Analysis of the Application Efficiency of TensorFlow and PyTorch in Convolutional Neural Network

Role of artificial intelligence, machine learning and deep learning models in corneal disorders - A narrative review

In the last decade, artificial intelligence (AI) has significantly impacted ophthalmology, particularly in managing corneal diseases, a major reversible cause of blindness. This review explores AI's transformative role in the corneal subspecialty, which has adopted advanced technology for superior clinical judgment, early diagnosis, and personalized therapy. While AI's role in anterior segment diseases is less documented compared to glaucoma and retinal pathologies, this review highlights its integration into corneal diagnostics through imaging techniques like slit-lamp biomicroscopy, anterior segment optical coherence tomography (AS-OCT), and in vivo confocal biomicroscopy. AI has been pivotal in refining decision-making and prognosis for conditions such as keratoconus, infectious keratitis, and dystrophies. Multi-disease deep learning neural networks (MDDNs) have shown diagnostic ability in classifying corneal diseases using AS-OCT images, achieving notable metrics like an AUC of 0.910. AI's progress over two decades has significantly improved the accuracy of diagnosing conditions like keratoconus and microbial keratitis. For instance, AI has achieved a 90.7% accuracy rate in classifying bacterial and fungal keratitis and an AUC of 0.910 in differentiating various corneal diseases. Convolutional neural networks (CNNs) have enhanced the analysis of color-coded corneal maps, yielding up to 99.3% diagnostic accuracy for keratoconus. Deep learning algorithms have also shown robust performance in detecting fungal hyphae on in vivo confocal microscopy, with precise quantification of hyphal density. AI models combining tomography scans and visual acuity have demonstrated up to 97% accuracy in keratoconus staging according to the Amsler-Krumeich classification. However, the review acknowledges the limitations of current AI models, including their reliance on binary classification, which may not capture the complexity of real-world clinical presentations with multiple coexisting disorders. Challenges also include dependency on data quality, diverse imaging protocols, and integrating multimodal images for a generalized AI diagnosis. The need for interpretability in AI models is emphasized to foster trust and applicability in clinical settings. Looking ahead, AI has the potential to unravel the intricate mechanisms behind corneal pathologies, reduce healthcare's carbon footprint, and revolutionize diagnostic and management paradigms. Ethical and regulatory considerations will accompany AI's clinical adoption, marking an era where AI not only assists but augments ophthalmic care.

Deep Learning Based on ResNet-18 for Classification of Prostate Imaging-Reporting and Data System Category 3 Lesions

RATIONALE AND OBJECTIVES

To explore the classification and prediction efficacy of the deep learning model for benign prostate lesions, non-clinically significant prostate cancer (non-csPCa) and clinically significant prostate cancer (csPCa) in Prostate Imaging-Reporting and Data System (PI-RADS) 3 lesions.

MATERIALS AND METHODS

From January 2015 to December 2021, lesions diagnosed with PI-RADS 3 by multi-parametric MRI or bi-parametric MRI were retrospectively included. They were classified as benign prostate lesions, non-csPCa, and csPCa according to the pathological results. T2-weighted images of the lesions were divided into a training set and a test set according to 8:2. ResNet-18 was used for model training. All statistical analyses were performed using Python open-source libraries. The receiver operating characteristic curve (ROC) was used to evaluate the predictive effectiveness of the model. T-SNE was used for image semantic feature visualization. The class activation mapping was used to visualize the area focused by the model.

RESULTS

A total of 428 benign prostate lesion images, 158 non-csPCa images and 273 csPCa images were included. The precision in predicting benign prostate disease, non-csPCa and csPCa were 0.882, 0.681 and 0.851, and the area under the ROC were 0.875, 0.89 and 0.929, respectively. Semantic feature analysis showed strong classification separability between csPCa and benign prostate lesions. The class activation map showed that the deep learning model can focus on the area of the prostate or the location of PI-RADS 3 lesions.

CONCLUSION

Deep learning model with T2-weighted images based on ResNet-18 can realize accurate classification of PI-RADS 3 lesions.

