Artificial intelligence and identifying prognostic patterns in upper tract urothelical carcinoma (UTUC).
03 medical and health sciences
0302 clinical medicine
DOI:
10.1200/jco.2024.42.16_suppl.e16616
Publication Date:
2024-06-06T17:50:54Z
AUTHORS (7)
ABSTRACT
e16616 Background: Artificial intelligence tools (AI) are increasingly used in all scientific fields. UTUC is rare form of kidney cancer and prognostic factors seem unclear. We sought to understand if AI could help identify parameters for overall survival aid the development a score. Methods: n=231 consecutive patients who underwent surgery between 2005 2020 were included. De-personalized data included age, gender, ECOG performance status, smoking habits, baseline creatinine platelets, histology, laterality, pathological clinical T-stage, grading, cis, TNM stage, necrosis, lymphovascular invasion, vascular resection margin, tumor size, duration surgery, surgeon, surgical approach, blood loss, complications (graded by Clavien-Dindo), complete removal ureter including bladder cuff, until recurrence death reason death. Several publicly accessible (chatGPT, Tableau, Julius AI, Microsoft Power BI, Polymer) asked first develop score RCC outcome. Statistical analysis was done using chi-square test, cox regression Kaplan-Meier estimation. analyze set based on proportional log rank testing performed estimate accuracy. Results: Median age 70.9 years (43.7-81.8). 61% (n=160) male. stats 0 42.1% (n=82), 1 27.2% (n=52). 53.6% (n=105) non-smoker 24.5% (n=48) smoker 21.9% (n=43) former smokers. 10.3% (n=12) had synchronous cancer. 60.4% (n=131) suffered from gross hematuria 19.7% (n=42) flank pain. follow-up 34.5 months (0-174). (OS) 58.9 (95% CI 46.2-71.5). Bellmunt risk factors, BMI creatine CRP levels correlated significantly with adverse (p>0.001). Tumors localized ureter, larger than 5cm, carcinoma situ high grade significant worse OS. Resecting cuff improved OS 86.7 65-108.4) compared 39.9 27.9-51.9, p=0.012) without. revealed nodal grade, expertise resecting be did not conventional relevant scoring system predict PFS. Conclusions: can outcomes factors. Human adjustment still necessary. Interestingly, that influenced should considered future evaluations models besides characteristics like grading status can’t medically influenced.
SUPPLEMENTAL MATERIAL
Coming soon ....
REFERENCES (0)
CITATIONS (1)
EXTERNAL LINKS
PlumX Metrics
RECOMMENDATIONS
FAIR ASSESSMENT
Coming soon ....
JUPYTER LAB
Coming soon ....