Index

As this ebook edition doesn't have fixed pagination, the page numbers below are hyperlinked for reference only, based on the printed edition of this book.

A

accurate extraction of semi-structured document types, Amazon Textract

JSON API response 54

accurate extraction of structured document types, Amazon Textract

JSON API response 56

accurate extraction of unstructured document types, Amazon Textract

JSON API response 52

AI stack 5-9

Amazon Augmented AI (Amazon A2I)

about 128

for human review process 128-130

Amazon Comprehend

IDP extraction stage 73-76

PII detection, on AWS Console 115-117

using, for accurate data extraction 67-73

using, for custom entities extraction 77, 78

using, for extraction 208-210

Amazon Comprehend custom classifier

using, for document classification 29, 30

Amazon Comprehend custom entity recognizer

inference result 82-84

trained model performance, checking 81

training 78-81

Amazon Comprehend Medical

about 93-105

for PHI information 117-124

using, for accurate extraction of medical entities 93

using, for medical ontology 105-108

Amazon Comprehend Real-time analysis

entities, detecting 70

key phrases, detecting 71

language, detecting 72, 73

Amazon HealthLake

prior authorization form filling, automating with 158-162

used, for performing FHIR operations 138

used, for performing PUT operation 140

used, for performing READ operation 140, 141

using, as heath data store 134-138

Amazon Rekognition 38

Amazon Rekognition Custom Label 39

Amazon SageMaker

Jupyter notebook instance, creating 17

Amazon Simple Storage Service (Amazon S3 )

about 19

used, for data capture 20

Amazon Textract

about 47

accurate extraction, of ID document 57, 58

accurate extraction, of semi-structured document types 53, 54

accurate extraction, of structured document types 55

accurate extraction, of unstructured document types 48-51

ID document (US passport) accurate extraction 58-60

invoice document accurate extraction 62, 63

receipt document accurate extraction 60-62

Amazon Textract Queries 162

Amazon Web Services (AWS)

about 5

account, setting up 16

account, signing up 15

Application Programming Interfaces (APIs) 68, 93

ArtificiaI Intelligence (AI) 5

automated ML (AutoML) 77

AWS Boto3 library 21

AWS Free Tier

reference link 16

AWS Management Console

reference link 16

AWS ML 5-9

B

bag of words (BoW) 46

benefits enrollment process

automating, with IDP 178-185

bounding box feature 212

C

CDS hooks

reference link 157

Centers for Medicare and Medicaid Services (CMS) 156

claims adjudication 168

claims processing 168

completeness check

in IDP pipeline 112-114

Comprehend custom classification model

document classification, based on sensitive data 38

inference 35-37

result 37

training 31-35

consumption stage 13

Coverage Requirements Discovery (CRD) 157

CREATE, READ, UPDATE, and DELETE (CRUD) 134

custom entity recognition 208

D

data capture stage

about 198

data format 10

data sources 10

data store 10

data capture, with Amazon S3

about 20

data sources 20-26

data store 20

sensitive document processing 27

digitalization 9

digitization 9

document categorization

with computer vision 38-43

document classification 11, 29

document extraction

about 11

in IDP pipeline 46, 47

document processing

financial services use cases 214, 215

document review process, with human-in-the-loop

about 127

Amazon Augmented AI (Amazon A2I), for human review process 128-130

validation, with automation 127, 128

validation, without automation 127

document splitter 11

E

Electronic Health Records (EHRs) 134, 156

Electronic Medical Record (EMR) 4

enterprise resource planning (ERP) system 46

F

Fast Healthcare Interoperability Resources (FHIR) 134

FHIR data store

used, for handling documents 145-151

FHIR operations

performing, with Amazon HealthLake 138

H

healthcare prior authorization

IDP, using 155

process 156-158

HealthLake delete request

using 144, 145

HealthLake GET request

using 143

HealthLake PUT request

using 142, 143

Hierarchical Condition Category (HCC) 174

Home Owner Association (HOA) 77

I

ID document (driver's license)

accurate extraction, with Amazon Textract 57, 58

ID document (US passport)

accurate extraction, with Amazon Textract 58-60

Identity and Access Management (IAM) 17

IDP flow, insurance claims processing

data capture 185

document classification 185, 186

document enrichment 189

document extraction 186-188

review and validation 189-191

IDP pipeline

about 9

consumption 12

data capture 10

document classification 11

document enrichment 11

document extraction 11, 46, 47

document post-processing (review and verification 12

mortgage processing review and verification stage 213

IDP pipeline implementation, with Amazon A2I

reference link 130

insurance claims processing

extraction, with IDP 185

for invalid claims form 191-193

integration stage 13

Intelligent Document Processing (IDP)

benefits enrollment process, automating with 178-185

document enrichment stage 90, 92

industry use cases 4

insurance claims processing, extracting with 185

learning, with pharmacy receipt automation 162-167

mortgage processing data capture and data categorization, automating 196, 197

using, for healthcare claims processing and risk adjustment 168-174

using, with healthcare prior authorization 155

using, with mortgage processing extraction and enrichment 198

invoice document

accurate extraction, with Amazon Textract 62, 63

K

KYC standards 214

L

legacy document extraction

challenges 45-47

M

Machine Learning (ML) 5, 67

manual processing 46

medical ontology

Amazon Comprehend Medical, using 105-108

ML model

versus pre-trained AI services 47, 48

mortgage application processing

document enrichment 210-212

mortgage processing data capture and data categorization

automating, with IDP 196, 197

mortgage processing extraction and enrichment

IDP, using 198-207

N

Named Entity Recognition (NER) 11

natural language models

used, for transforming medical text 141

Natural Language Processing (NLP) 67, 93

notebook execution role, Amazon SageMaker

IAM permissions and trust relationships, modifying for 17-19

O

optical character recognition (OCR) 46

P

Personal Identifiable Information (PII)

about 12, 115

reference link 115

pharmacy receipt automation

IDP, learning with 162-167

post-processing

sensitive data 115

using, for completeness check 112-114

pre-trained AI services

versus ML model 47, 48

Protected Health Information (PHI) 12

PUT operation

performing, with Amazon HealthLake 138-140

R

READ operation

performing 140, 141

receipt document

accurate extraction, performing with Amazon Textract 60-62

S

Securities and Exchange Commission (SEC) 214

semi-structured documents

about 10, 53

accurate extraction, with Amazon Textract 53, 54

sensitive data, post-processing

about 115

Amazon Comprehend Medical, for PHI information 117-124

Amazon Comprehend PII detection, on AWS Console 115-117

redaction, with structural elements 125-127

sensitive document processing

about 27, 28

diversity 28, 29

preprocessing 29

Social Security Card (SSC) 47

Social Security Number (SSN) 47

structured documents

about 10, 55

accurate extraction, with Amazon Textract 55

Systems, Applications, and Products (SAP) 46

T

Textract asynchronous operation 24

U

unstructured documents

about 10, 48

accurate extraction, with Amazon Textract 48-51