HBV-Liver-Severity: A Clinical Dataset for Liver Damage Prediction among Hepatitis B Patients in Bangladesh

Description

HBV-Liver-Severity is a well-structured, high quality, balanced clinical dataset particularly designed to satisfy the important need around the world to predict the level of liver damage in patients diagnosed with Hepatitis B (HBV). The data is carefully extracted from the official Blood Supply Register of National Gastro-Liver Institute & Hospital (NGIH), Dhaka, Bangladesh from the clinical records up to April 2025. This set of data also includes details of the demographic data and relevant serologic and viral markers and offers a detailed picture of the liver pathology in people with hepatitis B infection. It has been designed to complement the collection of clinical data with the latest AI diagnostics, offering a unique offering that can add value to patient monitoring, risk stratification, and clinical decision-making in the unique healthcare context of Bangladesh. Preprocessed Dataset Specifications: Data Type: The data is a balanced type data having 1050 samples. Target Classes: There are three classes of liver health – Mild Damage, Moderate Damage and Severe Damage – with a sample size of 350 for each class. Key Features (Description): Age: This is the patient's chronological age, and is a very important part of the evaluation of liver disease status. Gender: Sex of the patient, to look at any possible difference in the susceptibility of liver damage. Blood Group: Check with patient's blood group (O+, O-, B+) and see if there is any correlation with liver health. Total Anti-HBc: Total antibody to Hepatitis B core antigen (HBCAg) is a test that detects whether someone has been exposed to Hepatitis B in the past or recently. Serological marker IgM Anti-HBc: it is used to diagnose recent or acute Hepatitis B viral infection. Viral high infectivity in a Hepatitis B person will be a very significant marker—HBeAg. Anti-HBe: This antibody means that the infection is in a less virally active state, typically a better liver status. Viral coinfections: This diagnostic feature is used to identify the Hepatitis C virus, which is used to differentiate viral coinfections. HIV: this label will show if the patient is infected with the human immunodeficiency virus or not; it is important to understand the effect of HIV on liver pathology. Damage_Severity is the "target" feature that specifies liver's health status of Mild, Moderate or Severe Damage. Objective: The main purpose of this dataset is to enable developing predictive Artificial Intelligence (AI) models that could be beneficial to healthcare workers to assess the progression of liver health in patients suffering from Hepatitis B, leading to more accurate diagnosis and care of Hepatitis B patients in the healthcare system in Bangladesh.

Institutions

• Daffodil International University

  Dhaka Division, Dhaka

Categories

Licence