Ramalingam K, Yadalam PK, Ramani P, Krishna M, Hafedh S, Badnjević A, Cervino G, Minervini G. Light gradient boosting-based prediction of quality of life among oral cancer-treated patients.
BMC Oral Health 2024;
24:349. [PMID:
38504227 PMCID:
PMC10949789 DOI:
10.1186/s12903-024-04050-x]
[Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [Key Words] [MESH Headings] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Received: 01/07/2024] [Accepted: 02/19/2024] [Indexed: 03/21/2024] Open
Abstract
BACKGROUND AND INTRODUCTION
Statisticians rank oral and lip cancer sixth in global mortality at 10.2%. Mouth opening and swallowing are challenging. Hence, most oral cancer patients only report later stages. They worry about surviving cancer and receiving therapy. Oral cancer severely affects QOL. QOL is affected by risk factors, disease site, and treatment. Using oral cancer patient questionnaires, we use light gradient Boost Tree classifiers to predict life quality.
METHODS
DIAS records were used for 111 oral cancer patients. The European Organisation for Research and Treatment of Cancer's QLQ-C30 and QLQ-HN43 were used to document the findings. Anyone could enroll, regardless of gender or age. The IHEC/SDC/PhD/OPATH-1954/19/TH-001 Institutional Ethical Clearance Committee approved this work. After informed consent, patients received the EORTC QLQ-C30 and QLQ-HN43 questionnaires. Surveys were in Tamil and English. Overall, QOL ratings covered several domains. We obtained patient demographics, case history, and therapy information from our DIAS (Dental Information Archival Software). Enrolled patients were monitored for at least a year. After one year, the EORTC questionnaire was retaken, and scores were recorded. This prospective analytical exploratory study at Saveetha Dental College, Chennai, India, examined QOL at diagnosis and at least 12 months after primary therapy in patients with histopathologically diagnosed oral malignancies. We measured oral cancer patients' quality of life using data preprocessing, feature selection, and model construction. A confusion matrix was created using light gradient boosting to measure accuracy.
RESULTS
Light gradient boosting predicted cancer patients' quality of life with 96% accuracy and 0.20 log loss.
CONCLUSION
Oral surgeons and oncologists can improve planning and therapy with this prediction model.
Collapse