The treatment dashboard reports using Amazon Bedrock, Langchain, and streamlit

In health care, immediate analyzing and interpreting important medical reports to both health care providers. While medical reports contain important information, usually remain under due to its complex nature and comprehensive analysis process. This complex reflects on many ways: The translation of the parameters and their relationships (such as different blood calculations), comparing the test results against the regular references, and the need to analyze health parameters over time. Dealing with this challenge, analyzing reports that show that health suppliers can promote their interactions with medical data for the implementation of the sample

In this regard, the Dashboard created represents the Amazon Bedrock Advanced Advanced Advanced Ai, and Stream's Processed Interface. By applying this technology, we have only created a system not only in stores and indicates medical reports, but actively helps translate environmental and strong sighting.

Looking for everything

The solution is a variety of languages ​​found in Amazon Bedrock, including Anthropic's Claudes series and Amazon Foundation Model. You can choose from the options such as Claude Opus 4.1, Claude 3.7 Sonnet, Amazon Nova Pro, and each one, are prepared for different performance work and energy needs. The selected model processes natural language questions by awareness of the medical context, making a detailed clarification of health data. Through this variable situation, you can measure things as accuracy, speed, and cost depending on your specific needs. This improvement of documents in Langchain processes, treating the return system and maintain the context of the discussion, making accurate and appropriate answers.

The remedy of the remedy database with medical reports maintained by the simzon easy-to-Amazon storage service (Amazon S3), which is made with Langchain's DEPY SYSTEM. When contacting the streamlit frontlend, your questions are analyzed by Amazon Bedrock, and Langchain keeps the context of the discussion and handles the restoration of the document. The system processes this information and shows the results in a visible form of view.

This looks, accidentally emphasizes, including the comparisons that clearly show comparisons with the real prices, bar charts parameters, and tracking lines change over time. The streamlit indicator binds everything together, providing real time interactions through AI program while in the control of the user session status and the history of chat history. This is the perfect option to ensure that medical professionals can get in quickly, analyze, and interpret their medical reports on natural language questions while they view visual data.

The following is a drawing of the construction of four layer:

• SERVICE SERVICE BACKGROUND: Web site's direction, dialog for chat interface, data for visual data
• To process the layer: Langchain documents, retrieve chain conversations, data parsing
• AI / ML layer: Amazon Bedrock, Amazon Bedrock Embeddings, In-Memory Vector Store
• Last Background: Amazon S3 of Medical Reports, Dialogum Question Memory

Requirements

Before sending medical analysis reports Dashboard, you need:

We will use the visual Python (VEW) environment for this project to provide clean and one-person. The visual areas help to avoid package conflicts between projects and make the administration of dependent on the exact purpose. While using the Built-in Venv, you can use a manner in a way or other environmental managers.

Submission

To start the submission, enter the required packages in a local machine.

1. Clone The repository:

git clone 

2. Navigate to a project identifier.
3. Create and activate visual nature (recommended):

With Mac / Linux:

python3 -m venv venv
source venv/bin/activate

In Windows:

python3 -m venv venv
venvScriptsactivate

4. Update PIP to the latest version:

python3 -m pip install --upgrade pip

5. Enter the required packages:

pip install -r requirements.txt

The project leaning is written in requirements.txt:

• Boto3
• Support
• premature
• Langchain-AWS
• Langchain-Community
• Pings to the head
• seldom
• destruction
• sight

These packages will manage new integration, web synplogum, data processing, and viewing. They will be included in our visible location during the submission process. This setup helps to ensure that the components are properly installed and separated from the visible environment.

6. Follow the natural variable of AWS CLI to stop AWS authentication.

export AWS_ACCESS_KEY_ID='your-access-key'
export AWS_SECRET_ACCESS_KEY='your-secret-key'

7. Add CSV sample files to S3 bucket made from the Presentsiates section:

Our last location has two sample files:

• basic_test.csv: Complete the blood transition with 15 parameter
• blood_test.csv With basic parameters

The following content of basic_test.csv:

Parameter,Value,Reference_Range,Unit
Hemoglobin,13.8,13.5-17.5,g/dL
RBC,4.8,4.5-5.9,million/µL
WBC,8500,4000-11000,cells/µL
Glucose,92,70-100,mg/dL
Creatinine,1.0,0.7-1.3,mg/dL

Run the following instructions to upload sample files in S3 buck:

aws s3 cp basic_test.csv s3://BUCKET_NAME/

aws s3 cp blood_test.csv s3://BUCKET_NAME/

Go to the app.py line 68 and update the S3 Bucket name in the app.py to match your actual S3 name.

