SlideShare a Scribd company logo
1 of 17
http://www.free-powerpoint-templates-design.com
Prediction of Arterial Blood
Gases(ABG) by Using Neural
Network In Trauma Patients
Milad Shayan
Mohammad Sabouri
Dr. Shahram Paydar
Leila Shayan
Dr. Shahram Paydar
Leila Shayan
Milad Shayan
Mohammad Sabouri
Use of artificial intelligence and neural network algorithms to predict arterial
blood gas items in trauma victims
Milad Shayan, Mohammad Sabouri, Leila Shayan, Shahram Paydar
doi: https://doi.org/10.1101/2020.03.14.20035584
Applied Control & Robotics Research Laboratory of Shiraz University
https://sites.google.com/view/acrrl/home
1
2
3
4
Table of
Contents
Introduction
Medical description and our goal and phase alignment
Data Preprocessing
Choose I/O and What we do
Neural Network
Initial design and prediction
Result
Analysis and result of 1st phase
3
1
INTRODUCTION
Trauma and ABG
5
Traumatic injury
Serious injury to the body, as from physical violence or an accident
Psychological trauma
Severe emotional or mental distress caused by an experience
Arterial Blood Gases
An ABG is a blood test that measures the acidity, or pH, and the levels of oxygen (O2)
and carbon dioxide (CO2) and HCO3 from an artery
01
02
03
3 Phase of Project
6
ABG
Arterial Blood Gases include
PH & PO2 & PCO2 & HCO3
BE
Base Excess or Base Deficit
Fibrinogen
Level of Fibrinogen
One of the Coagulating factor of blood
Reason of Prediction of ABG
On arrival
Age, Gender, Vital signs
After 1 Hour
First ABG and the other
parameter
After 2 Hour
Second ABG and Required
parameter
Transfer to
Surgery room, CCU, ICU & …
7
2
DATA
PREPROCESSING
Our Step
Del/Rep Missing
Low Missing
Use Average
Correct Data
Scatter chart
Normalize Data
[0,1]
Export to NN
Ready to use
9
Normalize Data
𝑁𝑜𝑟𝑚𝑎𝑙𝑖 =
𝑋𝑖 − 𝑀𝑖𝑛𝑖
𝑀𝑎𝑥𝑖 − 𝑀𝑖𝑛𝑖
Formula 1
12
Input/Output Information
13
Row data Normalize from normal distribution Normalize from row data
variance average min max variance average min max variance average min max
input
Age 17.70459 35.925439 2 95 0.166769 0.5299989 0 1 0.190372 0.36479 0 1
Blood pressure systol 21.7673 125.84837 40 210 0.128043 0.5049904 0 1 0.128043 0.50499 0 1
Blood pressure dystol 14.68896 79.112782 20 146 0.116579 0.4691491 0 1 0.116579 0.469149 0 1
Respiratory Rate 3.62063 19.620301 8 45 0.097855 0.3140622 0 1 0.097855 0.314062 0 1
Pulse Rate 22.36556 100.48183 17 190 0.129281 0.4825539 0 1 0.129281 0.482554 0 1
output
pH 0.122997 7.3606266 6.69 9.94 0.037845 0.2063466 0 1 0.037845 0.206347 0 1
Pressure of CO2 10.39696 40.989286 11 99 0.118147 0.3407873 0 1 0.118147 0.340787 0 1
Pressure of O2 41.10154 44.673058 0.4 333.3 0.122121 0.6550968 0 1 0.123465 0.132992 0 1
Hco3 4.636405 22.924311 3.4 48.5 0.102803 0.4329115 0 1 0.102803 0.432912 0 1
3
NEURAL NETWORK
Table of Neural Network
15
name of
network
network type
training
function
adaption learning
function
performance
function
number of
layers
number of
neurons
transfer
function
network1 feed-forward backprop trainLM LearnGDM MSE 3 50 logsig
network2 feed-forward backprop trainCGB LearnGD MSE 3 50 logsig
network3 feed-forward backprop trainCGB LearnGDM MSEREG 3 50 logsig
