Identifying Key Survival AI-Based Predictors in Breast Cancer for Indian Women: A Retrospective Cohort Analysis
摘要
Breast cancer is one of the most prevalent malignancies worldwide and a leading cause of cancer-related morbidity and mortality. In India, it has surpassed cervical cancer as the most common cancer in women, with rising incidence due to urbanization, lifestyle changes, and delayed diagnosis. Survival rates remain lower than in developed countries due to late-stage presentation, limited awareness, and unequal healthcare access. This study analyses demographic, clinical, and treatment-related factors influencing breast cancer survival outcomes. Key AI-based predictors—tumor stage, receptor status, and treatment adherence—are examined, along with socioeconomic and lifestyle factors affecting prognosis. Evidence-based recommendations for improving early detection and survival in Indian women are proposed. A retrospective cohort study (2021–2023) examined 800 breast cancer cases from tertiary healthcare centers in India. Tumor stage, receptor status, treatment adherence, and survival outcomes were analyzed. Kaplan-Meier survival analysis assessed overall survival (OS) and progression-free survival (PFS). Cox proportional hazards regression identified independent survival predictors. Ethical approval was obtained, ensuring data confidentiality. Late-stage diagnosis remains a challenge, with only 15% diagnosed at Stage I. Five-year overall survival (OS) was highest in Stage I (~90%) and lowest in Stage III (~55%). ER/PR+ tumors had the best prognosis (PFS ~65%), while triple-negative breast cancer (TNBC) had the worst (~40%). Improved screening, equitable healthcare access, and novel therapies, particularly for aggressive subtypes, like TNBC, are crucial for improving breast cancer outcomes in India.