PACS Company

AI Integration

Deployment of AI at a PACS Company

Situation

A PACS company wants to provide AI to its customers (for example - MediLab). Note - the PACS company has the ability to themselves select which algorithm they want to offer - all they have to do is change the inference project ID in their API.

Workflow

Create Deployment project

- Create account on CARPL
- Create Dataset
  1. Name: Deployment at Site_1
  2. Description: The dataset for deployment of AI Algorithm #2 at Site_1
  3. Modality: <Choose appropriate>
  4. Click on "Create Dataset"
  5. Click on "Go to Dataset Manager"
  6. Upload studies either by
    - Browser upload
    - API Upload ( https://documenter.getpostman.com/view/12410320/T1LV8iyD?version=latest#ae0df535-eeb5-47b0-9576-475a75a1b489 )
    - Enable Dicom Receive to get AE title and port to push data directly from PACS
- Create Inferencing Project
  1. Name: Deployment at Site_1_Algo#2
  2. Description: An inference project to drive the deployment of Algo#2 at Site_1
  3. Algorithm: <Choose appropriate - Algo#2>
  4. Dataset: Deployment at Site_1

Execute commands (If Executing through API)

- One-time checks

import requestsurl = "https://demo.carpl.ai/api/v1/login"payload = "{\r\n \"emailid\":\"YOUR_EMAIL_ID\",\r\n \"password\":\"YOUR_PASSWORD\"\r\n}"headers= {}response = requests.request("POST", url, headers=headers, data = payload)print(response.text.encode('utf8'))

POST: Show datasets >> get the dataset_ID of Deployment at Site_1

import requestsurl = "https://demo.carpl.ai/api/v1/datasets"payload = {}headers= {}response = requests.request("POST", url, headers=headers, data = payload)print(response.text.encode('utf8'))

POST: Show inference projects >> get the inference project ID of Deployment at Site_1_Algo#2

import requestsurl = "https://demo.carpl.ai/api/v1/inf_projects"payload = {}headers= {}response = requests.request("POST", url, headers=headers, data = payload)print(response.text.encode('utf8'))POST: Generate Auth Token >> get authorization token every hour

- Upload data to Deployment at Site_1

POST: Upload studies >> use dataset_ID from 2.1.1

import requestsurl = "https://demo.carpl.ai/api/v1/upload"payload = {'dataset_id': '<dataset_id>','anon': '0/1'}files = [('file', open('/path/to/file','rb'))]headers = {'token': 'eyJ0eXAiOiJKV1QiLCJhbGciOiJIUzI1NiJ9.eyJlbWFpbGlkIjoicm9oaXQudGFraGFyQGNhcmluZy1yZXNlYXJjaC5jb20iLCJ1c2VyaWQiOjEsImV4cCI6MTU5NjM3OTY0OH0.kz6sHNdcu5HUdCOAIEjCYYpkbloTEklmUcdOK0RyDH8'}response = requests.request("POST", url, headers=headers, data = payload, files = files)print(response.text.encode('utf8'))

- Run inference in Inferencing Project

POST: Run algorithm will start automatically

import requestsurl = "https://demo.carpl.ai/api/v1/run_algorithm"payload = {'project_id': '<inference_id>', 'studies': '["d21f59437e74298df2d4ab453ee2c9f0e8319f1283369175b58054532d5fcd70"]'}files = []headers= {}response = requests.request("POST", url, headers=headers, data = payload, files = files)print(response.text.encode('utf8'))

POST: Download Inference of study

import requestsurl = "https://demo.carpl.ai/api/v1/inf_download_study"payload = {'study_id': '967b5e49f491a4a33a1903f0c4f4697be187a3488e1ff841de8d35e0a0503a46','project_id': '<inference_id>'}files = []headers= {}response = requests.request("POST", url, headers=headers, data = payload, files = files)print(response.text.encode('utf8'))

Repeat steps 2.2 and 2.3 every-time new data is being pushed to CARPL for inference.

Page updated

Report abuse