Utilizing LSTM and K-NN for Anatomical Localization of Tuberculosis: A Solution for Incomplete Data
Bibliographic record
Abstract
Tuberculosis (TB) is a prevalent lung disease that significantly contributes to mortality rates, with an estimated 98,000 fatalities observed in Indonesia alone. TB can be classified into two categories based on its anatomical location: pulmonary, when detected in lung parenchyma tissue, and extrapulmonary, when identified in organs outside the lungs. Current diagnostic procedures necessitate numerous laboratory tests and manual assessments, which are both time-consuming and susceptible to data incompleteness, thereby potentially influencing the diagnostic outcomes. This necessitates the development of a rapid and accurate classification system for the anatomical location of TB, which could aid medical professionals in diagnosis. In this study, we propose a novel classification system that utilizes the K-Nearest Neighbors (K-NN) algorithm to handle missing data, and the Synthetic Minority Over-sampling Technique (SMOTE) for data balancing. For the classification of pulmonary and extrapulmonary TB, the study employs the Long Short-Term Memory (LSTM) method, the performance of which is compared with other models, namely Na ve Bayes, Support Vector Machine (SVM), and Backpropagation. Although all four models demonstrated high levels of accuracy, the LSTM method outperformed the others, achieving 100% accuracy compared to Na ve Bayes (99.4%), SVM (99.3%), and Backpropagation (99.7%). These results were obtained after implementing imputation and class balancing stages, and optimizing LSTM features such as the tanh activation function, learning rate of 0.01, 100 LSTM units, and the ADAM optimizer. The proposed system thus presents an effective solution for the rapid and accurate classification of TB based on anatomical location.
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 | 0.001 | 0.003 |
| Meta-epidemiology (narrow) | 0.001 | 0.000 |
| Meta-epidemiology (broad) | 0.001 | 0.001 |
| Bibliometrics | 0.000 | 0.001 |
| Science and technology studies | 0.001 | 0.000 |
| Scholarly communication | 0.001 | 0.002 |
| Open science | 0.002 | 0.001 |
| Research integrity | 0.002 | 0.001 |
| Insufficient payload (model declined to judge) | 0.001 | 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 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".