Spermiogenesis or Spermateleosis or metamorphosis of spermatid
Ai based mobile application to fight antibiotic resistance ppt
1. AI-BASED MOBILE APPLICATION TO
FIGHT ANTIBIOTIC RESISTANCE
KOLLA SRIVALLI
201709010
M.SC. SYSTEMS BIOLOGY
MANIPAL SCHOOL OF
LIFESCIENCES
2. INTRODUCTION
• Antibiotic resistance is the ability to defeat the drugs which are designed to kill the
microorganisms.
• Antimicrobial resistance occurs when microbes evolve mechanisms that protect them
from the effects of antimicrobials.
• Techniques which are used to detect the antibiotic resistance of microorganisms
(Antibiotic Susceptibility Tests) are –
- dilution method
-disc diffusion method
-E-test
-genotypic methods such as PCR and DNA hybridization methods
3. DISC DIFFUSION
TECHNIQUE
Fig 1 : Disc diffusion
technique
• Principle – the antibiotic in the disc get diffused
radially into agar and inhibit the growth of
microorganisms in the area till where the diffusion
occurs.
• Zone of inhibition can be used to measure the
susceptibility of bacteria to antibiotic.
• Presence – bacteria are susceptible to antibiotic
- suitable antibiotic
• Absence – bacteria are resistant to antibiotic
- not suitable antibiotic
4. DISADVANTAGES-
1. Labor intensive and time consuming.
2. Require proficient technicians for quality plate
preparation and for measuring the diameter of
zone of inhibition.
3. Interpretative reading is required based on rules
published by EUCAST ,Europe or CLSI.
EUCAST – European Committee on Antibiotic
Susceptibility Testing
CLSI – Clinical and Laboratory Standards Institute
Fig 2 : EUCAST
5. DEVELOPMENTS IN
TECHNIQUE
Fig 3: mobile application
developed against antibiotic
resistance
Manual
AntibiogramJ
(used in
computers)
Antibiogo
(mobile app)
• The need for the development of mobile
application is usually identified by Medicins Sans
Frontiers (MSF).
• MSF often operates in low and middle income
countries(LMIC) where AST is difficult to
implement.
6. WORKING OF APP
Interpretation of results
Image analysis
Acquisition setup
(for clicking clear picture)
Making of antibiograms
Fig 4 :working model Fig 5 : result
interpretation
7. CREATION OF DATA SETS
• AST groups A1 and A2 consist of 570 and 75 antibiograms prepared during working
routine in the microbiology laboratory of the University Hospital in Creteil, France.
• AST set A3 consists of eight Petri dishes prepared in the Hospital of Medecins Sans
Frontieres in Amman, Jordan. In the case of this set, the plates were inoculated with
microorganisms purchased from the American Type Culture Collection (ATCC) and
routinely used for quality control.
8. IMAGE ANALYSIS
• If for more than two antibiotics in the
picture we found an absolute diameter
difference between the App and control
values of more than 20 mm, we
considered the picture problematic,
otherwise, it was considered standard.
• The problematic images are often
associated with plates with defects or
show very low inhibition-to-bacteria
intensity contrast.
Fig 6: (a) and images with very poor visible
contrast between the bacteria and the inhibition
(b) Some inhibition zones are hard to isolate,
even by eye.
(c) The standard image
Susceptibility characterization :
• Very major disagreement – categorization of antibiotic as susceptible(S) while control is resistant (R)
• Major disagreement –categorization of resistant (R) with control susceptible(S)
• Minor – categorization of error involving intermediate(I)
9. ABNORMAL IMAGES
A1) non-circular inhibition shape
A2) complete (or) no inhibition
A3) light reflection
A4) colonies within inhibition zone
A5) double inhibition zones
A6) hazy borders
A7) low contrast
A8) defined as the difference between the
inhibition and bacteria intensity value, compared to
a high contrast
A9) image of dataset
Fig 7 : images of dataset and some
of its problematic discs visualization
10. LINKS FOR ACCESS
• Data is available at http://stat.genopole.cnrs.fr/ast.zip
• App can be downloaded from https://form.typeform.com/to/qEGVBzbu with a request
form.
• Live performed demo video is available in https://youtu.be/0hNr9zTu6ig .
• For the matter of hardware compatibility, as of now they have tested three
smartphone models (Google Pixel 3A, Honor 6x,Samsung A10) ranging from high-
to low-end.
• The whole reading of one antibiogram (12 megapixels picture, 16 antibiotic disks)
takes less than 1 s on a PC using one 2.3 GHz Intel Core i5 processor, 1.5 s on a
high-end smartphone (Pixel 3 released in 2018), and 6.6 s on a low-end smartphone
(Samsung A10), still much faster than manual reading.
11. REFERENCES
• EUCAST. Eucast disk diffusion method. European Committee on
Antimicrobial Susceptibility Testing (EUCAST)
https://eucast.org/ast_of_bacteria/disk_diffusion_methodology
• Hudzicki, J. Kirby- bauer disk diffusion susceptibility test protocol
https://www.asmscience.org/content/education/protocol/protocol .