Science Animated – Explaining Uses of AI in Medicine

Science meets media design in an interdisciplinary project to strengthen the exchange about applications of artificial intelligence.

How can AI sup­port the selec­tion of suit­able med­ic­a­tion, make med­ic­al dia­gnoses more effi­cient or help in the ana­lys­is of MRI images? Tech­no­lo­gies based on arti­fi­cial intel­li­gence are gain­ing more and more influ­ence in vari­ous areas of med­ic­al research and prac­tice – and are becom­ing increas­ingly rel­ev­ant for our society.

To address this devel­op­ment, the RHET AI Cen­ter col­lab­or­ated with part­ners from medi­cine and anim­a­tion research to devel­op an inter­dis­cip­lin­ary teach­ing cooper­a­tion aimed at stim­u­lat­ing dia­logue and dis­cus­sion: Sci­ence Anim­ated – Explain­ing Uses of AI in Medi­cine. For one semester, ten stu­dents from the fields of medi­cine, bio­logy, media stud­ies and rhet­or­ic worked togeth­er in inter­dis­cip­lin­ary teams. Their com­mon goal: strength­en­ing their skills in visu­al sci­ence com­mu­nic­a­tion – and fos­ter­ing the pub­lic dis­course on AI applic­a­tions with vivid videos. As a res­ult, a total of four explain­er videos on the top­ic of "AI in medi­cine" were created.

The explainer videos

Dur­ing the devel­op­ment of the videos, accom­pa­ny­ing work­shops and peer feed­back ele­ments provided a struc­tur­ing frame­work for the design pro­cess. How­ever, the stu­dent teams were com­pletely free in their spe­cif­ic focus, artist­ic design and choice of tar­get group. While one team, for example, set itself the goal of appeal­ing primar­ily to chil­dren ("A Voy­age Through a Crowded Cell World"), oth­er teams focused on broad­er tar­get groups. All of this con­trib­uted to the diversity of the cre­at­ive pro­cesses and res­ults – and led to four unique videos:

Which Pill to Choose?
How Machine Learn­ing Can Improve Med­ic­al Pre­scrip­tions
by Felix Freuer / Wikt­or­ia Palka

How Does Dis­trib­uted Ana­lyt­ics Work?
by Alex­an­der Flisiak / Aina Segura

A Voy­age Through a Crowded Cell World:
How Artific­al Intel­li­gence May Sup­port Med­ic­al Dia­gnos­is
by Neus Gil Noguera / Kim Ben­jamin Rösen­er / Min Zhou

Machine Learn­ing in Radi­ology:
How Neur­al Net­works Can Help in Ana­lyz­ing Med­ic­al Images
by Den­nis Brun­neck­er / Alex­an­der Kempf

Strengthening the dialog on applications of artificial intelligence in medicine

The explan­at­ory videos developed by the stu­dents are inten­ded as an invit­a­tion to fur­ther exchange – and to pro­mote dia­log between sci­ence and soci­ety. The videos have already been presen­ted in a vari­ety of con­texts and formats: In a first step, they were shown as part of the "Cyber and the City" exhib­i­tion at the Tübin­gen City Museum (Feb­ru­ary 2023 – Janu­ary 2024). There, togeth­er with oth­er exhib­its, they pur­sued the goal of mak­ing applic­a­tions of arti­fi­cial intel­li­gence tan­gible and cre­at­ing a space for dis­cus­sion. In a fur­ther step, the videos also found their way onto the big screen: the par­ti­cip­at­ing stu­dents and the course lead­er team presen­ted the videos at an event at the Arsen­al cinema in Tübin­gen in April 2023. In addi­tion to the present­a­tion of the explan­at­ory videos, there were also insights into the cre­at­ive pro­cesses behind the videos as well as a vari­ety of oppor­tun­it­ies to talk about the top­ic and join in the discussion.

Direct links to current research

In order to estab­lish dir­ect links to cur­rent research top­ics, four extern­al experts research­ing AI applic­a­tions in medi­cine presen­ted key aspects of their work to the par­ti­cip­at­ing stu­dents at the start of the course. These impulses formed the start­ing point from which the stu­dents developed their respect­ive explan­at­ory videos in inter­dis­cip­lin­ary teams. The four experts were also avail­able to advise the stu­dents – in addi­tion to over­all guid­ance giv­en by the course lead­er team – on research related ques­tions dur­ing the devel­op­ment of the videos:

• Dr. Stephanie Bier­gans
  (Uni­ver­sity Hos­pit­al Tübin­gen):
  Data Man­age­ment & Dis­trib­uted Analytics
• Prof. Dr. med. Ser­gios Gatid­is
  (Max Planck Insti­tute for Intel­li­gent Sys­tems, Tübin­gen):
  Deep Learn­ing in Med­ic­al Imaging
• Dr. med. Alex­an­der Toli­os
  (Med­ic­al Uni­ver­sity, Vienna)
  AI Mod­el­ling in (Pre)Clinical Settings
• Prof. Dr. Man­fred Claassen
  (Uni­ver­sity Hos­pit­al Tübin­gen):
  Machine Learn­ing in Trans­la­tion­al Single-Cell Biology

Background: a (teaching) cooperation between rhetoric and medicine

The prac­tic­al course was part of the sem­in­ar series Visu­al Sci­ence Com­mu­nic­a­tion in Medi­cine offered since the sum­mer semester 2021 by Michael Pelzer (Research Cen­ter for Sci­ence Com­mu­nic­a­tion) and PD Dr. Markus Löffler (Uni­ver­sity Hos­pit­al Tübin­gen). The inter­dis­cip­lin­ary teach­ing and skills devel­op­ment concept developed as part of this sem­in­ar series formed the basic frame­work of the course.

With regard to the spe­cif­ic train­ing of stu­dents in the field of cre­at­ing explan­at­ory videos, the course lead­er team was fur­ther sup­ple­men­ted and sup­por­ted by Naima Alam (Insti­tute of Media Stud­ies), who con­trib­uted her extens­ive expert­ise in the field of anim­ated film design. Alex­an­dra Geßn­er (Machine Learn­ing ⇌ Sci­ence Col­ab­or­at­ory) was also involved as an expert in the field of machine learn­ing in the first half of the semester – and provided valu­able insights and advice on the spe­cial­ist background.