Abstract 6250: Clinical significance of co localization index in rectal cancer: a deep learning approach to cell interaction analysis
Notice bibliographique
Résumé
Abstract Background: Tumor tissue is not just a cluster of cancer cells, but an organized ecosystem in which various cells, extracellular matrices, and various types of liquid factors interact with each other. There are high technical barriers to processing this complex ecosystem in large quantities and quantitatively. Therefore, we have established a histological analysis method using deep-learning-based imaging cytometry (DL-IC) (Abe T, Yamashita K, et al. Anticancer Res. 43:3755 2023). Identification of the constituent cells of tumor tissue and spatial understanding of the tumor immune microenvironment using DL-IC are important for predicting cancer prognosis and treatment efficacy. In this study, we propose a co-localization index (CLI) as a new indicator for analyzing the tumor immune microenvironment using AI. Purpose: We investigated the impact of the co-localization of cytotoxic lymphocytes and cancer cells on the prognosis of rectal cancer after surgery. Subjectsand Methods: Forty rectal cancer surgical specimens were analyzed using the DL-IC system Cu-Cyto. A bit pattern kernel filtering algorithm was implemented to prevent duplicate cell counting, and its performance on immunohistochemistry (IHC) specimens was evaluated. The accuracy was compared between analyses with and without the algorithm. Cell-cell interactions were quantified using CLI, and the usefulness of CLI as a prognostic indicator was evaluated, particularly focusing on interactions between cancer cells and CD8+ T cells. Additionally, a comprehensive analysis of CLI was conducted using combinations of multiple cell types. Results: In the training process, where the data size of the training data was scaled, the F1 scores for adenocarcinoma cells, lymphocytes, and CD8+ T cells were 0.589, 0.889, and 0.911, respectively, at a cell number of 1013. The introduction of a bit pattern kernel filtering algorithm improved the accuracy of the determination of each cell type. The 5-year disease-free survival rate was significantly prolonged in the high CLI group between cancer cells and CD8+ T cells (P=0.041), and multivariate analysis also showed that it was an independent prognostic factor. On the other hand, there was no significant difference in the 5-year overall survival rate (P=0.41). Furthermore, it was found that the CLI of the three-way interaction between cancer cells, macrophages, and CD8+ T cells was also an independent prognostic factor. Conclusion: Cu-Cyto has achieved highly accurate cell determination by introducing a bit pattern kernel filtering algorithm. CLI has shown correlation with patient prognosis as an objective and reproducible quantitative evaluation method for cell-cell interactions. In the future, it will be necessary to elucidate the oncological and biological significance of the combinations of each cell type. Citation Format: Kimihiro Yamashita, Tomoki Abe, Toru Nagasaka, Masayuki Ando, Takao Tsuneki, Yukari Adachi, Takaaki Tachibana, Hiroki Kagiyama, Tomoaki Aoki, Yasufumi Koterazawa, Ryuichiro Sawada, Hitoshi Harada, Yasunori Otowa, Naoki Urakawa, Hironobu Goto, Hiroshi Hasegawa, Shingo Kanaji, Takeru Matsuda, Ryohei Sasaki, Yoshihiro Kakeji. Clinical significance of co localization index in rectal cancer: a deep learning approach to cell interaction analysis [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2025; Part 1 (Regular Abstracts); 2025 Apr 25-30; Chicago, IL. Philadelphia (PA): AACR; Cancer Res 2025;85(8_Suppl_1):Abstract nr 6250.
Récupéré en direct depuis OpenAlex et désinversé. Les résumés ne sont pas conservés dans cette base de données : les index inversés représentent 8,6 Go des 9,3 Go de texte de la base, et le serveur dispose de 13 Go libres.
Comment cette classification a été obtenuedéplier
Prédiction machine sur la base complète
Imitation des enseignantsNi prévalence calibrée, ni vérité terrain. Validation humaine à venir. Le volet Gemma est une étiquette directe du modèle pour chaque travail de la base, lue sur la notice réduite au titre. Le volet Codex est un classifieur appris des 10 348 étiquettes directes de Codex et calibré sur les taux pondérés de l'échantillon; les champs sans appui suffisant ne portent aucun appel Codex. Le mode candidate est l'union des deux volets; le consensus est leur intersection. Ces sorties portent le statut machine_predicted_unvalidated et ne sont pas des étiquettes humaines.
Scores du classifieur distillé par catégorie (deux têtes)
| Catégorie | Codex | Gemma |
|---|---|---|
| Métarecherche | 0,001 | 0,003 |
| Méta-épidémiologie (sens strict) | 0,000 | 0,000 |
| Méta-épidémiologie (sens large) | 0,000 | 0,000 |
| Bibliométrie | 0,002 | 0,001 |
| Études des sciences et des technologies | 0,000 | 0,000 |
| Communication savante | 0,001 | 0,000 |
| Science ouverte | 0,000 | 0,001 |
| Intégrité de la recherche | 0,001 | 0,001 |
| Charge utile insuffisante (le modèle a refusé de juger) | 0,002 | 0,000 |
Scores machine (provisoires)
Les deux têtes enseignantes du modèle étudiant, lues sur ce travail. Un score ordonne la base pour la relecture; il n'affirme jamais une catégorie, et le statut de validation accompagne chaque rangée tel quel.
Scores de référence d'un modèle non mature (critères de maturité non atteints, 7 itérations). Un score ordonne; il n'affirme jamais une catégorie.
score_only:v0-immature-baseline · tel quel depuis la passe de notation : score_only signifie que le nombre peut ordonner les travaux, et qu'aucune étiquette de catégorie n'en découleClassification
machine, non validéePrédiction automatique; un appel candidat d’une seule source (Gemma direct ou Codex distillé), pas un consensus.
Le détail, modèle par modèle et score par score, se trouve en fin de page sous « Comment cette classification a été obtenue ».