An Alternative Diagnostic Method for C. neoformans: Preliminary Results of Deep-Learning Based Detection Model

Çağatan, Ayşe Seyer; Mustapha, Mubarak Taiwo; Bagkur, Cemile; Şanlıdağ, Tamer; Ozsahin, Dilber Uzun

doi:10.3390/diagnostics13010081

Cited by 10 publications

(9 citation statements)

References 31 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Furthermore, conducting tests on a more extensive variety of fungal species could yield valuable insights. In a study by Cagatan et al, a VGG16-based method was introduced for identifying Cryptococcus neoformans, a fungal pathogen, in patient samples [5]. The initial dataset for this study comprised only 63 images, later augmented to 1000 through the generation of synthetic images using data augmentation techniques.…”

Section: Discussionmentioning

confidence: 99%

Residual neural networks in single instance-driven identification of fungal pathogens

Wyszyński,

Struniawski

2023

MG&V

View full text Add to dashboard Cite

The rise in fungal infections, attributed to various factors including medical interventions and compromised immune systems, necessitates rapid and accurate identification methods. While traditional mycological diagnostics are time-consuming, machine learning offers a promising alternative. Nevertheless, the scarcity of well-curated datasets is a significant obstacle. To address this, a novel approach for identifying fungi in microscopic images using Residual Neural Networks and a subimage retrieval mechanism is proposed, with the final step involving the implementation of majority voting. The new method, applied to the Digital Images of Fungus Species database, surpassed the original patch-based classification using Convolutional Neural Networks, obtaining an overall classification accuracy of 94.7% compared to 82.4% with AlexNet FV SVM. The observed MCC metric exceeds 0.9, while AUC is near to one. This improvement is attributed to the optimization of hyperparameters and top layer architecture, as well as the effectiveness of the Mish activation function in ResNet-based architectures. Noteworthy, the proposed method achieved 100% accurate classification for images from 8 out of 9 classes after majority voting and is high resistant to overfitting, highlighting its potential for rapid and accurate fungal species identification in medical diagnostics and research.

show abstract

Section: Discussionmentioning

confidence: 99%

Residual neural networks in single instance-driven identification of fungal pathogens

Wyszyński,

Struniawski

2023

MG&V

View full text Add to dashboard Cite

show abstract

“…The next step was to arrange the dataset according to the order that would enable evaluation. During the pre-processing phase [ 22 ], a table of records was converted into a more usable format through a series of steps: Data from two countries (Greece and Israel) were collected from the overall dataset to enable us to carry out the evaluation. Columns and rows containing no valid data were deleted.…”

Section: Methodsmentioning

confidence: 99%

“… For our prediction, only datasets from Israel were used to train the model. Data normalization [ 21 , 22 ] was carried out prior to modeling using Equation (1).

where

is the measured data

and

are the minimum and maximum values, respectively.…”

Section: Methodsmentioning

confidence: 99%

“…If mistakes are made, they are passed around the system until the desired result is achieved. The FFNN-BP algorithm’s central idea is to minimize the network’s error so that it can fully understand the training data and make more precise predictions of the true value [ 22 ]. During operation, the initial weights are multiplied by the inputs, and the resulting value is transferred to the second layer, where it remains until it reaches the output layer, as shown in the following equation:

where x ij is an illustration of the input, y i is the consequent sum of outputs from the ith node, and zi is the weight transferred from the j th input to the i th node.…”

