LungXplain: Chunk - Level Modeling for Precise Detection and Localization of Respiratory Events
Bibliographic record
Abstract
The automation of detecting adventitious lung sounds, such as crackles and wheezes, could enhance the diagnosis and management of respiratory diseases. However, current methods often struggle with using broad, segment-level labels and lack explainability. In this work, we introduce an audio processing framework that performs fine-grained chunk-level classification to capture the temporal and spectral dynamics of respiratory events. Our approach processes the ICBHI 2017 dataset by segmenting each annotated respiratory event into shorter fixed-duration audio chunks, extracting relevant features, and classifying each chunk individually using an XGBoost model. These chunk predictions are then aggregated to infer the overall label of the original respiratory segment using rule-based strategies. The proposed system achieves an ICBHI score of 95.71 percent on the official test split, significantly outperforming prior state-of-the-art approaches. Additionally, we introduce a visual explanation mechanism that highlights the specific time intervals during which pathological sounds are detected, providing insight into the model's decisions.
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 | 0.002 |
| Meta-epidemiology (narrow) | 0.001 | 0.000 |
| Meta-epidemiology (broad) | 0.001 | 0.001 |
| Bibliometrics | 0.001 | 0.000 |
| Science and technology studies | 0.000 | 0.000 |
| Scholarly communication | 0.001 | 0.001 |
| Open science | 0.002 | 0.001 |
| Research integrity | 0.001 | 0.001 |
| Insufficient payload (model declined to judge) | 0.003 | 0.002 |
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".