BUCKET_NAME = "your-bucket-name"

8. Run the app:

Dashboard will be available at http://localhost:8501. You can now contact your medical reports by using a web sync.

Using dashboard

This section moves in important factors and shows how we can use the dashboard to analyze medical data.

Visual view of dashboard

The following figures reflect the full dashboard where the selected medical report blood_test.csv From repo indicating the fate of submission and main content. The first figure is also showing the first two graphs.

The following figure shows the second graph of the three of these included in the dashboard.

Dashboard interface is organized for three medical reporting phases:

1. Selection of Docs and Choice of Model (navigation Fact)
   • Choosing the Amazon Children (for example: Claude opus 4.1, claude 3.7 Sonnet, or Amazon Nova Pro)
   • Medical Medical lists available in the drop-down menu
   • Currently to analyze blood_test.csv
   • Telephone usage display display (input, outgoing, and full tokens)
2. Description of analysis of the discussion
   • A clean chat display of natural language questions
   • The History of Disciplinary Chat History
   • Wipe off the format
3. View
   • The range of matching chart showing normal compared to real prices
   • Bar chart showing parameters
   • Trend's plans for many parameters

The question of the question of being able to think

The AI ​​powerful dashboard program program shows the complex understanding of natural negotiations. Here is a sequence of communication skills.

Question 1: First question with hemoglobin:

What is the hemoglobin level in report?

QUESTION 2: The question of following a question that displays a context awareness:

How does this compare to other parameters in the report? Are there any that stand out?

Question 3: Application for complex analysis:

Can you analyze the distribution patterns of percentage-based measurements versus absolute values in this report, and identify any notable patterns in their reference ranges?

The program keeps the context of the discussion while providing a detailed understanding of medical reports, supporting the relevant data information.

The solution can be more developed by preparing the basic model in certain medical information, health questions, and domain technology. This special training helps the model better understand the names of medical, common protocols, and habits relevant to institutions. In addition, organizations may use the previously trained medical professional available AWS market, which is specially prepared for health use cases. When combined with the skills in the system, these special models can provide the relevant medical responses while maintaining compliance with the requirements of health data management requirements.

Amazon Bedrock Guardrails should be prepared to limit the model to providing medical advice, instructions, or diagnosis, Conviction is limited to data and interpretation.

Safety Assessment

While our current shipment uses Dummy medical data for display purposes, it is important to look for safety measures and compliance with real health applications. Here are the recommendations to improve the safety area:

Data privacy:

• Compliance with HIPAA: Use HIPAA-harmony, including access controls and audit routes.
• Encrypt: Use Amazon S3 Server-Side Encryption (Sse-S3) of rest data and TLLs of data in transit.
• Personal Information (PII) Protection:
• Amazon S3 Configuration: Protect your medical data in the following S3 Bluket settings
  • Enable conversion to maintain a complete audit trend and protect from accident or conversion
  • Prevents social access to both bucket levels and accounts
  • Use solid bucket policies that limit access to certain IAM rates and forcing encryption
  • Prepare encryption (AES-256 or KMS) of all loaded items in bucket

Recommended for AWS security implementation:

These are normal recommendations. For a request for production of Healthcare, contact the security expert and make risk assessment to ensure that all the relevant appropriate levels are met.

Clean

To avoid ongoing AWS costs, follow these steps to clean up resources created:

1. Remove Amazon S3 bucket created
2. Delete local resources made:

# Deactivate virtual environment
deactivate
# Remove project directory and virtual environment
rm -rf medical-analysis-dashboard/

Store

In this regard, we demonstrated the development of analyzing reports of medical medical reports that include Amazon Bedrock Ai Ai, Langchain documents, and active networking. The solution is changing complex medical data into an accessible key discussion program provided by large languages ​​of languages ​​available through Amazon Bedrock and stronger health parameters.

The project indicates how the cloud and Ai Technologies can be included in health analysis, which makes interpreting accurate medical report and functional. While our implementation uses dummy data for display purposes, the construction of buildings provides basis for building safe safety programs, associated with health care that can be developed to fulfill the needs of the organization and safety policies.

About the authors

ADIGYA RANJAN Is a delivery consultation for AWs, experts in construction organizations performed in construction and indigenous cloud solutions. It works with customers to designate and use well-prepared technology solutions using the latest AWS technology, including AI services, enabling them to fulfill their business goals and intentions.

Shubam thi Is solving a solution to AWs responsible for modernitarism, containers and security. He has been helping customers to powerful, strong, and cost-based costs in AWS.