Journal: Journal of Imaging

Article Title: Automated Detection and Diagnosis of Diabetic Retinopathy: A Comprehensive Survey

doi: 10.3390/jimaging7090165

Figure Lengend Snippet: Classification-based studies in DR detection using fundus imaging.

Article Snippet: Samanta, 2020 [ ] , EyePACS , Grade DR based on ICDR scale , Yes , DCNN: DenseNet121 based , 84.1% , NA , NA , NA.

Techniques: Imaging, Modification, Extraction

Journal: Scientific Reports

Article Title: Deep visual detection system for oral squamous cell carcinoma

doi: 10.1038/s41598-025-34332-5

Figure Lengend Snippet: Proposed deep visual detection system using DenseNet121 for binary classification on the Kaggle OSCC dataset.

Article Snippet: DenseNet121 showed moderate performance, especially on the Kaggle dataset with 86.91% accuracy, but failed to generalize effectively on the more complex NDB-UFES dataset.

Techniques:

Journal: Scientific Reports

Article Title: Deep visual detection system for oral squamous cell carcinoma

doi: 10.1038/s41598-025-34332-5

Figure Lengend Snippet: Key Hyperparameters of the DenseNet121 model (Kaggle Binary Class OSCC Dataset).

Article Snippet: DenseNet121 showed moderate performance, especially on the Kaggle dataset with 86.91% accuracy, but failed to generalize effectively on the more complex NDB-UFES dataset.

Techniques:

Journal: Scientific Reports

Article Title: Deep visual detection system for oral squamous cell carcinoma

doi: 10.1038/s41598-025-34332-5

Figure Lengend Snippet: Key Hyperparameters of the DenseNet121 model (NDB-UFES Multiclass OSCC Dataset).

Article Snippet: DenseNet121 showed moderate performance, especially on the Kaggle dataset with 86.91% accuracy, but failed to generalize effectively on the more complex NDB-UFES dataset.

Techniques:

Journal: Scientific Reports

Article Title: Deep visual detection system for oral squamous cell carcinoma

doi: 10.1038/s41598-025-34332-5

Figure Lengend Snippet: Training and validation loss & accuracy curve—DenseNet121 (Epochs = 20, Batch Size = 64) Kaggle Binary Class OSCC Dataset.

Article Snippet: DenseNet121 showed moderate performance, especially on the Kaggle dataset with 86.91% accuracy, but failed to generalize effectively on the more complex NDB-UFES dataset.

Techniques: Biomarker Discovery

Journal: Scientific Reports

Article Title: Deep visual detection system for oral squamous cell carcinoma

doi: 10.1038/s41598-025-34332-5

Figure Lengend Snippet: Training and validation loss & accuracy curve—DenseNet121 (Epochs = 20, Batch Size = 64) NDB-UFES Multiclass OSCC Dataset.

Article Snippet: DenseNet121 showed moderate performance, especially on the Kaggle dataset with 86.91% accuracy, but failed to generalize effectively on the more complex NDB-UFES dataset.

Techniques: Biomarker Discovery

Journal: Scientific Reports

Article Title: Deep visual detection system for oral squamous cell carcinoma

doi: 10.1038/s41598-025-34332-5

Figure Lengend Snippet: Confusion matrix of DenseNet121 model (Epochs = 20, Batch Size = 64)—Kaggle Binary Class OSCC dataset.

Article Snippet: DenseNet121 showed moderate performance, especially on the Kaggle dataset with 86.91% accuracy, but failed to generalize effectively on the more complex NDB-UFES dataset.

Techniques:

Journal: Scientific Reports

Article Title: Deep visual detection system for oral squamous cell carcinoma

doi: 10.1038/s41598-025-34332-5

Figure Lengend Snippet: Confusion matrix of DenseNet121 model (Epochs = 20, Batch Size = 64)—NDB-UFES Multiclass OSCC dataset.

Article Snippet: DenseNet121 showed moderate performance, especially on the Kaggle dataset with 86.91% accuracy, but failed to generalize effectively on the more complex NDB-UFES dataset.

