Impact of Introducing Artificial Intelligence on Colonoscopy: A Retrospective Study on Potential Benefits and Drawbacks
Bibliographic record
Abstract
BACKGROUND AND AIM: Computer-aided detection (CAD) can improve adenoma detection rates (ADRs); however, the impact of its introduction into real-world practice remains unclear. This study investigated the effect of CAD's introduction on colonoscopy in a hospital. METHODS: This retrospective study included 1314 patients who underwent colonoscopy between January and December 2023 at a single facility where CAD was introduced in three of four endoscopy rooms. ADR, polyp detection rate (PDR), and sessile serrated lesion detection rate (SSLDR) were first compared between patients who underwent colonoscopy without CAD before introduction to the facility (pre-intervention non-CAD group) and those who underwent colonoscopy with CAD after introduction (CAD group). Subsequently, cases without CAD were analyzed to evaluate endoscopists' performance by comparing the detection rates between the pre-intervention non-CAD group and patients who underwent colonoscopy without CAD after introduction (post-intervention non-CAD group). RESULTS: ADR (49.3% vs. 31.6%, p < 0.001) and PDR (57.9% vs. 39.8%, p < 0.001) were significantly higher in the CAD group than in the pre-intervention non-CAD group; SSLDR (4.4% vs. 2.8%, p = 0.14) was comparable between groups. ADR (31.6% vs. 13.5%, p < 0.001) and PDR (39.8% vs. 18.2%, p < 0.001) were significantly lower in the post-intervention non-CAD group than in the pre-intervention non-CAD group. CONCLUSIONS: The introduction of CAD-assisted colonoscopy significantly improved ADR and PDR. However, CAD reliance may lead to lapses in attention toward independent lesion detection by endoscopists. It is essential to consider how CAD should be utilized in clinical practice to maximize its benefits.
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) | 0.001 | 0.000 |
| Bibliometrics | 0.000 | 0.000 |
| Science and technology studies | 0.000 | 0.000 |
| Scholarly communication | 0.000 | 0.000 |
| Open science | 0.000 | 0.000 |
| Research integrity | 0.000 | 0.000 |
| 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 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".