← Search

Data · dataset · 2026

DMCH-PneumoX: Pneumonia Chest X-ray Annotated Dataset

Listed in ScienceDB

The DMCH Pneumonia Chest X-ray Dataset (DMCH-PneumoX) is a curated collection of clinical chest radiographs designed to support research on pneumonia detection and thoracic disease analysis using computer vision and machine learning techniques.

Description

The dataset was compiled from routine diagnostic imaging performed at Dhaka Medical College and Hospital (DMCH), a major tertiary healthcare facility in Bangladesh. All radiographs were originally acquired in Digital Imaging and Communications in Medicine (DICOM) format and were subsequently converted to JPG/JPEG while preserving the grayscale characteristics typical of clinical chest radiography.The current release contains 6,000 chest radiographs, representing a broad range of patient presentations encountered in routine clinical practice.

Among these images, 426 radiographs contain expert annotations, while 5,574 images remain unannotated and can serve as additional data for semi-supervised or unsupervised learning approaches. The annotated subset includes 552 bounding-box annotations identifying pathological regions associated with pneumonia and pneumonia-related complications.Annotations were created using the Computer Vision Annotation Tool (CVAT) following a standardized protocol developed in consultation with radiology experts.

Read the rest (2 more)

The annotations focus on two primary categories of pathological findings: pneumonia and pneumonia with associated thoracic complications. Within the annotated subset, 314 bounding boxes correspond to pneumonia findings, while 238 bounding boxes correspond to pneumonia accompanied by related thoracic abnormalities, including radiographic patterns such as consolidation, pleural effusion, infiltration, atelectasis, and bronchial wall thickening.

At the image level, 253 radiographs contain pneumonia annotations and 175 radiographs contain pneumonia-with-complications annotations. A small number of images (two radiographs) contain annotations belonging to both categories.

Links

Where it is published

Catalogue records · 1

Topics

Inferred from text
Clinical sciences 70% · Disease 75% · Image 75% · Imaging 75%
Provenance · 1 source records, 15 field assertions
SourceKeyLast seenRaw
ScienceDB10.57760/sciencedb.3043612 d agoJSON v1
FieldAssertionExtractorEvidence
access_levelsource · scidb cnconnector:scidb_cn@1.0.0
concepts[disease].local:disease:diseaseenrichment · scidb cnkeyword-concept-rules@1.0.0title+description (75%)
concepts[field].anzsrc:group:3202enrichment · scidb cntaxonomy-embedding@1.0.0title+keywords+description (70%)
concepts[field].local:field:earth-environmentalmapping · scidb cnconnector:scidb_cn@1.0.0
concepts[field].local:field:engineeringmapping · scidb cnconnector:scidb_cn@1.0.0
concepts[field].local:field:humanitiesmapping · scidb cnconnector:scidb_cn@1.0.0
concepts[field].local:field:life-sciencesmapping · scidb cnconnector:scidb_cn@1.0.0
concepts[field].local:field:medicine-healthmapping · scidb cnconnector:scidb_cn@1.0.0
concepts[field].local:field:social-sciencemapping · scidb cnconnector:scidb_cn@1.0.0
concepts[modality].local:modality:imageenrichment · scidb cnkeyword-concept-rules@1.0.0title+description (75%)
concepts[modality].local:modality:imagingenrichment · scidb cnkeyword-concept-rules@1.0.0title+description (75%)
descriptionsource · scidb cnconnector:scidb_cn@1.0.0/metadata/dc/description
license_textsource · scidb cnconnector:scidb_cn@1.0.0
publication_datesource · scidb cnconnector:scidb_cn@1.0.0
titlesource · scidb cnconnector:scidb_cn@1.0.0/metadata/dc/title