Section: Methodsmentioning

confidence: 99%

See 1 more Smart Citation

COVID-19 Prediction Using Black-Box Based Pearson Correlation Approach

Ozsahin

Onakpojeruo²,

Duwa³

et al. 2023

Diagnostics

View full text Add to dashboard Cite

The novel coronavirus (COVID-19), also known as SARS-CoV-2, is a highly contagious respiratory disease that first emerged in Wuhan, China in 2019 and has since become a global pandemic. The virus is spread through respiratory droplets produced when an infected person coughs or sneezes, and it can lead to a range of symptoms, from mild to severe. Some people may not have any symptoms at all and can still spread the virus to others. The best way to prevent the spread of COVID-19 is to practice good hygiene. It is also important to follow the guidelines set by local health authorities, such as physical distancing and quarantine measures. The World Health Organization (WHO), on the other hand, has classified this virus as a pandemic, and as a result, all nations are attempting to exert control and secure all public spaces. The current study aimed to (I) compare the weekly COVID-19 cases between Israel and Greece, (II) compare the monthly COVID-19 mortality cases between Israel and Greece, (III) evaluate and report the influence of the vaccination rate on COVID-19 mortality cases in Israel, and (IV) predict the number of COVID-19 cases in Israel. The advantage of completing these tasks is the minimization of the spread of the virus by deploying different mitigations. To attain our objective, a correlation analysis was carried out, and two distinct artificial intelligence (AI)-based models—specifically, an artificial neural network (ANN) and a classical multiple linear regression (MLR)—were developed for the prediction of COVID-19 cases in Greece and Israel by utilizing related variables as the input variables for the models. For the evaluation of the models, four evaluation metrics (determination coefficient (R2), mean square error (MSE), root mean square error (RMSE), and correlation coefficient (R)) were considered in order to determine the performance of the deployed models. From a variety of perspectives, the corresponding determination coefficient (R2) demonstrated the statistical advantages of MLR over the ANN model by following a linear pattern. The MLR predictive model was both efficient and accurate, with 98% accuracy, while ANN showed 94% accuracy in the effective prediction of COVID-19 cases.

show abstract

“…With the advancement of computer vision technologies and the availability of large-scale data sets, powerful arti cial intelligence, especially machine learning, have been used to automatically generate diagnosis through algorithm for image analysis [4] . Deep learning as a branch of machine learning is composed of sequentially arranged multilayer neural networks that can learn directly from raw data by learning highly complex functions to develop its own pattern recognition, and its task of detection, prediction and classi cation in medical image diagnosis helps to overcome the shortcomings of physicians'subjective visual assessments.…”

Section: Introductionmentioning

confidence: 99%

Development of a deep learning-based grading model to evaluate fetal lung maturity from normal fetal lung ultrasound images

Zhao

Zhang

Wang

et al. 2023

Preprint

View full text Add to dashboard Cite

Objectives To develop a deep learning algorithm for quantifying fetal lung maturity from normal fetal lung ultrasound images. Methods This is a single-center analytical study that retrospectively cross-sectionally observes singleton pregnant woman without pregnancy complications at 20–41+ 6 weeks of gestation, and acquires axial images of their fetuses at the level of the four-chambered heart in order to create a model for evaluating the maturity of the fetal lungs using normal fetal lung ultrasound images. Assuming that the texture information of the fetal lung ultrasound image can represent the maturity of the fetal lung development, and the gestational age is proportional to the maturity of the fetal lung, the deep learning grading model based on the normal fetal lung ultrasound image is established with the gestational age as the baseline. The pictures were split into three classes according to the gestational weeks which were calculated at the last menstrual period of the pregnant woman as the reference standard: 20 to 29+ 6 weeks for class I, 30 to 36+ 6 weeks for class II, and 37 to 41+ 6 weeks for class III. Results This study had 350 pregnant women in total, and the deep learning model's classification accuracy for ultrasound images of fetal lungs at different gestational ages was 0.917. Conclusion The grading model based on deep learning can correctly and independently predict the fetal lung maturity from ultrasound images.

show abstract

An Alternative Diagnostic Method for C. neoformans: Preliminary Results of Deep-Learning Based Detection Model

Cited by 10 publications

References 31 publications

Residual neural networks in single instance-driven identification of fungal pathogens

Residual neural networks in single instance-driven identification of fungal pathogens

COVID-19 Prediction Using Black-Box Based Pearson Correlation Approach

Development of a deep learning-based grading model to evaluate fetal lung maturity from normal fetal lung ultrasound images

Contact Info

Product

Resources

About