network4 feed-forward backprop trainCGB LearnGD SSE 3 40 tansig
network5 feed-forward backprop trainCGB LearnGDM MSEREG 3 40 tansig
network6 feed-forward backprop trainCGF LearnGD SSE 3 40 logsig
network7 feed-forward backprop trainCGF LearnGDM MSEREG 3 40 tansig
network8 feed-forward backprop trainBR LearnGDM MSEREG 3 30 tansig
network9 feed-forward backprop trainSCG LearnGDM MSEREG 3 40 logsig
network10 feed-forward backprop trainRP LearnGDM MSEREG 3 50 logsig
4
RESULT
Table of Result
17
name of
network
Type of Data
Accuracy of Learn Accuracy of Test Average of Absolute Errors
ph pco2 po2 hco3 Total ph pco2 po2 hco3 Total ph pco2 po2 hco3
network1
Normalize Data 89.77% 74.77% 86.43% 83.13% 83.52% 90.06% 72.65% 88.09% 81.18% 83.00%Learn 0.069 7.5676 23.981 3.2935
Real Data 99.07% 81.54% 46.32% 85.63% 78.14% 99.10% 80.09% 48.73% 84.06% 78.00%Test 0.066 8.0582 25.998 3.5374
network2
Normalize Data 89.90% 75.71% 86.83% 83.72% 84.04% 89.67% 68.01% 86.39% 79.50% 80.89%Learn 0.077 7.7695 23.287 3.378
Real Data 99.08% 82.23% 47.87% 86.13% 78.83% 99.07% 76.71% 44.42% 82.64% 75.71%Test 0.068 9.4247 28.187 3.8533
network3
Normalize Data 89.92% 76.22% 86.94% 83.69% 84.19% 89.23% 70.15% 86.77% 79.51% 81.41%Learn 0.068 7.1319 23.173 3.1843
Real Data 99.08% 82.60% 48.13% 86.11% 78.98% 99.03% 78.27% 46.78% 82.65% 76.68%Test 0.071 8.7944 26.99 3.8522
network4
Normalize Data 89.63% 74.58% 86.48% 83.13% 83.46% 89.05% 72.57% 88.59% 81.38% 82.90%Learn 0.07 7.6233 23.911 3.294
Real Data 99.06% 81.40% 46.48% 85.63% 78.14% 99.01% 80.03% 50.43% 84.23% 78.42%Test 0.073 8.0824 25.137 3.5006
network5
Normalize Data 88.73% 74.23% 86.29% 82.84% 83.02% 89.74% 72.48% 88.97% 81.57% 83.19%Learn 0.076 7.7287 24.164 3.3496
Real Data 98.97% 81.14% 45.91% 85.39% 77.85% 99.07% 79.96% 51.03% 84.39% 78.62%Test 0.068 8.1074 24.836 3.4644
network6
Normalize Data 89.12% 74.36% 86.29% 83.01% 83.20% 89.61% 72.90% 88.93% 81.61% 83.26%Learn 0.073 7.69 24.122 3.3174
Real Data 99.01% 81.24% 46.00% 85.53% 77.94% 99.06% 80.27% 50.94% 84.43% 78.68%Test 0.069 7.9844 24.878 3.4561
network7
Normalize Data 89.59% 74.75% 86.37% 83.17% 83.47% 90.13% 71.72% 88.70% 81.03% 82.90%Learn 0.07 7.5738 24.112 3.2861
Real Data 99.05% 81.52% 46.03% 85.67% 78.07% 99.11% 79.41% 50.31% 83.94% 78.19%Test 0.065 8.3327 25.199 3.5651
network8
Normalize Data 89.83% 74.75% 86.47% 83.06% 83.53% 90.51% 72.53% 88.63% 81.70% 83.34%Learn 0.068 7.572 23.96 3.3083
Real Data 99.07% 81.53% 46.37% 85.57% 78.13% 99.14% 80.00% 50.35% 84.51% 78.50%Test 0.063 8.0926 25.179 3.4393
network9
Normalize Data 89.65% 74.40% 86.49% 83.01% 83.39% 90.10% 72.42% 88.42% 81.59% 83.13%Learn 0.069 7.6787 23.883 3.317
Real Data 99.06% 81.27% 46.54% 85.53% 78.10% 99.11% 79.91% 49.61% 84.41% 78.26%Test 0.066 8.1274 25.554 3.4602
network10
Normalize Data 88.55% 74.36% 86.32% 82.82% 83.01% 88.42% 71.48% 88.12% 80.72% 82.18%Learn 0.077 7.6893 24.101 3.3541
Real Data 98.96% 81.24% 46.05% 85.37% 77.90% 98.96% 79.24% 41.01% 83.67% 75.72%Test 0.077 8.4018 29.914 3.6241
Real Data
𝑁𝑜𝑟𝑚𝑎𝑙𝑖 =
𝑋𝑖 − 𝑀𝑖𝑛𝑖
𝑀𝑎𝑥𝑖 − 𝑀𝑖𝑛𝑖
Formula 2
18
⇒ 𝑋𝑖 = 𝑁𝑜𝑟𝑚𝑎𝑙𝑖 × 𝑀𝑎𝑥𝑖 − 𝑀𝑖𝑛𝑖 + 𝑀𝑖𝑛𝑖
Formula 3 %𝐴𝑐𝑐𝑢𝑟𝑎𝑐𝑦 = 1 −
𝐴𝑣𝑔 |𝑋𝑖 − 𝑋
∧
𝑖| 𝑛
𝐴𝑣𝑔 𝑋𝑖
× 100
Thank You for Watching!

