Derin Öğrenme ile Alzheimer Hastalığı Teşhisi İçin Model Önerisi
Why this work is in the frame
A frame that forgets how it found something cannot be audited. These are the routes that admitted this work.
Bibliographic record
Abstract
Alzheimer's disease is one of the greatest health problems of our time. Since there is no cure, it is necessary to diagnose the disease in the early stages and to apply preventive treatments. However, early diagnosis of the disease is very difficult, so most people can be diagnosed after significant and irreversible effects occur. Various studies are carried out by researchers around the world for the early diagnosis of the disease. Deep learning has recently gained importance in the early diagnosis of Alzheimer's disease. With the use of models created using deep learning, the success of early diagnosis has reached high levels. In this study, the stages of Alzheimer's disease and the changes that occur were examined. A literature review was conducted for various techniques used in the diagnosis of Alzheimer's and the use of imaging techniques in the early diagnosis of Alzheimer's was investigated. Due to its widespread use, the MRI technique has been emphasized, and mostly studies using MRI have been examined. Concepts used in deep learning are explained, innovations and results are presented. The architectures used in deep learning and the innovations they bring to this field are revealed, and deep learning models that have been created and tested in current studies are examined. The innovations and success rates brought by various studies have been revealed. Efforts have been made to develop a fast, efficient and successful model that provides ease of use. For this, the scheduler structure, MONAI framework, Data loader structure and various techniques are presented with simple use. Also, the model is optimized to run smoothly on Google Colab. In addition, the tools in the FSL library, which are very important in preprocessing, were studied and optimal parameters were found for the "Bias field and Neck Clean Up", "Standard Brain Extraction Using BET2" and "Robust Brain Center Estimation" tools. By using this library, optimal brain images can be obtained for any model. The DenseNet121 model was used as a basis in the model and it was presented in a structure that can be easily changed. The model can directly use 3D MR images, thus preventing the loss of various spatial information.
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.
Full frame distilled prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. Learned from the 10,348 direct Codex labels and 10,348 direct Gemma labels. Candidate is the union of thresholded teacher heads; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels or direct frontier model labels.
Codex and Gemma teacher scores by category
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.002 | 0.000 |
| Meta-epidemiology (narrow) | 0.000 | 0.000 |
| Meta-epidemiology (broad) | 0.000 | 0.000 |
| Bibliometrics | 0.001 | 0.002 |
| Science and technology studies | 0.001 | 0.001 |
| Scholarly communication | 0.000 | 0.000 |
| Open science | 0.001 | 0.000 |
| Research integrity | 0.000 | 0.000 |
| Insufficient payload (model declined to judge) | 0.000 | 0.000 |
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 it