Tuberculosis Dataset for Intelligent and Adaptive Medical Diagnostic System

Tuberculosis is a communicable chronic disease and one of the top ten causes of death worldwide according to World Health Organization (WHO). With availability of clean and well encoded clinical data from tuberculosis patients, artificial intelligence and machine learning algorithms would be able to transform the management of tuberculosis patients through intelligent prediction and intervention. This dataset contains four hundred and thirty (430) clinical data from patients with tuberculosis at Tuberculosis and Leprosy Hospital, Eku, Delta State, Nigeria. The dataset was gathered through validated and structured questionnaire administered using random sampling after obtaining the patients' consent. The collated dataset was pre-processed and encoded with variables (features) for prediction which include cough, night sweat, breathing difficulty, fever, chest pain, sputum, immune suppression, loss of pleasure, chill, lack of concentration, irritation, loss of appetite, loss of energy, lymph node enlargement, systolic blood pressure and BMI. Prediction of tuberculosis based on the clinical data from patients' features would play an essential role in diagnosis, intervention and management of tuberculosis patient.

结核病作为一种传染性慢性疾病，位居世界卫生组织（WHO）公布的十大死因之列。凭借来自结核病患者的高质量、编码规范的临床数据，人工智能与机器学习算法有望通过智能预测与干预手段，革新结核病患者的管理方式。本数据集收录了来自尼日利亚 Delta 州 Eku 结核病和麻风病医院的430份结核病患者临床数据。数据采集过程遵循了经过验证的结构化问卷，并采用随机抽样方法在获得患者同意后进行。所收集的数据集经过预处理并编码，以预测变量（特征）为依据，包括咳嗽、夜间出汗、呼吸困难、发热、胸痛、痰液、免疫功能低下、愉悦感丧失、寒战、注意力不集中、瘙痒、食欲减退、能量下降、淋巴结肿大、收缩压和BMI等。基于患者特征的临床数据对结核病的预测，在结核病的诊断、干预和管理中扮演着至关重要的角色。