More Related Content

Similar to Prediction of Arterial Blood Gases(ABG) by Using Neural Network In Trauma Patients

Question1.xlsxAnova ResultsSUMMARY OUTPUTRegression Statistic.docx
Question1.xlsxAnova ResultsSUMMARY OUTPUTRegression Statistic.docxQuestion1.xlsxAnova ResultsSUMMARY OUTPUTRegression Statistic.docx
Question1.xlsxAnova ResultsSUMMARY OUTPUTRegression Statistic.docx
catheryncouper
 
Soft And Handling
Soft And HandlingSoft And Handling
Soft And Handling
hiratufail
 
Survival Analysis On Kidney Failure of Kidney Tranplant Patients
Survival Analysis On Kidney Failure of Kidney Tranplant PatientsSurvival Analysis On Kidney Failure of Kidney Tranplant Patients
Survival Analysis On Kidney Failure of Kidney Tranplant Patients
Dwaipayan Mukhopadhyay
 
Presentation Stage Graphe de Connectivité du Cerveau 2014 (FR)
Presentation Stage Graphe de Connectivité du Cerveau 2014 (FR)Presentation Stage Graphe de Connectivité du Cerveau 2014 (FR)
Presentation Stage Graphe de Connectivité du Cerveau 2014 (FR)
Romain Chion
 

Similar to Prediction of Arterial Blood Gases(ABG) by Using Neural Network In Trauma Patients (20)

BDW Chicago 2016 - Don Deloach, CEO and President, Infobright - Rethinking Ar...
BDW Chicago 2016 - Don Deloach, CEO and President, Infobright - Rethinking Ar...BDW Chicago 2016 - Don Deloach, CEO and President, Infobright - Rethinking Ar...
BDW Chicago 2016 - Don Deloach, CEO and President, Infobright - Rethinking Ar...
 
First trimester report regional veterinary laboratory pokhara
First trimester report regional veterinary laboratory pokharaFirst trimester report regional veterinary laboratory pokhara
First trimester report regional veterinary laboratory pokhara
 
04.3 heterogeneous debt portfolios
04.3   heterogeneous debt portfolios04.3   heterogeneous debt portfolios
04.3 heterogeneous debt portfolios
 
healthcare healthcare statistics.pdf
healthcare healthcare statistics.pdfhealthcare healthcare statistics.pdf
healthcare healthcare statistics.pdf
 
Chemistry to Clinic: The Nanosyn Approach to Bridging the Gaps in Translation...
Chemistry to Clinic: The Nanosyn Approach to Bridging the Gaps in Translation...Chemistry to Clinic: The Nanosyn Approach to Bridging the Gaps in Translation...
Chemistry to Clinic: The Nanosyn Approach to Bridging the Gaps in Translation...
 
Mohammed alharbi 2 e (1)
Mohammed alharbi 2 e (1)Mohammed alharbi 2 e (1)
Mohammed alharbi 2 e (1)
 
Question1.xlsxAnova ResultsSUMMARY OUTPUTRegression Statistic.docx
Question1.xlsxAnova ResultsSUMMARY OUTPUTRegression Statistic.docxQuestion1.xlsxAnova ResultsSUMMARY OUTPUTRegression Statistic.docx
Question1.xlsxAnova ResultsSUMMARY OUTPUTRegression Statistic.docx
 
Soft And Handling
Soft And HandlingSoft And Handling
Soft And Handling
 
Chromatography: Meeting the Challenges of EU regulations with up-to-date Conf...
Chromatography: Meeting the Challenges of EU regulations with up-to-date Conf...Chromatography: Meeting the Challenges of EU regulations with up-to-date Conf...
Chromatography: Meeting the Challenges of EU regulations with up-to-date Conf...
 
