UNIGE Faculty of MedicineFacMed Hack #1Digital Phenotyping of Mental Health
FR

The challenges

6 challenges, grown from the needs you submitted. Rank the 3 you want to work on.

We do our best to follow everyone’s wishes, but voting guarantees neither a place in one of your preferred challenges nor a seat at the hackathon.

Sign in to rank the challenges

  1. #1
    Physiology you can trust: tracking change between visitsBetween two consultations, is a patient getting better or quietly getting worse?
    More

    Between two consultations, is a patient getting better or quietly getting worse? Wearables and phones stream heart rhythm, sleep and movement around the clock; the hard part is trusting them and telling stress from a coffee, a run or a bad night. The team will turn these everyday signals into a measure of change from each person's own baseline, decide who should hear about a meaningful shift and when, and keep alert fatigue out of the loop. Open wearable datasets, foundation models for physiological signals and language models are on the table, along with the outline of a first study with patients. One lead worth chasing: the body clock as an early signal.

    Resources

    Data

    Skills sought

    Cliniciendeveloperscience communicator or designerSignal processingainsi que des cliniciens dans la santé mentalawareness of medical-device rulescardiologistCardiologyCliniciananxiety specialistsapp developmentclinician who treats anxietydesignergeriatrician or neuropsychologistmachine learningmixed-effects or Bayesian hierarchical modellingmobile developerneurological conditionsbiosignal foundation modelswearable signal quality

    Contribute →

  2. #2
    Beyond the human eye: behaviour from video and audioSome of the most telling signs in mental health last less than a second: a micro-movement, a glance held too long, the moment words and face tell different stories.
    More

    Some of the most telling signs in mental health last less than a second: a micro-movement, a glance held too long, the moment words and face tell different stories. Trained clinicians catch some of them; coding them by hand is slow and varies between coders. Here, video and audio of play, conversation and therapy become behavioural signatures precise enough to follow one person across sessions, a step towards precision and personalised medicine. Over two days, vision-language models, audio-visual models and pose tracking meet open footage and human coders, and clinicians question and correct the labels in plain language. Leads: change across sessions, movement patterns shared by overlapping conditions.

    Resources

    Skills sought

    AI expertbehavioral sciencebias auditingclassificationcomputer vision posespeech / audio MLmultimodal / vision-language modelshuman–computer interaction3D reconstruction / synthetic datadevelopmental behaviour coding

    Contribute →

  3. #3
    What speech reveals: memory and mood in clinical audioA recorded memory, a therapy session, a parent playing with a child: each carries pauses, prosody and word choices that say something about memory and mood, and no clinician can score them at scale.
    More

    A recorded memory, a therapy session, a parent playing with a child: each carries pauses, prosody and word choices that say something about memory and mood, and no clinician can score them at scale. How much of that signal can be captured from French clinical audio, compared over time, and shared without exposing the speaker? The team will build the pieces: who spoke when, voice and language markers, and anonymisation that protects identity while keeping the signal. Speech foundation models, language models and on-device processing work on open corpora, and the two days end with a plan to bring them to the clinic. Still wide open: markers that survive a phone call.

    Resources

    Skills sought

    NLPmachine learningspeechapp developmentNeurosciencesoftware developmentaccessibilityaudio engineeraudio ML skillscareful alert design that prompts without alarmingspeaker diarisationvoice and text anonymisationFrench clinical linguisticsneuropsychology of memory

    Contribute →

  4. #4
    From camera to cortex: everyday markers, anchored in the brainPicture a frail patient at home, a phone propped on the table, and measures of gaze, pupil and gait as good as a lab's.
    More

    Picture a frail patient at home, a phone propped on the table, and measures of gaze, pupil and gait as good as a lab's. Eye trackers, motion capture and EEG rarely leave the lab, and many patients can rarely come to them. This challenge brings lab-grade measures to ordinary cameras, checks them against lab references, and anchors them in brain activity, with EEG kept clean while people talk and walk and pipelines that hold across sites. Markerless tracking, EEG foundation models and open EEG and imaging data are the starting kit for a first prototype and a study plan. The open question at its heart: what neurophysiology/imaging adds beyond the camera.

    Resources

    Technical

    Skills sought

    computer visionsoftware developmentbiomechanicsComputer-vision engineerdata analysisData scientisteye trackingEye-tracking analystApp developmentclinical psychologistclinical sessionsClinical trial coordinatorclinicianclinician to keep the session tolerablecoding skillscoding skills e.gComputational skillsData engineer for multimodal alignmentmarkerless pose / webcam eye trackingEEG signal processingMRI segmentation / neuroimagingmultimodal synchronisation (Lab Streaming Layer)neuromodulation

    Contribute →

  5. #5
    Pain, pleasure and touch: one readout from mice to humansCan we read pain and pleasure the same way in a mouse and in a person, without touching either?
    More

    Can we read pain and pleasure the same way in a mouse and in a person, without touching either? It matters for both: handling stresses animals, while rating scales miss what patients cannot put into words. The aim is one hands-off readout of pain, pleasure and social behaviour that travels across species, from pose, face and movement on camera, with touch delivered by robots and behaviour discovered with as little human bias as possible. Pretrained pose models, behaviour transformers and vision-language models start from open video; the two days end with a prototype and the outline of a translational study. Barely explored so far: how sight and sound reshape touch.

    Resources

    Data

    Skills sought

    3D printingpose estimationaccess to HPCadapting existing modelsanalysing errorsanimal welfare or veterinary inputAudioaudio processing — diarisationbehavioral science. Knowledge on AI VLMcoding — implementingrobotics / stimulus deliveryunsupervised behaviour discoverycross-species transfer learningsomatosensory neuroscience

    Contribute →

  6. #6
    Interventions that measure: XR, games and smart toysWhat if every moment of play could also be a measurement?
    More

    What if every moment of play could also be a measurement? A game in augmented reality, a virtual scene for practising social cues, a toy that responds to how a child plays: each can train attention, social skills or emotion recognition, and each can record gaze, movement, choices and physiology along the way, so clinicians see more than a final score and can tailor the next session. Headsets with built-in eye tracking, language-model characters and sensor-equipped toys are within reach for a two-day build, together with a plan to test the best idea with patients. Worth a first look: augmented-reality cues for attention after stroke.

    Resources

    Data
    Technical

    Skills sought

    audioBroad knowledge in neurosciencechild-centered iterative design with therapistclinicienComputational skillscomputer visiondata analysisdata analystedata managementdata managerXR development (Unity, Godot, WebXR)game designembedded / tangible hardwareconversational AIneurorehabilitation

    Contribute →

← website, children, informationdecline, depression, early →↑ gaze, scene, camera11. Physiology you can trust: tracking change between visits22. Beyond the human eye: behaviour from video and audio33. What speech reveals: memory and mood in clinical audio44. From camera to cortex: everyday markers, anchored in the brain55. Pain, pleasure and touch: one readout from mice to humans66. Interventions that measure: XR, games and smart toys
  1. 1Physiology you can trust: tracking change between visits
  2. 2Beyond the human eye: behaviour from video and audio
  3. 3What speech reveals: memory and mood in clinical audio
  4. 4From camera to cortex: everyday markers, anchored in the brain
  5. 5Pain, pleasure and touch: one readout from mice to humans
  6. 6Interventions that measure: XR, games and smart toys
Each challenge grew out of the needs you submitted. A bubble sits at the centre of its needs (the small dots, anonymous) and its size says how many it gathers. Challenges close together address related questions. Hover a bubble or a title for its pitch and the area its needs cover, click to go to the challenge.