Transcript NEW APPLICATION INFRARED SPECTROSCOPY IN MEDICAL DIAGNOSIS

NEW APPLICATION OF
INFRARED SPECTROSCOPY
IN MEDICAL DIAGNOSIS
Prof. János Mink
Department of Molecular Spectroscopy
Institute of Structural Chemistry
Chemical Research Center of the
Hungarian Academy of Sciences
H-1025-Budapest
Pusztaszeri str. 59-67.
Chemical Research Center of HAS
Department of Molecular Spectroscopy

Laboratories:




Infrared and Raman Spectroscopic Lab.
ESR Spectroscopic Lab.
NMR Spectroscopic Lab.
Laser Spectroscopic Lab.
 Professors:







László Biczók
Gábor Keresztury
János Mink
Lászó Nemes
Antal Rockenbauer
Tamás Vidóczy
28 – 16 young scientists / PhD students
University of Pannonia (Veszprém)

Research group: 5 (4) PhD / young scientists

PhD School: Material Science and Technology
Infrared and Raman Laboratory
Budapest:






2 FT-Raman systems
FTIR Microscope with FPA detector (micro ATR)
FTIR Microscope
2 FTIR spectrometer (ATR)
FTIR emission system
FTIR-GC system
Veszprém:




FTIR research grade
2 FTIR routine
Far-IR system
Open-path FTIR
S 13 spectrometer
Extended Collaborations

Surface study: ATR; IRRAS; DRIFTS; EMISSION; LBfilms; coatings; polymer surfaces; dispersed catalysts;
archeological samples


Analytical studies: HPLC-FTIR; GC-FTIR;TLC-Raman
Environmental studies: atmospheric pollutants;
aerosol-microscopy

Theoretical studies: DFT; ab initio; Chemometrsics;
etc.