Nanostructured Lipid Carrier based Dry Powder Inhaler (DPI) of Anti TB drug.
Nanostructured Lipid Carrier based Dry Powder Inhaler (DPI) of Anti TB drug. Nanostructured Lipid Carrier based Dry Powder Inhaler (DPI) of Anti TB drug.
Nanostructured Lipid Carrier based Dry Powder Inhaler (DPI) of Anti TB drug.
 
IDNADEX: Improving DNA Data Exchange Validation Studies of a Global STR System
IDNADEX: Improving DNA Data Exchange Validation Studies of a Global STR SystemIDNADEX: Improving DNA Data Exchange Validation Studies of a Global STR System
IDNADEX: Improving DNA Data Exchange Validation Studies of a Global STR System
 
Survival Analysis On Kidney Failure of Kidney Tranplant Patients
Survival Analysis On Kidney Failure of Kidney Tranplant PatientsSurvival Analysis On Kidney Failure of Kidney Tranplant Patients
Survival Analysis On Kidney Failure of Kidney Tranplant Patients
 
Survival analysis on kidney failure of kidney transplant patients
Survival analysis on kidney failure of kidney transplant patientsSurvival analysis on kidney failure of kidney transplant patients
Survival analysis on kidney failure of kidney transplant patients
 
Purification optimization and characterization of protease from Bacillus va...
Purification optimization and characterization of  protease from  Bacillus va...Purification optimization and characterization of  protease from  Bacillus va...
Purification optimization and characterization of protease from Bacillus va...
 
Kshivets astana wscts2017
Kshivets astana wscts2017Kshivets astana wscts2017
Kshivets astana wscts2017
 
Presentation Stage Graphe de Connectivité du Cerveau 2014 (FR)
Presentation Stage Graphe de Connectivité du Cerveau 2014 (FR)Presentation Stage Graphe de Connectivité du Cerveau 2014 (FR)
Presentation Stage Graphe de Connectivité du Cerveau 2014 (FR)
 
Kshivets wscts2015
Kshivets wscts2015Kshivets wscts2015
Kshivets wscts2015
 
Database Marketing - Dominick's stores in Chicago distric
Database Marketing - Dominick's stores in Chicago districDatabase Marketing - Dominick's stores in Chicago distric
Database Marketing - Dominick's stores in Chicago distric
 
Chang Sha, China
Chang Sha, ChinaChang Sha, China
Chang Sha, China
 
2019 Triangle Machine Learning Day - Machine Learning from De-Identified Code...
2019 Triangle Machine Learning Day - Machine Learning from De-Identified Code...2019 Triangle Machine Learning Day - Machine Learning from De-Identified Code...
2019 Triangle Machine Learning Day - Machine Learning from De-Identified Code...
 

More from Mohammad Sabouri

Extremely low-cost lower limb prostheses_G12.pptx
Extremely low-cost lower limb prostheses_G12.pptxExtremely low-cost lower limb prostheses_G12.pptx
Extremely low-cost lower limb prostheses_G12.pptx
Mohammad Sabouri
 

More from Mohammad Sabouri (15)

Extremely low-cost lower limb prostheses_G12.pptx
Extremely low-cost lower limb prostheses_G12.pptxExtremely low-cost lower limb prostheses_G12.pptx
Extremely low-cost lower limb prostheses_G12.pptx
 
MECHANICAL DESIGN METHODS IN ROBOTICS.pptx
MECHANICAL DESIGN METHODS IN ROBOTICS.pptxMECHANICAL DESIGN METHODS IN ROBOTICS.pptx
MECHANICAL DESIGN METHODS IN ROBOTICS.pptx
 
Human Computer Interaction (HCI).pptx
Human Computer Interaction (HCI).pptxHuman Computer Interaction (HCI).pptx
Human Computer Interaction (HCI).pptx
 
Intelligent Decision Making Assistant (IDMA) for SAL improvement.pptx
Intelligent Decision Making Assistant (IDMA) for SAL improvement.pptxIntelligent Decision Making Assistant (IDMA) for SAL improvement.pptx
Intelligent Decision Making Assistant (IDMA) for SAL improvement.pptx
 
Introducing the services of Iran Patent Center- PDF
Introducing the services of Iran Patent Center- PDFIntroducing the services of Iran Patent Center- PDF
Introducing the services of Iran Patent Center- PDF
 
Introduction to Lens database -in Persian (powerful site for searching)
Introduction to Lens database -in Persian (powerful site for searching)Introduction to Lens database -in Persian (powerful site for searching)
Introduction to Lens database -in Persian (powerful site for searching)
 
CV_ nov.2019
CV_ nov.2019CV_ nov.2019
CV_ nov.2019
 
Icbme2020- Use of neural network algorithms to predict arterial blood gas ite...
Icbme2020- Use of neural network algorithms to predict arterial blood gas ite...Icbme2020- Use of neural network algorithms to predict arterial blood gas ite...
Icbme2020- Use of neural network algorithms to predict arterial blood gas ite...
 
