Dermoscopic Image Classification for Skin Cancer Diagnosis Using Deep Learning and Cuckoo Search With 3D Shearlet
Bibliographic record
Abstract
Melanoma, the most serious kind of skin cancer, is formed by a mutation in melanocytes.An early diagnosis is very important to reduce mortality.The proposed system categorizes dermoscopic images for identifying skin malignancies using Deep Learning and Cuckoo Search (DLCS) in conjunction with 3D Shearlet Transform (3DST).There are four different modules that make up the DLCS system.These modules include preprocessing, representation of dermoscopic images, selection of directional sub-bands and features, and classification.Using a straightforward median filtering strategy, the initial step eliminates the undesirable information which degrades the system's performance.These details include noise and hair in skin images.The pre-processed image is decomposed using 3D ST during the feature extraction step to retrieve the textural characteristics at varying scales and directions.The DLCS technique is used to choose a certain proportion of features, and then, a straightforward DL architecture with ten hidden layers is used to create a classification system for the dermoscopic image.Experimental results on PH 2 and ISIC databases show that the DLCS-3DST system's performances are affected by the features from different Levels (L) and Directions (D).Training the classifier using the selected features from 3L-8D provides the highest accuracy of 99.22% for PH 2 database and 99.39% for ISIC database.It is also observed that when dermoscopic images are decomposed by 4L with 32D, there is an increase in redundant information, which negatively impacts the performance of the classifier.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
How this classification was reachedexpand
Full frame machine prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. The Gemma side is a direct model label for every work in the frame, read from the title-only record. The Codex side is a classifier learned from the 10,348 direct Codex labels and calibrated to design-weighted sample rates; fields without enough sample support carry no Codex call. Candidate is the union of the two sides; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels.
Distilled classifier scores by category (both heads)
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.000 | 0.001 |
| Meta-epidemiology (narrow) | 0.001 | 0.000 |
| Meta-epidemiology (broad) | 0.001 | 0.001 |
| Bibliometrics | 0.002 | 0.001 |
| Science and technology studies | 0.000 | 0.000 |
| Scholarly communication | 0.001 | 0.001 |
| Open science | 0.001 | 0.001 |
| Research integrity | 0.001 | 0.001 |
| Insufficient payload (model declined to judge) | 0.002 | 0.001 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedMachine predicted; a candidate call from one source (direct Gemma or distilled Codex), not a consensus.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".