Comprehend Medical Samples
Sample code to start using Amazon Comprehend with C# and Dotnet Core Console application. This sample uses test files in png and jpg formats that are read by Amazon Textract for text detection. The detected text is then passed to Amazon Comprehend that detects Entities and PHI. The detected Entities and PHI are then printed on screen.
Amazon Comprehend Medical Samples for .NET Core
.NET Core Samples for Amazon Comprehend Medical. Amazon Comprehend Medical is a natural language processing service that makes it easy to use machine learning to extract relevant medical information from unstructured text. Using Amazon Comprehend Medical, you can quickly and accurately gather information, such as medical condition, medication, dosage, strength, and frequency from a variety of sources like doctors’ notes, clinical trial reports, and patient health records.
This sample uses a two step process:
- It uses Amazon Textract AWS .NET SDK to detect text from a PNG file (sample.png) located in test-files/ folder. You can use your own PNG or JPG file. Amazon Textract can also read from PDF but that is out of scope for this sample, and
- Amazon Comprehend Medical AWS .NET SDK is then used to analyze the detected text to detect entities like attributes, categories, and traits, and detect PHI information
Prerequisites
To run the code
Browse the folder where you have cloned/downloaded the project then run the following
dotnet run
Dependencies
The project file (.csproj) of this sample lists all the dependencies
<ItemGroup>
<PackageReference Include="AWSSDK.ComprehendMedical" Version="3.3.100.32" />
<PackageReference Include="AWSSDK.Extensions.NETCore.Setup" Version="3.3.100.1" />
<PackageReference Include="AWSSDK.Textract" Version="3.3.101.24" />
<PackageReference Include="Microsoft.Extensions.Configuration" Version="2.2.0" />
<PackageReference Include="Microsoft.Extensions.Configuration.Json" Version="2.2.0" />
</ItemGroup>
Example output
==Non-PHI==
12 entities found
Attributes
Categories
PROTECTED_HEALTH_INFORMATION:0.9932775:Unicorn Smith
Traits
Attributes
Categories
PROTECTED_HEALTH_INFORMATION:0.9966829:Unicorn Smith
Traits
Attributes
Categories
PROTECTED_HEALTH_INFORMATION:0.9953083:Unicorn Smith
Traits
Attributes
Categories
PROTECTED_HEALTH_INFORMATION:0.9998387:January 32, 1901
Traits
Attributes
Categories
PROTECTED_HEALTH_INFORMATION:0.76162:Captain Marvel
Traits
Attributes
Categories
PROTECTED_HEALTH_INFORMATION:0.2977315:Stark Towers
Traits
Attributes
Categories
MEDICAL_CONDITION:0.9962941:Nausea
Traits
SYMPTOM:0.8347825
Attributes
Categories
MEDICAL_CONDITION:0.9984876:vomiting
Traits
SYMPTOM:0.8194339
Attributes
Categories
MEDICAL_CONDITION:0.9421097:high fever
Traits
SYMPTOM:0.7858275
Attributes
Categories
MEDICAL_CONDITION:0.3872434:Suicide
Traits
Attributes
DOSAGE:0.9819972:200 mg
FREQUENCY:0.9816223:3 times a day
DURATION:0.9941246:10 days
Categories
MEDICATION:0.9929543:Ibuprofen
Traits
Attributes
DOSAGE:0.9688048:150 mg
FREQUENCY:0.9989164:twice daily
Categories
MEDICATION:0.9990332:Ranitidine
Traits
==PHI==
6 entities found
Attributes
Categories
PROTECTED_HEALTH_INFORMATION:0.9932775:Unicorn Smith
Traits
Attributes
Categories
PROTECTED_HEALTH_INFORMATION:0.9966829:Unicorn Smith
Traits
Attributes
Categories
PROTECTED_HEALTH_INFORMATION:0.9953083:Unicorn Smith
Traits
Attributes
Categories
PROTECTED_HEALTH_INFORMATION:0.9998387:January 32, 1901
Traits
Attributes
Categories
PROTECTED_HEALTH_INFORMATION:0.76162:Captain Marvel
Traits
Attributes
Categories
PROTECTED_HEALTH_INFORMATION:0.2977315:Stark Towers
Traits
License Summary
This sample code is made available under the MIT-0 license. See the LICENSE file.
React
Typescript
Django
Laravel
Facebook
Microsoft
Google
Alibaba
D3
Tencent