Predicting Diffusion Coefficients in Nafion Membranes during the Soaking Process Using a Machine Learning Approach

Nafion, a versatile polymer used in electrochemistry and membrane technologies, exhibits complex behaviors in saline environments. This study explores Nafion membrane's IR spectra during soaking and subsequent drying processes in salt solutions at various concentrations. Utilizing the principles of Fick's second law, diffusion coefficients for these processes are derived via exponential approximation. By harnessing machine learning (ML) techniques, including the optimization of neural network hyperparameters via a genetic algorithm (GA) and leveraging various regressors, we effectively pinpointed the optimal model for predicting diffusion coefficients. Notably, for the prediction of soaking coefficients, our model is composed of layers with 64, 64, 32, and 16 neurons, employing ReLU, ELU, sigmoid, and ELU activation functions, respectively. Conversely, for drying coefficients, our model features two hidden layers with 16 and 12 neurons, utilizing sigmoid and ELU activation functions, respectively.

Deep learning-based automatic segmentation of meningioma from T1-weighted contrast-enhanced MRI for preoperative meningioma differentiation using radiomic features

BACKGROUND

This study aimed to establish a dedicated deep-learning model (DLM) on routine magnetic resonance imaging (MRI) data to investigate DLM performance in automated detection and segmentation of meningiomas in comparison to manual segmentations. Another purpose of our work was to develop a radiomics model based on the radiomics features extracted from automatic segmentation to differentiate low- and high-grade meningiomas before surgery.

MATERIALS

A total of 326 patients with pathologically confirmed meningiomas were enrolled. Samples were randomly split with a 6:2:2 ratio to the training set, validation set, and test set. Volumetric regions of interest (VOIs) were manually drawn on each slice using the ITK-SNAP software. An automatic segmentation model based on SegResNet was developed for the meningioma segmentation. Segmentation performance was evaluated by dice coefficient and 95% Hausdorff distance. Intra class correlation (ICC) analysis was applied to assess the agreement between radiomic features from manual and automatic segmentations. Radiomics features derived from automatic segmentation were extracted by pyradiomics. After feature selection, a model for meningiomas grading was built.

RESULTS

The DLM detected meningiomas in all cases. For automatic segmentation, the mean dice coefficient and 95% Hausdorff distance were 0.881 (95% CI: 0.851-0.981) and 2.016 (95% CI:1.439-3.158) in the test set, respectively. Features extracted on manual and automatic segmentation are comparable: the average ICC value was 0.804 (range, 0.636-0.933). Features extracted on manual and automatic segmentation are comparable: the average ICC value was 0.804 (range, 0.636-0.933). For meningioma classification, the radiomics model based on automatic segmentation performed well in grading meningiomas, yielding a sensitivity, specificity, accuracy, and area under the curve (AUC) of 0.778 (95% CI: 0.701-0.856), 0.860 (95% CI: 0.722-0.908), 0.848 (95% CI: 0.715-0.903) and 0.842 (95% CI: 0.807-0.895) in the test set, respectively.

CONCLUSIONS

The DLM yielded favorable automated detection and segmentation of meningioma and can help deploy radiomics for preoperative meningioma differentiation in clinical practice.

[Comparative evaluation of the accuracy of 3D TMJ analysis performed by different methods of processing computed tomograms]

OBJECTIVE

The aim of this study. Comparison of the accuracy of segmentation of TMJ elements in different ways and assessment of the suitability of the data obtained for the diagnosis of TMJ dysfunction.

MATERIALS AND METHODS

To study the segmentation of the bone elements of the TMJ (articular fossa, head of the LF), 60 computed tomograms of the maxillofacial region of patients were randomly selected in various ways (archival material). In group 1, the results of CT processing by AI diagnostics algorithms (Russia) were collected; in group 2, the results of CT processing based on the semi-automatic segmentation method in the Avantis3D program. The results of CT processing by Avantis3D AI algorithms (Russia) with different probability modes - 0.4 and 0.9, respectively, were selected for the third and fourth groups. Visually, the coincidence of the contours of the LF heads and articular pits isolated using different methods with their contours on all possible sections of the original CT itself was evaluated. The time spent on TMJ segmentation according to CT data was determined and compared using the methods described above.