Traffic monitoring using drone_ACRRL_Shiraz University
Traffic monitoring using drone_ACRRL_Shiraz UniversityTraffic monitoring using drone_ACRRL_Shiraz University
Traffic monitoring using drone_ACRRL_Shiraz University
 
Robotic introduction
Robotic introductionRobotic introduction
Robotic introduction
 
Recurrent Neural Network
Recurrent Neural NetworkRecurrent Neural Network
Recurrent Neural Network
 
Labview2_Computer Applications in Control_ACRRL
Labview2_Computer Applications in Control_ACRRLLabview2_Computer Applications in Control_ACRRL
Labview2_Computer Applications in Control_ACRRL
 
Labview1_ Computer Applications in Control_ACRRL
Labview1_ Computer Applications in Control_ACRRLLabview1_ Computer Applications in Control_ACRRL
Labview1_ Computer Applications in Control_ACRRL
 
Spoofing attack on PMU (Phasor measurement unit)
Spoofing attack on PMU (Phasor measurement unit)Spoofing attack on PMU (Phasor measurement unit)
Spoofing attack on PMU (Phasor measurement unit)
 
Haptic technology ppt
Haptic technology pptHaptic technology ppt
Haptic technology ppt
 

Recently uploaded

scipt v1.pptxcxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx...
scipt v1.pptxcxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx...scipt v1.pptxcxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx...
scipt v1.pptxcxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx...
HenryBriggs2
 
Introduction to Robotics in Mechanical Engineering.pptx
Introduction to Robotics in Mechanical Engineering.pptxIntroduction to Robotics in Mechanical Engineering.pptx
Introduction to Robotics in Mechanical Engineering.pptx
hublikarsn
 
Standard vs Custom Battery Packs - Decoding the Power Play
Standard vs Custom Battery Packs - Decoding the Power PlayStandard vs Custom Battery Packs - Decoding the Power Play
Standard vs Custom Battery Packs - Decoding the Power Play
Epec Engineered Technologies
 
Integrated Test Rig For HTFE-25 - Neometrix
Integrated Test Rig For HTFE-25 - NeometrixIntegrated Test Rig For HTFE-25 - Neometrix
Integrated Test Rig For HTFE-25 - Neometrix
Neometrix_Engineering_Pvt_Ltd
 
XXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXX
XXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXX
XXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXX
ssuser89054b
 
1_Introduction + EAM Vocabulary + how to navigate in EAM.pdf
1_Introduction + EAM Vocabulary + how to navigate in EAM.pdf1_Introduction + EAM Vocabulary + how to navigate in EAM.pdf
1_Introduction + EAM Vocabulary + how to navigate in EAM.pdf
AldoGarca30
 

Recently uploaded (20)

scipt v1.pptxcxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx...
scipt v1.pptxcxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx...scipt v1.pptxcxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx...
scipt v1.pptxcxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx...
 
8086 Microprocessor Architecture: 16-bit microprocessor
8086 Microprocessor Architecture: 16-bit microprocessor8086 Microprocessor Architecture: 16-bit microprocessor
8086 Microprocessor Architecture: 16-bit microprocessor
 
Basic Electronics for diploma students as per technical education Kerala Syll...
Basic Electronics for diploma students as per technical education Kerala Syll...Basic Electronics for diploma students as per technical education Kerala Syll...
Basic Electronics for diploma students as per technical education Kerala Syll...
 
8th International Conference on Soft Computing, Mathematics and Control (SMC ...
8th International Conference on Soft Computing, Mathematics and Control (SMC ...8th International Conference on Soft Computing, Mathematics and Control (SMC ...
8th International Conference on Soft Computing, Mathematics and Control (SMC ...
 
Path loss model, OKUMURA Model, Hata Model
Path loss model, OKUMURA Model, Hata ModelPath loss model, OKUMURA Model, Hata Model
Path loss model, OKUMURA Model, Hata Model
 
Max. shear stress theory-Maximum Shear Stress Theory ​ Maximum Distortional ...
Max. shear stress theory-Maximum Shear Stress Theory ​  Maximum Distortional ...Max. shear stress theory-Maximum Shear Stress Theory ​  Maximum Distortional ...
Max. shear stress theory-Maximum Shear Stress Theory ​ Maximum Distortional ...
 
Introduction to Geographic Information Systems
Introduction to Geographic Information SystemsIntroduction to Geographic Information Systems
Introduction to Geographic Information Systems
 
Introduction to Robotics in Mechanical Engineering.pptx
Introduction to Robotics in Mechanical Engineering.pptxIntroduction to Robotics in Mechanical Engineering.pptx
Introduction to Robotics in Mechanical Engineering.pptx
 
HOA1&2 - Module 3 - PREHISTORCI ARCHITECTURE OF KERALA.pptx
HOA1&2 - Module 3 - PREHISTORCI ARCHITECTURE OF KERALA.pptxHOA1&2 - Module 3 - PREHISTORCI ARCHITECTURE OF KERALA.pptx
HOA1&2 - Module 3 - PREHISTORCI ARCHITECTURE OF KERALA.pptx
 
Post office management system project ..pdf
Post office management system project ..pdfPost office management system project ..pdf
Post office management system project ..pdf
 
Worksharing and 3D Modeling with Revit.pptx
Worksharing and 3D Modeling with Revit.pptxWorksharing and 3D Modeling with Revit.pptx
Worksharing and 3D Modeling with Revit.pptx
 
Standard vs Custom Battery Packs - Decoding the Power Play
Standard vs Custom Battery Packs - Decoding the Power PlayStandard vs Custom Battery Packs - Decoding the Power Play
Standard vs Custom Battery Packs - Decoding the Power Play
 
Integrated Test Rig For HTFE-25 - Neometrix
Integrated Test Rig For HTFE-25 - NeometrixIntegrated Test Rig For HTFE-25 - Neometrix
Integrated Test Rig For HTFE-25 - Neometrix
 
Unit 4_Part 1 CSE2001 Exception Handling and Function Template and Class Temp...
Unit 4_Part 1 CSE2001 Exception Handling and Function Template and Class Temp...Unit 4_Part 1 CSE2001 Exception Handling and Function Template and Class Temp...
Unit 4_Part 1 CSE2001 Exception Handling and Function Template and Class Temp...
 
Linux Systems Programming: Inter Process Communication (IPC) using Pipes
Linux Systems Programming: Inter Process Communication (IPC) using PipesLinux Systems Programming: Inter Process Communication (IPC) using Pipes
Linux Systems Programming: Inter Process Communication (IPC) using Pipes
 
Online electricity billing project report..pdf
Online electricity billing project report..pdfOnline electricity billing project report..pdf
Online electricity billing project report..pdf
 
Signal Processing and Linear System Analysis
Signal Processing and Linear System AnalysisSignal Processing and Linear System Analysis
Signal Processing and Linear System Analysis
 
Convergence of Robotics and Gen AI offers excellent opportunities for Entrepr...
Convergence of Robotics and Gen AI offers excellent opportunities for Entrepr...Convergence of Robotics and Gen AI offers excellent opportunities for Entrepr...
Convergence of Robotics and Gen AI offers excellent opportunities for Entrepr...
 
XXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXX
XXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXX
XXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXX
 
1_Introduction + EAM Vocabulary + how to navigate in EAM.pdf
1_Introduction + EAM Vocabulary + how to navigate in EAM.pdf1_Introduction + EAM Vocabulary + how to navigate in EAM.pdf
1_Introduction + EAM Vocabulary + how to navigate in EAM.pdf
 

Prediction of Arterial Blood Gases(ABG) by Using Neural Network In Trauma Patients

  • 1. http://www.free-powerpoint-templates-design.com Prediction of Arterial Blood Gases(ABG) by Using Neural Network In Trauma Patients Milad Shayan Mohammad Sabouri Dr. Shahram Paydar Leila Shayan
  • 2. Dr. Shahram Paydar Leila Shayan Milad Shayan Mohammad Sabouri Use of artificial intelligence and neural network algorithms to predict arterial blood gas items in trauma victims Milad Shayan, Mohammad Sabouri, Leila Shayan, Shahram Paydar doi: https://doi.org/10.1101/2020.03.14.20035584 Applied Control & Robotics Research Laboratory of Shiraz University https://sites.google.com/view/acrrl/home
  • 3. 1 2 3 4 Table of Contents Introduction Medical description and our goal and phase alignment Data Preprocessing Choose I/O and What we do Neural Network Initial design and prediction Result Analysis and result of 1st phase 3
  • 5. Trauma and ABG 5 Traumatic injury Serious injury to the body, as from physical violence or an accident Psychological trauma Severe emotional or mental distress caused by an experience Arterial Blood Gases An ABG is a blood test that measures the acidity, or pH, and the levels of oxygen (O2) and carbon dioxide (CO2) and HCO3 from an artery
  • 6. 01 02 03 3 Phase of Project 6 ABG Arterial Blood Gases include PH & PO2 & PCO2 & HCO3 BE Base Excess or Base Deficit Fibrinogen Level of Fibrinogen One of the Coagulating factor of blood
  • 7. Reason of Prediction of ABG On arrival Age, Gender, Vital signs After 1 Hour First ABG and the other parameter After 2 Hour Second ABG and Required parameter Transfer to Surgery room, CCU, ICU & … 7
  • 9. Our Step Del/Rep Missing Low Missing Use Average Correct Data Scatter chart Normalize Data [0,1] Export to NN Ready to use 9
  • 10. Normalize Data 𝑁𝑜𝑟𝑚𝑎𝑙𝑖 = 𝑋𝑖 − 𝑀𝑖𝑛𝑖 𝑀𝑎𝑥𝑖 − 𝑀𝑖𝑛𝑖 Formula 1 12
  • 11. Input/Output Information 13 Row data Normalize from normal distribution Normalize from row data variance average min max variance average min max variance average min max input Age 17.70459 35.925439 2 95 0.166769 0.5299989 0 1 0.190372 0.36479 0 1 Blood pressure systol 21.7673 125.84837 40 210 0.128043 0.5049904 0 1 0.128043 0.50499 0 1 Blood pressure dystol 14.68896 79.112782 20 146 0.116579 0.4691491 0 1 0.116579 0.469149 0 1 Respiratory Rate 3.62063 19.620301 8 45 0.097855 0.3140622 0 1 0.097855 0.314062 0 1 Pulse Rate 22.36556 100.48183 17 190 0.129281 0.4825539 0 1 0.129281 0.482554 0 1 output pH 0.122997 7.3606266 6.69 9.94 0.037845 0.2063466 0 1 0.037845 0.206347 0 1 Pressure of CO2 10.39696 40.989286 11 99 0.118147 0.3407873 0 1 0.118147 0.340787 0 1 Pressure of O2 41.10154 44.673058 0.4 333.3 0.122121 0.6550968 0 1 0.123465 0.132992 0 1 Hco3 4.636405 22.924311 3.4 48.5 0.102803 0.4329115 0 1 0.102803 0.432912 0 1
  • 13. Table of Neural Network 15 name of network network type training function adaption learning function performance function number of layers number of neurons transfer function network1 feed-forward backprop trainLM LearnGDM MSE 3 50 logsig network2 feed-forward backprop trainCGB LearnGD MSE 3 50 logsig network3 feed-forward backprop trainCGB LearnGDM MSEREG 3 50 logsig network4 feed-forward backprop trainCGB LearnGD SSE 3 40 tansig network5 feed-forward backprop trainCGB LearnGDM MSEREG 3 40 tansig network6 feed-forward backprop trainCGF LearnGD SSE 3 40 logsig network7 feed-forward backprop trainCGF LearnGDM MSEREG 3 40 tansig network8 feed-forward backprop trainBR LearnGDM MSEREG 3 30 tansig network9 feed-forward backprop trainSCG LearnGDM MSEREG 3 40 logsig network10 feed-forward backprop trainRP LearnGDM MSEREG 3 50 logsig