Medical applications: human hair and skin
One fingerprint
Four fingerprints
Discovery of the spectral effects
Hair:
 Narcotics in hair
 Effects of bleaching
 Effects of UV-radiation
Skin:
 Moisture content
 Lipid content
 Effects of cosmetics
(Literature)
Unexpected spectral
differences
Spectral parameters
• Band position
• Band intensities
• Band widths
• Spectral subtraction
• First derivative
• Second derivative
• Band deconvolution
Statistical data processing
• artificial neural network (ANN),
• principal component analysis (PCA) and factor analysis (FA),
• principal component regression (PCR),
• partial minimum squares (PMS),
• Cluster analysis (CA),
• generic algorithms (GA).
SKIN
Difference IR spectrum of two
healthy patients is very small
1.4
1.2
Absorbance
1
.8
.6
.4
.2
difference
0
1750
1700
1650
1600
Wavenumber (cm-1)
1550
1500
1450
Spectral deviations between
“healthy” and “ill” patients,
exhibiting different RIAR values
.2
1550 RIAR
.15
720 RIAR
20 RIAR
.1
Arbitrary
.05
0
-.05
-.1
-.15
1700
1600
1500
1400
1300
-1
Wavenumber(cm )
1200
1100
1000
900
RIAR - Relative Infrared Analytical Risk
RIAR
values
Classifications
Category of healthy
condition
0-400
Healthy
1
400-600
Weakly endangered
2
600-800
Endangered
3
800-1800
Highly endangered
4
Deviation of a healthy patient’s spectra from the
averaged spectra of healthy patients
(minimum 20-40 persons)
RIAR – Relative Infrared Analytical Risk
180 0
A ra nk a
160 0
140 0
R IA R
120 0
100 0
80 0
60 0
40 0
20 0
0
1
3
4
7
9
da y
10
11
12
14
Process of recovery monitored by infrared spectra of the
skin belonging to a patient suffering in pancreas cancer.
Green line is the limit value of healthy patients
1800
1704
1600
1400
1220
RIAR
1200
1000
800
600
400
164
200
0
1
64
day
147
Distribution between different types of
diseases of 820 “ill” patients measured by
FTIR spectra of their skin
Diseases of cardiovascular system
287 persons
(35%)
Cancerous diseases
270 persons
(33%)
Diseases of gastrointestinal segments
82 persons
(10%)
Lesions of nervous system
74 persons
(9%)
Diabetes
67 persons
(8%)
Allergy
40 persons
(5%)
Total
820 persons
Results of measurements for 380 patients
declaring themselves as healthy in
accordance of their genial state of health
Qualification
Numbers
Percent
Healthy or weakly
152
endangered
40%
Endangered
137
36%
Highly
endangered
91
24%
Results of 39 diabetic patients and 59
healthy persons analyzed by Principal
Component Analysis (PCA)
0.6
healthy
0.4
0.2
0
diabetic
-0.2
-0.4
-0.6
-0.8
-1
-1.2
-1.4
8
8.2
8.4
8.6
8.8
9
9.2
HAIR
Infrared spectra of untreated (lower trace) and bleached (upper trace) female hair sample.
The difference spectrum is given below (dashed line, with 5x ordinate expansion)
Absorbance
Comparison of human hair infrared spectra of a
healthy (lower trace, full line) and a breast cancer
patient (upper trace, dashed line)
0.5 A.U.
1800
1700
1600
1500
1400
1300
1200
Wavenumber(cm-1)
1100
1000
900
Difference spectrum of healthy and
breast cancer patient’s hair samples
Absorbance
0.02 A.U.
1800
1700
1600
1500
1400
1300
1200
Wavenumber(cm-1)
1100
1000
900
Plot of PCA scores for infrared
spectra of hair samples
◊-Healthy female scores (negative), □-healthy male scores (positive), ○-breast
cancer patients’ scores
Partners
• Stockholm University (Electron Microscopy)
• Institute für Innovative Diagnostik (Ulm)
• Municipal Clinic (XVI District, Budapest)
• LaborTrading Ltd./Varian (Instrumentation)
PRE-DIAGNOSIS BY OPTICAL METHODS
Scheme I
Chemometric
(Statistical data processing)
Preliminary evaluation
Measurement
(Involves WP 1.1 - WP 1.9 work packages)
EXPERIMENTAL METHODS IN LABORATORIES
WP 1.1
MEDICAL CONTROL IN HOSPITAL
IR measurements
WP 1.3
Skin
Hair
Patient
Full diagnosis
and therapy
(in situ)
(laboratory)
Repeating the
measurement
checking the
diagnosis
Categories of health conditions
WP 1.2
WP 1.4
Skin
1
2
3
4
Hair
1
2
3
4
Preliminary comparison
with medical control of
health conditions
(4 categories)
Classification
RIAR software
Disagreement
Agreement
Category 1 = „healthy”  library (averaging)
(Selected samples)
EM structures
(Selected samples)
Chemometric
Categories 2-4
Chemometric
Structures
Classes
of similarities
Classes
of similarities
Classes
of similarities
Hair
B
C
D
-
Hair
A
B
C
D
-
IR spectra
WP 1.5
WP 1.6
WP 1.7
Skin
A
B
C
D
-
WP 1.8
Category 2-4
Hair
B
D
E
-
IR images
Illness specific libraries
WP 1.9
SUMMARY OF THE
RESULTS
SUMMARY
OF TWO
THEYEARS
RESULTS
Illness diagnosis from hospitals
Type of illnesses
A
B
C
D
E
F
Transmission FT-IR imaging of
human hair
Cuticle
Cortex
Medulla
A
B
Images based on the absorption of the CH stretching band (collagen) (A),
the Amide I band (protein) (B), of hair
ADVANTAGES




Detection (health condition)
Prevention (early diagnosis)
Monitoring
Diagnosis (under investigation)
CHALLANGING






Simple methods
Quick measurement (hair: ~1 min, skin: ~ ½ min)
Not invasive and painless
Relative cheap
Screening test in mass measures
Patients showed very positive attitudes
PhD students and
young co-workers
László Kocsis
Éva Pfeifer
Veronika Gombás
Viktória Komlósi
Bernadett Illés
László Hajba
Co-workers
Zsuzsanna Kovács
Dr. Judith Mihály
Csaba Németh
Dr. Zoltán Bacsik
Dr. Gábor Keresztury
Katalin Módly
Dr. Károly Héberger
It was my great pleasure to
be with you:
Thanks for Jenny, Carolyn, Jennifer, Kevin, Imre, Glenn,
Chris, Mark, George, Baijn and Sam
Special thanks for April H. Foley, Ambassador of
United States
Thanks for your kind
attention!!!