AI-KODA score application for cleanliness assessment in video capsule endoscopy frames
Bibliographic record
Abstract
Background Currently, there is no automated method for assessing cleanliness in video capsule endoscopy (VCE). Our objectives were to automate the process of evaluating and collecting medical scores of VCE frames according to the existing KOrea-CanaDA (KODA) scoring system by developing an easy-to-use mobile application called artificial intelligence-KODA (AI-KODA) score, as well as to determine the inter-rater and intra-rater reliability of the KODA score among three readers for prospective AI applications, and check the efficacy of the application.Method From the 28 patient capsule videos considered, 1539 sequential frames were selected at five-minute intervals, and 634 random frames were selected at random intervals during small bowel transit. The frames were processed and shifted to AI-KODA. Three readers (gastroenterology fellows), who had been trained in reading VCE, rated 2173 frames in duplicate four weeks apart after completing the training module on AI-KODA. The scores were saved automatically in real time. Reliability was assessed for each video using estimate of intra-class correlation coefficients (ICCs). Then, the AI dataset was developed using the frames and their respective scores, and it was subjected to automatic classification of the scores via the random forest and the k-nearest neighbors classifiers.Results For sequential frames, ICCs for inter-rater variability were ‘excellent’ to ‘good’ among the three readers, while ICCs for intra-rater variability were ‘good’ to ‘moderate’. For random frames, ICCs for inter-rater and intra-rater variability were ‘excellent’ among the three readers. The overall accuracy achieved was up to 61% for the random forest classifier and 62.38% for the k-nearest neighbors classifier.Conclusions AI-KODA automates the process of scoring VCE frames based on the existing KODA score. It saves time in cleanliness assessment and is user-friendly for research and clinical use. Comprehensive benchmarking of the AI dataset is in process.
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 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 | 0.000 | 0.000 |
| Meta-epidemiology (narrow) | 0.000 | 0.000 |
| Meta-epidemiology (broad) | 0.000 | 0.000 |
| Bibliometrics | 0.000 | 0.001 |
| Science and technology studies | 0.000 | 0.000 |
| Scholarly communication | 0.000 | 0.000 |
| Open science | 0.000 | 0.000 |
| Research integrity | 0.000 | 0.001 |
| 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 itClassification
machine, unvalidatedMachine predicted; a candidate call from one teacher head, 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".