Techniques:

Journal: Scientific Reports

Article Title: Deep visual detection system for oral squamous cell carcinoma

doi: 10.1038/s41598-025-34332-5

Figure Lengend Snippet: Performance comparison of EfficientNetB3, DenseNet121, and ResNet50 on Kaggle Binary-Class and NDB-UFES multiclass OSCC datasets.

Article Snippet: DenseNet121 showed moderate performance, especially on the Kaggle dataset with 86.91% accuracy, but failed to generalize effectively on the more complex NDB-UFES dataset.

Techniques: Comparison

Journal: Frontiers in Big Data

Article Title: Artificial intelligence for the detection of acute myeloid leukemia from microscopic blood images; a systematic review and meta-analysis

doi: 10.3389/fdata.2024.1402926

Figure Lengend Snippet: Types of models used and their specifications.

Article Snippet: Sakthiraj ( ) , Hybrid Convolutional Neural Network with Interactive Autodidactic School (HCNN-IAS) algorithm , HCNN-IASO , No , Softmax-CNN layer classifier (based on ResNet-34 and DenseNet-121) , IV , -Before augmentation: Healthy: 190 CML: 58 CLL: 30 AML: 56 ALL: 182 -After augmentation: Healthy: 1,291 CML: 1,244 CLL: 845 AML: 1,198 ALL: 1,082 , The proposed approach is used to generate results and to accurately identify and detect them. The data augmentation technique involved is utilized to practice big datasets and thus it processes large Leukemia images. The features from Leukemia datasets are extracted by using our proposed HCNN and further the attention layer in the HCNN is exploited to fuse the extracted features. The softmax layer of HCNN acts as a classifier and therefore it classifies the leukemia dataset into several subtypes. Furthermore, the accuracy of classification is optimized by utilizing Interactive autodidactic school optimization techniques. Finally, the optimized outcomes are sent to the medical institution/hospital via an IoMT platform for further processing. Based on the results retrieved, the physician/doctor provides a diagnosis to the patients..

Techniques: Biomarker Discovery, Derivative Assay, Microscopy, Staining, Diagnostic Assay, Control, Generated

Journal: Frontiers in Public Health

Article Title: A framework to distinguish healthy/cancer renal CT images using the fused deep features

doi: 10.3389/fpubh.2023.1109236

Figure Lengend Snippet: Classification results achieved for raw renal CT slice with a SoftMax classifier.

Article Snippet: DenseNet121 , SoftMax , 129 , 5 , 127 , 7 , 95.5224 , 94.8529 , 96.2687 , 94.7761 , 95.5556 , 91.0549.

Techniques:

Journal: Frontiers in Public Health

Article Title: A framework to distinguish healthy/cancer renal CT images using the fused deep features

doi: 10.3389/fpubh.2023.1109236

Figure Lengend Snippet: Classification results achieved for processed renal CT slice with a SoftMax classifier.

Article Snippet: DenseNet121 , SoftMax , 129 , 5 , 127 , 7 , 95.5224 , 94.8529 , 96.2687 , 94.7761 , 95.5556 , 91.0549.

Techniques:

Journal: Frontiers in Public Health

Article Title: A framework to distinguish healthy/cancer renal CT images using the fused deep features

doi: 10.3389/fpubh.2023.1109236

Figure Lengend Snippet: Overall results achieved with the proposed framework for individual and fused features.

Article Snippet: DenseNet121 , SoftMax , 129 , 5 , 127 , 7 , 95.5224 , 94.8529 , 96.2687 , 94.7761 , 95.5556 , 91.0549.

Techniques:

Journal: Frontiers in Public Health

Article Title: A framework to distinguish healthy/cancer renal CT images using the fused deep features

doi: 10.3389/fpubh.2023.1109236

Figure Lengend Snippet: Spider plot achieved using the results of . (A) VGG19. (B) DenseNet121. (C) Fused deep features (VGG+DN).

Article Snippet: DenseNet121 , SoftMax , 129 , 5 , 127 , 7 , 95.5224 , 94.8529 , 96.2687 , 94.7761 , 95.5556 , 91.0549.

Techniques: