ABC: Artificial Intelligence for Bladder Cancer grading system
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
Bladder cancer tissue grading, which assigns a numerical grade reflecting how aggressive a tumor looks under a microscope, is essential to determine the proper course of treatment, design a therapeutic plan and determine prognosis. The major problem is that there are considerable and clinically relevant variations in grading by pathologists – as they are humans with different opinions and experience – including in bladder cancer. This work presents a solution, i.e., Artificial Intelligence for Bladder Cancer grading (ABC) system, that is developed based on deep neural network architectures to provide a more reliable and accurate diagnosis for patients affected by this deadly disease and ultimately improve management and clinical outcomes. Whole Slide Images (WSI) are split up into equally-sized square tiles and annotated to build a training dataset. ABC introduces a new grading system concept that can provide a percentage distribution of each different grade in a specific tumor, unlike the current numerical grade value between 1 and 3 based on the general impression of the pathologist. This new approach aims to provide a more granular grading of bladder cancer tissues and better capture tumor grade heterogeneity. This new concept may offer a more precise prognosis and optimize management in the future. The ABC learning model is fully configurable, and any deep architecture model can be trained and used by ABC. Some trained models developed by ABC have shown high accuracy and consistency in grading and intra-observer variability. The combination of a loosely coupled architecture and fully integrated tiles’ utilization makes ABC a universal, scalable, and versatile system that could be configured and deployed worldwide.
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 | 0.000 | 0.000 |
| Meta-epidemiology (narrow) | 0.000 | 0.000 |
| Meta-epidemiology (broad) | 0.000 | 0.000 |
| Bibliometrics | 0.000 | 0.000 |
| Science and technology studies | 0.001 | 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 it