Computer Vision Data Extraction for Malaria Logbooks

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Healthcare workers in resource-limited settings face challenges in digitizing malaria-related data, as they often have to stop providing medical services to manually encode information, due to difficulties in using mobile phones for data entry, which can lead to inefficiencies and reduced patient care.

Innovation Solution

A system that allows healthcare workers to take photos of logbook pages, using computer vision techniques to extract data from digital images, enabling the use of a supervised or unsupervised framework for document type classification, such as the Retrieval, Learning, and Matching (RLM) algorithm, to accurately identify and extract data from images of forms, even with handwritten or machine-printed characters.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Loss of information

If healthcare workers manually encode malaria-related data from logbooks using mobile phones, then data can be digitized and stored, but healthcare workers must stop providing medical services and spend significant time on data entry

Engineering Contradiction:
Improvedata digitizationVSAvoidtime for data entry
Core Design Contradiction:
Loss of informationVSLoss of time

Solution Approach 1:

The system creates a digital copy of the logbook page image and uses computer vision algorithms to extract data from this copy, eliminating the need for manual typing while preserving all information

Inventive Principle:
Principle #26Copying

Solution Approach 2:

The patent replaces the mechanical process of manual data encoding with automated computer vision and optical character recognition systems that can read both printed and handwritten text

Inventive Principle:
Principle #28Mechanics substitution (Replace mechanical system)

2Extent of automation

If healthcare workers use mobile phones for data entry, then data can be transmitted digitally, but the complexity of encoding information creates barriers for workers with limited formal education

Engineering Contradiction:
Improvedigital data reportingVSAvoiddata encoding difficulty
Core Design Contradiction:
Extent of automationVSEase of operation

Solution Approach 1:

The system performs the complex data encoding and digitization tasks automatically without requiring healthcare workers to understand or perform the encoding process, making the system self-sufficient

Inventive Principle:
Principle #25Self-service

Solution Approach 2:

The patent introduces an intermediary system (computer vision software) that bridges the gap between the logbook and digital storage, handling the complex transformation without requiring user intervention

Inventive Principle:
Principle #24Intermediary (Mediator)

3Reliability

If healthcare workers stop providing medical services to digitize data, then accurate data tracking can be maintained, but patient care quality deteriorates

Engineering Contradiction:
Improvedata accuracyVSAvoidpatient care provision
Core Design Contradiction:
ReliabilityVSProductivity

Solution Approach 1:

The system extracts and stores data from logbook images before they are physically stored, creating a digital record in advance that maintains accuracy without requiring dedicated data entry time during patient care

Inventive Principle:
Principle #10Preliminary action

Data Source

PatentUS20240168993A1Analyzing content of digital images
Publication Date: 2024.05.23 DST TECHNOLOGIES INC
  • US20240168993A1 patent drawing
  • US20240168993A1 patent drawing
  • US20240168993A1 patent drawing

AI summary

Methods, apparatuses, and embodiments related to analyzing the content of digital images. A computer extracts multiple sets of visual features, which can be keypoints, based on an image of a selected object. Each of the multiple sets of visual features is extracted by a different visual feature extractor. The computer further extracts a visual word count vector based on the image of the selected object. An image query is executed based on the extracted visual features and the extracted visual word count vector to identify one or more candidate template objects of which the selected object may be an instance. When multiple candidate template objects are identified, a matching algorithm compares the selected object with the candidate template objects to determine a particular candidate template of which the selected object is an instance.