RESULTS

Of the 240 objects, only 7.5% of the cases showed a slight discrepancy between the contours of the original CT in group b1, which was the lowest of all. A slight discrepancy in the TMJ contours to be corrected is characteristic of the semi-automatic method of segmentation by optical density was detected in 50.4% (group 2). The largest percentage of significant errors not subject to correction was noted in the first group, which made it impossible to perform a full 3D analysis of the TMJ, and the smallest in the second and fourth. The magnitude of the error in determining the width of the articular gap in different groups is comparable to the size of one voxel per CT. When segmentation is carried out using AI, the difference between segmented objects is close to zero values. The average time spent on TMJ segmentation in group 1 was 10.2±1.23 seconds, in group 2 - 12.6±1.87 seconds, in groups 3 and 4 - 0.46±0.12 seconds and 0.46±0.13 seconds, respectively.

CONCLUSION

The developed automated method for segmenting TMJ elements using AI is obviously more suitable for practical work, since it requires minimal time, and is almost as accurate as other methods under consideration.

Identification of antigen-presentation related B cells as a key player in Crohn's disease using single-cell dissecting, hdWGCNA, and deep learning

Crohn's disease (CD) arises from intricate intercellular interactions within the intestinal lamina propria. Our objective was to use single-cell RNA sequencing to investigate CD pathogenesis and explore its clinical significance. We identified a distinct subset of B cells, highly infiltrated in the CD lamina propria, that expressed genes related to antigen presentation. Using high-dimensional weighted gene co-expression network analysis and nine machine learning techniques, we demonstrated that the antigen-presenting CD-specific B cell signature effectively differentiated diseased mucosa from normal mucosa (Independent external testing AUC = 0.963). Additionally, using MCPcounter and non-negative matrix factorization, we established a relationship between the antigen-presenting CD-specific B cell signature and immune cell infiltration and patient heterogeneity. Finally, we developed a gene-immune convolutional neural network deep learning model that accurately diagnosed CD mucosa in diverse cohorts (Independent external testing AUC = 0.963). Our research has revealed a population of B cells with a potential promoting role in CD pathogenesis and represents a fundamental step in the development of future clinical diagnostic tools for the disease.

Open-Source Machine Learning in Computational Chemistry

The field of computational chemistry has seen a significant increase in the integration of machine learning concepts and algorithms. In this Perspective, we surveyed 179 open-source software projects, with corresponding peer-reviewed papers published within the last 5 years, to better understand the topics within the field being investigated by machine learning approaches. For each project, we provide a short description, the link to the code, the accompanying license type, and whether the training data and resulting models are made publicly available. Based on those deposited in GitHub repositories, the most popular employed Python libraries are identified. We hope that this survey will serve as a resource to learn about machine learning or specific architectures thereof by identifying accessible codes with accompanying papers on a topic basis. To this end, we also include computational chemistry open-source software for generating training data and fundamental Python libraries for machine learning. Based on our observations and considering the three pillars of collaborative machine learning work, open data, open source (code), and open models, we provide some suggestions to the community.

Transverse Deflection for Extreme Ultraviolet Pellicles

Defect control of extreme ultraviolet (EUV) masks using pellicles is challenging for mass production in EUV lithography because EUV pellicles require more critical fabrication than argon fluoride (ArF) pellicles. One of the fabrication requirements is less than 500 μm transverse deflections with more than 88% transmittance of full-size pellicles (112 mm × 145 mm) at pressure 2 Pa. For the nanometer thickness (thickness/width length (t/L) = 0.0000054) of EUV pellicles, this study reports the limitation of the student's version and shear locking in a commercial tool-based finite element method (FEM) such as ANSYS and SIEMENS. A Python program-based analytical-numerical method with deep learning is described as an alternative. Deep learning extended the ANSYS limitation and overcame shear locking. For EUV pellicle materials, the ascending order of transverse deflection was Ru<MoSi2=SiC<SiNx<ZrSr2<p-Si<Sn in both ANSYS and a Python program, regardless of thickness and pressure. According to a neural network, such as the Taguchi method, the sensitivity order of EUV pellicle parameters was Poisson's ratio<Elastic modulus<Pressure<Thickness<Length.