  • 15. Table of Result 17 name of network Type of Data Accuracy of Learn Accuracy of Test Average of Absolute Errors ph pco2 po2 hco3 Total ph pco2 po2 hco3 Total ph pco2 po2 hco3 network1 Normalize Data 89.77% 74.77% 86.43% 83.13% 83.52% 90.06% 72.65% 88.09% 81.18% 83.00%Learn 0.069 7.5676 23.981 3.2935 Real Data 99.07% 81.54% 46.32% 85.63% 78.14% 99.10% 80.09% 48.73% 84.06% 78.00%Test 0.066 8.0582 25.998 3.5374 network2 Normalize Data 89.90% 75.71% 86.83% 83.72% 84.04% 89.67% 68.01% 86.39% 79.50% 80.89%Learn 0.077 7.7695 23.287 3.378 Real Data 99.08% 82.23% 47.87% 86.13% 78.83% 99.07% 76.71% 44.42% 82.64% 75.71%Test 0.068 9.4247 28.187 3.8533 network3 Normalize Data 89.92% 76.22% 86.94% 83.69% 84.19% 89.23% 70.15% 86.77% 79.51% 81.41%Learn 0.068 7.1319 23.173 3.1843 Real Data 99.08% 82.60% 48.13% 86.11% 78.98% 99.03% 78.27% 46.78% 82.65% 76.68%Test 0.071 8.7944 26.99 3.8522 network4 Normalize Data 89.63% 74.58% 86.48% 83.13% 83.46% 89.05% 72.57% 88.59% 81.38% 82.90%Learn 0.07 7.6233 23.911 3.294 Real Data 99.06% 81.40% 46.48% 85.63% 78.14% 99.01% 80.03% 50.43% 84.23% 78.42%Test 0.073 8.0824 25.137 3.5006 network5 Normalize Data 88.73% 74.23% 86.29% 82.84% 83.02% 89.74% 72.48% 88.97% 81.57% 83.19%Learn 0.076 7.7287 24.164 3.3496 Real Data 98.97% 81.14% 45.91% 85.39% 77.85% 99.07% 79.96% 51.03% 84.39% 78.62%Test 0.068 8.1074 24.836 3.4644 network6 Normalize Data 89.12% 74.36% 86.29% 83.01% 83.20% 89.61% 72.90% 88.93% 81.61% 83.26%Learn 0.073 7.69 24.122 3.3174 Real Data 99.01% 81.24% 46.00% 85.53% 77.94% 99.06% 80.27% 50.94% 84.43% 78.68%Test 0.069 7.9844 24.878 3.4561 network7 Normalize Data 89.59% 74.75% 86.37% 83.17% 83.47% 90.13% 71.72% 88.70% 81.03% 82.90%Learn 0.07 7.5738 24.112 3.2861 Real Data 99.05% 81.52% 46.03% 85.67% 78.07% 99.11% 79.41% 50.31% 83.94% 78.19%Test 0.065 8.3327 25.199 3.5651 network8 Normalize Data 89.83% 74.75% 86.47% 83.06% 83.53% 90.51% 72.53% 88.63% 81.70% 83.34%Learn 0.068 7.572 23.96 3.3083 Real Data 99.07% 81.53% 46.37% 85.57% 78.13% 99.14% 80.00% 50.35% 84.51% 78.50%Test 0.063 8.0926 25.179 3.4393 network9 Normalize Data 89.65% 74.40% 86.49% 83.01% 83.39% 90.10% 72.42% 88.42% 81.59% 83.13%Learn 0.069 7.6787 23.883 3.317 Real Data 99.06% 81.27% 46.54% 85.53% 78.10% 99.11% 79.91% 49.61% 84.41% 78.26%Test 0.066 8.1274 25.554 3.4602 network10 Normalize Data 88.55% 74.36% 86.32% 82.82% 83.01% 88.42% 71.48% 88.12% 80.72% 82.18%Learn 0.077 7.6893 24.101 3.3541 Real Data 98.96% 81.24% 46.05% 85.37% 77.90% 98.96% 79.24% 41.01% 83.67% 75.72%Test 0.077 8.4018 29.914 3.6241
  • 16. Real Data 𝑁𝑜𝑟𝑚𝑎𝑙𝑖 = 𝑋𝑖 − 𝑀𝑖𝑛𝑖 𝑀𝑎𝑥𝑖 − 𝑀𝑖𝑛𝑖 Formula 2 18 ⇒ 𝑋𝑖 = 𝑁𝑜𝑟𝑚𝑎𝑙𝑖 × 𝑀𝑎𝑥𝑖 − 𝑀𝑖𝑛𝑖 + 𝑀𝑖𝑛𝑖 Formula 3 %𝐴𝑐𝑐𝑢𝑟𝑎𝑐𝑦 = 1 − 𝐴𝑣𝑔 |𝑋𝑖 − 𝑋 ∧ 𝑖| 𝑛 𝐴𝑣𝑔 𝑋𝑖 × 100
  • 17. Thank You for Watching!

Editor's Notes

  1. no slide master 1. Delete icons 2. Move numbers
  2. no slide master 1. Delete icons 2. Move numbers