AU2020100067A4

AU2020100067A4 - A method to identify key medical device product information from device labels with optical character recognition

Info

Publication number: AU2020100067A4
Application number: AU2020100067A
Authority: AU
Inventors: Rob Kidd
Original assignee: Opologic Pty Ltd
Current assignee: Opologic Pty Ltd
Priority date: 2020-01-14
Filing date: 2020-01-14
Publication date: 2020-02-20
Anticipated expiration: 2028-01-14

Abstract

Abstract An OCR method for identifying the core product information in a medical device document or its product label. Specifically, the product code and the lot code using text markers to identify and extract associated data, preferably captured with a mobile electronic device. FIG.1 Typical product code identifiers FIG.2 Typical lot code identifiers REF REF LOT LOT bordered without border bordered without border FIG. 3 Identifier is often a preface of the associated data to be extracted. Identifier data region OCR is custom programmed to use these key markers to identify the region where the associated data can be read and extracted by searching corresponding region. FIG. 4 Example of product code configuration REF 111-222-333 Identifier product code example (alpha-numeric) FIG. 5 Example of lot code configuration Identifier lot code example (alpha-numeric)

Description

A METHOD TO IDENTIFY KEY MEDICAL DEVICE PRODUCT INFORMATION FROM DEVICE LABELS WITH OPTICAL CHARACTER RECOGNITION

Field

The present disclosure relates to techniques for extracting data from product documents and labels.

More specifically, the present disclosure relates to techniques for facilitating optical character recognition (OCR) by identifying one or more regions on labels or printed documents to perform the OCR and extract key product information.

Background

Data processing and exchange are essential for a variety of business needs. For example, manufacturers and distributors use product data to obtain and build reports regarding inventory, sales, and forecasting.

However, variations in the layout or design of documents may preclude efficient extraction and/or transfer of data from packaging, labels and supporting product documentation.

For example, a product label and it's supporting documentation from one manufacturer or distributor may include a single type of data format to facilitate efficient data extraction, such as a 1D or 2D barcode. Yet this same data on an alternate manufacturer or distributors documentation may be In a different format entirely, or often at best the same data type but with a different decoding configuration.

As a result, a device performing barcode scanning across multiple products from varying distributors and manufacturers may have difficulty extracting information from each given document or product label. Instead, the recipient may have to manually enter data into an application and bypass a more efficient scanning process altogether.

Within the medical industry this valuable product data typically includes one or both of the following unique identifiers, REF and LOT.

REF refers to the Product code, a unique identifier, assigned to each manufactured product which is ready to be marketed or for sale.

LOT refers to the lot code, a unique identifier assigned to all products in each production batch. A unique combination of letters and/or numbers, by which a unit of product can be traced back to a specific production run.

Summary

One embodiment presented herein includes an OCR method for Identifying the core product information in a medical device document or its product label. Specifically, the product code and the lot code.

The method includes extracting spatial attributes for a region and its corresponding value from the bounding coordinates of the region.

Through custom development of the OCR data extraction code, It is possible to identify these key markers (REF & LOT) commonly utilised across a majority of medical device documentation and use these markers to extract the associated data.

2020100067 14 Jan 2020

Description

With common access to portable devices with large screen sizes, in-built cameras, and ever-increasing computational capabilities, the smart phone or tablet is a popular choice for product data capture.

Therefore, a mobile device, such as a mobile telephone or computing tablet could access the server over network using a mobile web browser or local application executed on the mobile device, to run the computational functionality of the custom OCR code.

The OCR is custom coded to specifically identify these key markers. Those can then be used to identify the regions of the label or document that contain the required data, and extract the required data from each document, specially the product code and the lot code.

These codes are a combination of letters, numbers and symbols, with an average length of 4-20 characters. The OCR can be custom coded to search and extract the complete string of information and store the extracted information under Its appropriate category either locally on device or in an associated database.

Description of Drawings

FIG. 1 illustrates the typical product code identifier format, showing it with or without a border. The software will be programmed to look for either format.

FIG. 2 illustrates the typical lot code identifier format, showing it with or without a border. The software will be programmed to look for either format

Both above illustrations specifically show theses markers to be REF and LOT uppercase characters, that are often bordered by a graphical box, that preface the reference number and LOT number of the product.

FIG. 3 illustrates how the identifier is often a preface to the required data and therefore the custom coded OCR will be programmed to search the expected data region and extract any/ail required data.

FIG. 4 illustrates an example operation performed as part of a run-time process used for identifying the required information from a product label or document, in accordance with aspects of the present disclosure for the product code.

FIG. 5 illustrates an example operation performed as part of a run-time process used for identifying the required information from a product label or document, in accordance with aspects of the present disclosure for the lot code.

Claims

Claims

1. An electronic device, utilising OCR is programmed to identify the product code on medical device documentation and/or its product labels, and to extract the related product code data.
2. An electronic device as in claim 1, wherein said device is programmed to identify the lot code on medical device documentation and/or product labels and extract the related lot code data.
3. An electronic device as in claim 1 & 2, wherein said device extracts and stores the product code and/or lot code in a database for future reference, via export or API.
4. An electronic device as in claim 1 & 2, wherein said device extracts and stores the product code and or lot code locally in device memory.