A hybrid features fusion-based framework for classification of breast micronodules using ultrasonography.

BACKGROUND: Breast cancer is one of the leading diseases worldwide. According to estimates by the National Breast Cancer Foundation, over 42,000 women are expected to die from this disease in 2024. OBJECTIVE: The prognosis of breast cancer depends on the early detection of breast micronodules and the ability to distinguish benign from malignant lesions. Ultrasonography is a crucial radiological imaging technique for diagnosing the illness because it allows for biopsy and lesion characterization. The user's level of experience and knowledge is vital since ultrasonographic diagnosis relies on the practitioner's expertise. Furthermore, computer-aided technologies significantly contribute by potentially reducing the workload of radiologists and enhancing their expertise, especially when combined with a large patient volume in a hospital setting. METHOD: This work describes the development of a hybrid CNN system for diagnosing benign and malignant breast cancer lesions. The models InceptionV3 and MobileNetV2 serve as the foundation for the hybrid framework. Features from these models are extracted and concatenated individually, resulting in a larger feature set. Finally, various classifiers are applied for the classification task. RESULTS: The model achieved the best results using the softmax classifier, with an accuracy of over 95%. CONCLUSION: Computer-aided diagnosis greatly assists radiologists and reduces their workload. Therefore, this research can serve as a foundation for other researchers to build clinical solutions.

Nutritional Profiles and Their Links to Insulin Resistance and Anthropometric Variables in a Female Cohort.

This study investigates the relationship between dietary habits and metabolic health among women, emphasizing the role of anthropometric parameters as proxies for insulin resistance. We analyzed data from 443 women categorized into two groups based on the presence or absence of clinically diagnosed insulin resistance. Our assessments included dietary quality, socio-demographic characteristics, and a series of anthropometric measurements such as body weight, Body Mass Index (BMI), Waist-Hip Ratio (WHR), Abdominal Volume Index (AVI), and Body Adiposity Index (BAI). The results indicated significant disparities in these parameters, with the insulin-resistant group exhibiting higher average body weight (78.92 kg vs. 65.04 kg, p < 0.001), BMI (28.45 kg/m2 vs. 23.17 kg/m2, p < 0.001), and other related measures, suggesting a strong influence of dietary patterns on body composition and metabolic risk. The study underscores the importance of dietary management in addressing insulin resistance, advocating for personalized dietary strategies to improve metabolic health outcomes in women. This approach highlights the need for integrating dietary changes with lifestyle modifications and socio-demographic considerations to combat metabolic risks effectively.

Robust Ulcer Classification: Contrast and Illumination Invariant Approach.

Gastrointestinal (GI) disease cases are on the rise throughout the world. Ulcers, being the most common type of GI disease, if left untreated, can cause internal bleeding resulting in anemia and bloody vomiting. Early detection and classification of different types of ulcers can reduce the death rate and severity of the disease. Manual detection and classification of ulcers are tedious and error-prone. This calls for automated systems based on computer vision techniques to detect and classify ulcers in images and video data. A major challenge in accurate detection and classification is dealing with the similarity among classes and the poor quality of input images. Improper contrast and illumination reduce the anticipated classification accuracy. In this paper, contrast and illumination invariance was achieved by utilizing log transformation and power law transformation. Optimal values of the parameters for both these techniques were achieved and combined to obtain the fused image dataset. Augmentation was used to handle overfitting and classification was performed using the lightweight and efficient deep learning model MobilNetv2. Experiments were conducted on the KVASIR dataset to assess the efficacy of the proposed approach. An accuracy of 96.71% was achieved, which is a considerable improvement over the state-of-the-art techniques.