Inflammation of the lungs, called pneumonia, is a disease characterized by inflammation of the air sacs that interfere with the exchange of oxygen and carbon dioxide. It is caused by a variety of infectious organisms, including viruses, bacteria, fungus, and parasites. Pneumonia is more common in people who have pre-existing lung diseases or compromised immune systems, and it primarily affects small children and the elderly. Diagnosis of pneumonia can be difficult, especially when relying on medical imaging, because symptoms may not be immediately apparent. Convolutional neural networks (CNNs) have recently shown potential in medical imaging applications. A CNN-based deep learning model is being built as part of ongoing research to aid in the detection of pneumonia using chest X-ray images. The dataset used for training and evaluation includes images of people with normal lung conditions as well as photos of people with pneumonia. Various preprocessing procedures, such as data augmentation, normalization, and scaling, were used to improve the accuracy of pneumonia diagnosis and extract significant features. In this study, a framework for deep learning with four pre-trained CNN models—InceptionNet, ResNet, VGG16, and DenseNet—was used. To take use of its key advantages, transfer learning utilizing DenseNet was used. During training, the loss function was minimized using the Adam optimizer. The suggested approach seeks to improve early diagnosis and enable fast intervention for pneumonia cases by leveraging the advantages of several CNN models. The outcomes show that CNN-based deep learning models may successfully diagnose pneumonia in chest X-ray pictures.

Retinal disorders, such as diabetic retinopathy, glaucoma, macular edema, and vein occlusions, are significant contributors to global vision impairment. These conditions frequently remain symptomless until patients suffer severe vision deterioration, underscoring the critical importance of early diagnosis. Fundus images serve as a valuable resource for identifying the initial indicators of these ailments, particularly by examining various characteristics of retinal blood vessels, such as their length, width, tortuosity, and branching patterns. Traditionally, healthcare practitioners often rely on manual retinal vessel segmentation, a process that is both time-consuming and intricate, demanding specialized expertise. However, this approach poses a notable challenge since its precision and consistency heavily rely on the availability of highly skilled professionals. To surmount these challenges, there is an urgent demand for an automatic and efficient method for retinal vessel segmentation and classification employing computer vision techniques, which form the foundation of biomedical imaging. Numerous researchers have put forth techniques for blood vessel segmentation, broadly categorized into machine learning, filtering-based, and model-based methods. Machine learning methods categorize pixels as either vessels or non-vessels, employing classifiers trained on hand-annotated images. Subsequently, these techniques extract features using 7D feature vectors and apply neural network classification. Additional post-processing steps are used to bridge gaps and eliminate isolated pixels. On the other hand, filtering-based approaches employ morphological operators within morphological image processing, capitalizing on predefined shapes to filter out objects from the background. However, this technique often treats larger blood vessels as cohesive structures. Model-based methods leverage vessel models to identify retinal blood vessels, but they are sensitive to parameter selection, necessitating careful choices to simultaneously detect thin and large vessels effectively. Our proposed research endeavors to conduct a thorough and empirical evaluation of the effectiveness of automated segmentation and classification techniques for identifying eye-related diseases, particularly diabetic retinopathy and glaucoma. This evaluation will involve various retinal image datasets, including DRIVE, REVIEW, STARE, HRF, and DRION. The methodologies under consideration encompass machine learning, filtering-based, and model-based approaches, with performance assessment based on a range of metrics, including true positive rate (TPR), true negative rate (TNR), positive predictive value (PPV), negative predictive value (NPV), false discovery rate (FDR), Matthews's correlation coefficient (MCC), and accuracy (ACC). The primary objective of this research is to scrutinize, assess, and compare the design and performance of different segmentation and classification techniques, encompassing both supervised and unsupervised learning methods. To attain this objective, we will refine existing techniques and develop new ones, ensuring a more streamlined and computationally efficient approach.

The study evaluated 33 accessions of groundnut in the field, consisting of 23 landraces from Nasarawa communities in Nigeria and 10 inbred lines. Assessment entailed the determination of plant survivorship, yield related parameters and pathological indices while genetic diversity study was undertaken using SSR and RAPD molecular markers. Data analysis was done on the Minitab 17.0 software. Significant variability was noted in all traits except in pod sizes, seed sizes and % infected seeds. About 33.3% of the accessions had a survival rate of ≥ 70.0% where 9/10 Inbred lines were found with overall yield (kg/ha) ranging from 4.0 ± 1.6 in Akwashiki-Doma to 516.8 ± 46.9 kg/ha in Samnut 24 × ICGV–91328. Five accessions (15.5%) had pathological indices of zero indicating no traces of any disease of any type and they included one landrace called Agric-Dazhogwa and four Inbred lines: Samnut 25 × ICGV–91317, Samnut 26 × ICGV–19324, Samnut 26 × ICGV–91328 and Samnut 26 × ICGV–91319. Coefficients of yield determination R2 by survivorship and pathological index were 50.6% and 15%, respectively. A fit model was established (Yield in kg/ha = –146 − 7.94 × Pi + 5.88 × S). Susceptibility to diseases depends on the type of variety (χ2(32) = 127.67, P = 0.00). Yield was significantly affected by BNR@30 (F = 5.47, P = 0.025, P < 0.05) and DSV@60*RUST@60 interaction effect (F = 4.39, P = 0.044, P < 0.05). The similarity coefficient ranged from 28.57 to 100 in plant morphology. Four varieties had no amplified bands with SSR primers whereas amplified bands were present only in four landraces accessions using the RAPD primer. The dendrogram generated by molecular data gave three groups where genetic similarity ranged from 41.4 to 100.0. The Inbred lines were noted for their high survivorship, good yield and disease resistance. Samnut 24 × ICGV–91328, an inbred line, had the highest yield but was susceptible to diseases. Among the landraces, Agric-Musha, Bomboyi-Dugu and Agric-Dazhogwa were selected for high survivorship and disease resistance. The selected inbred lines and landraces are valuable genetic resources that may harbour useful traits for breeding and they should be presented to the growers based on their unique agronomic values. The highest yielding inbred lines should be improved for resistance to late leaf spot diseases while the outstanding landraces should be improved for yield.