Player is loading...

Embed

Copy embed code

Transcriptions

Note: this content has been automatically generated.
00:00:01
well what you tell me we are thank you i'm a happy to prance skylight too and i'm also really happy to
00:00:08
present after carl's uh because you uh presents the concept of using a i you know by medical setting really well
00:00:14
and that's also what skylight is doing so we're developing an artificial intelligence software
00:00:19
to extract knowledge from a new type of biomedical data which is single cell
00:00:23
data like show what a question we address or what need to address
00:00:28
so there is currently a huge problem in health care and this is the time
00:00:32
point of technologies especially in the context of complex diseases such as cancer
00:00:36
the survival chances of patients of cancer patients regardless of the type of cancer and existing treatments
00:00:42
a largely correlate with the time point of technologies late technologies means for survival
00:00:48
so the challenge here is to find a way to dine all these patients early in the disease
00:00:52
and in cancer any uh other complex diseases is the disease arises we
00:00:57
have in the beginning very few cells that are of normal
00:01:00
and if this is progress is the cell become more and more and take
00:01:03
over the for example blotches of the patient and then patient dies
00:01:07
or diagnostic methods they require certain concentration of uh of markers uh that that
00:01:12
can be measured in the sample any this would be a blood sample
00:01:15
uh you need to have already quite early stage of the disease to detect most of the markers that are currently tested for
00:01:21
with a single cell data we actually get date of high parameter data from years of individual cells
00:01:27
and by applying artificial intelligence analysis to this data we can actually narrowed down very rare cell populations
00:01:33
that are problematic for survival or tried response and this is what we all for
00:01:38
and that we we are very confident that uh we'll find a solution for a
00:01:42
more to uh for module diseases not just cancer but also immune disorders
00:01:46
and infectious diseases what we're actually building isn't a i saw per platform
00:01:51
that true analysts this type of data single cell data and it can be used in multiple contacts
00:01:56
so what we are launching next year is a software for research which will
00:01:59
be used in universities to extract knowledge out of single cell data
00:02:03
then later on we are planning to release the software also for farmer
00:02:06
companies which are increasingly using this type of technology in there
00:02:10
early research and discovery faces and this will become more and more important also in the drug development
00:02:15
phase so we offered him a platform to really interpret the dating automated way in much faster
00:02:21
and then finally and most importantly will we're developing an internal by marker discovery
00:02:26
pipeline so we're using our software and partnerships with clinicians here in switzerland
00:02:31
uh to launch by marker discover projects in various diseases and
00:02:34
find button by markers for early diagnostics of diseases
00:02:38
um we have a very competent experience in motivating that is standing behind the
00:02:42
company i have a life has background and more recently i'm actually selling
00:02:46
single so profiling devices on the swiss markets will have access to all the customers who produce this
00:02:51
kind of data and then we have moderate classes with a professor d. e. h.'s you're it
00:02:56
and he actually developed the algorithm which our product is based in have dahlia and
00:03:00
denise these are two software engineers were developing the software in industry great software
00:03:05
and we have on our our chief financial officer uh with really long your business experience my times all
00:03:11
and what we need a receptor the funding for the research uh part of the
00:03:14
business that we need funding for farmer here yes hi hi mark any right
00:03:26
ah there is a question here here we go
00:03:30
uh_huh differentials now from some thanks oh thank you
00:03:39
so was offered yeah you're right so we have some similarities with software genetic so they also
00:03:43
quite a high and we're going to find all my markers and use it for diagnostics
00:03:47
but the type of data that they're looking at it's actually d. n. a. sequencing data so
00:03:51
they look at d. n. a. sequences and it's not single cell data so that
00:03:54
for example patients sample blood sample that would be a media's of cells you take the
00:03:58
bulk of it and then you extract the d. n. a. out of it
00:04:01
so you lose the resolution of single cell so you don't see individual cells in there you just see all the d. n. a. that is in the sample
00:04:07
and of course you can look for rare mutations there and that's what software genetics is doing what
00:04:11
we have a completely different approach so we're actually looking at the cells our approaches so based
00:04:16
and that's what differences from so few genetics in other companies that are looking at d. n. a.
00:04:20
sequences and to actually the first day that we're looking is actually porting data not uni
00:04:25
right fella criteria

Share this talk: 


Conference Program

elthera
Oct. 25, 2017 · 11:12 a.m.
257 views
SunRegen
Oct. 25, 2017 · 11:16 a.m.
329 views
GRATXRAY
Oct. 25, 2017 · 11:20 a.m.
272 views
certus diagnostics
Oct. 25, 2017 · 11:29 a.m.
222 views
Trackcath?
Oct. 25, 2017 · 11:32 a.m.
retinai
Oct. 25, 2017 · 11:36 a.m.
334 views
scailyte
Oct. 25, 2017 · 11:40 a.m.
241 views
Ophtho robotics
Oct. 25, 2017 · 11:45 a.m.
233 views
Dicronis
Oct. 25, 2017 · 11:49 a.m.
318 views
IDUN
Oct. 25, 2017 · 11:53 a.m.
197 views
steriLux
Oct. 25, 2017 · 11:56 a.m.
143 views
T-CHAIR
Oct. 25, 2017 · 12:01 p.m.
129 views
AESYRA
Oct. 25, 2017 · 12:04 p.m.
476 views
FlavorWiki
Oct. 25, 2017 · 12:09 p.m.
367 views
Algrano
Oct. 25, 2017 · 12:12 p.m.
498 views
BioSens
Oct. 25, 2017 · 12:17 p.m.
171 views
Farmerline
Oct. 25, 2017 · 12:21 p.m.
104 views

Recommended talks