 207 were normal), and the second consisted of 516 chest X-ray images (102 with COVID-19 and 414 were normal). We assessed the model based on its sensitivity rate, specificity rate, confusion matrix, and F1-measure. Additionally, we present a receiver operating characteristic curve. The numerical simulation reveals that the model accuracy is 95.8% and 100% at dropouts of 0.3 and 0.4, respectively. The model was implemented using Keras and Python programming.

Ensemble Deep-Learning-Enabled Clinical Decision Support System for Breast Cancer Diagnosis and Classification on Ultrasound Images.

Clinical Decision Support Systems (CDSS) provide an efficient way to diagnose the presence of diseases such as breast cancer using ultrasound images (USIs). Globally, breast cancer is one of the major causes of increased mortality rates among women. Computer-Aided Diagnosis (CAD) models are widely employed in the detection and classification of tumors in USIs. The CAD systems are designed in such a way that they provide recommendations to help radiologists in diagnosing breast tumors and, furthermore, in disease prognosis. The accuracy of the classification process is decided by the quality of images and the radiologist's experience. The design of Deep Learning (DL) models is found to be effective in the classification of breast cancer. In the current study, an Ensemble Deep-Learning-Enabled Clinical Decision Support System for Breast Cancer Diagnosis and Classification (EDLCDS-BCDC) technique was developed using USIs. The proposed EDLCDS-BCDC technique was intended to identify the existence of breast cancer using USIs. In this technique, USIs initially undergo pre-processing through two stages, namely wiener filtering and contrast enhancement. Furthermore, Chaotic Krill Herd Algorithm (CKHA) is applied with Kapur's entropy (KE) for the image segmentation process. In addition, an ensemble of three deep learning models, VGG-16, VGG-19, and SqueezeNet, is used for feature extraction. Finally, Cat Swarm Optimization (CSO) with the Multilayer Perceptron (MLP) model is utilized to classify the images based on whether breast cancer exists or not. A wide range of simulations were carried out on benchmark databases and the extensive results highlight the better outcomes of the proposed EDLCDS-BCDC technique over recent methods.

Detection of microscopic glaucoma through fundus images using deep transfer learning approach.

Glaucoma disease in humans can lead to blindness if it progresses to the point where it affects the oculus' optic nerve head. It is not easily detected since there are no symptoms, but it can be detected using tonometry, ophthalmoscopy, and perimeter. However, advances in artificial intelligence approaches have permitted machine learning techniques to diagnose at an early stage. Numerous methods have been proposed using Machine Learning to diagnose glaucoma with different data sets and techniques but these are complex methods. Although, medical imaging instruments are used as glaucoma screening methods, fundus imaging specifically is the most used screening technique for glaucoma detection. This study presents a novel DenseNet and DarkNet combination to classify normal and glaucoma affected fundus image. These frameworks have been trained and tested on three data sets of high-resolution fundus (HRF), RIM 1, and ACRIMA. A total of 658 images have been used for healthy eyes and 612 images for glaucoma-affected eyes classification. It has also been observed that the fusion of DenseNet and DarkNet outperforms the two CNN networks and achieved 99.7% accuracy, 98.9% sensitivity, 100% specificity for the HRF database. In contrast, for the RIM1 database, 89.3% accuracy, 93.3% sensitivity, 88.46% specificity has been attained. Moreover, for the ACRIMA database, 99% accuracy, 100% sensitivity, 99% specificity has been achieved. Therefore, the proposed method is robust and efficient with less computational time and complexity compared to the literature available.