Emobot

Watch the whole compass. (bipolar I) Most mental health tools are built to watch for one storm. Bipolar I has two. A depression-tracking tool points one needle in one direction: down. But a bipolar I patient can swing into a depressive episode or a hypomanic one — sometimes within the same month, sometimes within the same week. A biweekly PHQ-9 is built to catch the first storm. It was never built to catch the second one, and it's definitely not built to catch the turn between them. What changes first usually isn't mood — it's what's underneath it. Sleep fragments. Activity climbs or drops. Speech speeds up or flattens out. Those are exactly the signals Emobot is built to track, continuously and passively, from facial affect, voice, and activity — the same passive signals validated at r=0.89 against MADRS in depression, now applied to anxiety and bipolar disorders. Zero effort for the patient after a 3-minute setup. No daily check-ins. The clinician sees the trend, and the patient sees their own trend too — gently nudged back in when it moves in either direction, not just toward depression. This case is a reminder that "monitoring" for bipolar I can't just mean watching for the dip. It means watching the whole compass. If you treat bipolar I patients in an interventional psychiatry setting and want to walk through a case like this one, our team is happy to set up time. Not diagnosis. Not prediction. Not replacement of clinical judgment. Dramatization — footage is AI-generated and does not depict a real patient. Clinical data is from a real, anonymized case shared with consent. #InterventionalPsychiatry #BipolarDisorder #DigitalPhenotyping #PsychiatricCare #TMS

Très fier de présenter nos premiers résultats cliniques aujourd'hui aux Journées Neurosciences Psychiatrie Neurologie, au Palais des Congrès de Paris. 🎉 Suivre une dépression dans la durée reste étonnamment difficile. Entre deux consultations, le patient est seul et le médecin avance à l'aveugle. Notre « Thermomètre de la dépression » mesure la sévérité des symptômes en continu, à partir du smartphone et sans rien demander au patient. Et les données parlent d'elles-mêmes. Confronté aux échelles cliniques de référence, sur trois études prospectives chez des adultes souffrant de dépression majeure et de trouble bipolaire : → r = 0,895 de concordance intra-individuelle avec la MADRS cotée par le clinicien → ρ = 0,834 de sensibilité au changement vs PHQ-9 → ρ = 0,61 à 0,83 sur les 5 échelles MADRS, HAM-D₁₇, PHQ-9, GAD-7, BDI-II Ce qui me touche le plus, c'est ce que ça change pour les patients. Le signal qu'on capte chaque jour rejoint ce qu'un psychiatre observe en consultation, mais en temps réel, sans rien peser sur les épaules de quelqu'un qui va déjà mal. La théorie est derrière nous : l'app tourne déjà sur le téléphone de centaines de patients, notamment aux États-Unis en adjonction aux cures de TMS et d'eskétamine. Les équipes soignantes voient jour après jour si le traitement fonctionne, ajustent plus vite et accompagnent les patients sur la durée plutôt que de les perdre en cours de route. Merci à mes co-auteurs Tanel Petelot et Pr. Renaud Seguier (CentraleSupélec / IETR), et aux équipes cliniques sans qui rien de tout cela n'existerait. Vous êtes au Palais des Congrès ? Venez échanger devant le P-25 👇 #JNPN2026 #DigitalHealth #Psychiatrie #SantéMentale #DeepTech

A TMS patient starts a new protocol. For the next few weeks, most clinics are flying blind until the next visit. Here's what Emobot saw instead. Dr. Carlene MacMillan, MD, FCTMSS,DFAACAP shared a case with us. The patient began an Ampa TMS protocol on April 29. Passive mood monitoring captured the full arc between appointments: → a sharp, immediate lift starting the same day as initiation → that climb sustained over the following days into early May → mood then settling into a stable range and holding there The signal came from passive, multimodal monitoring running quietly in the background — the kind of data that normally just disappears between visits. No surveys. No daily check-ins. The patient installed the app once, in about 3 minutes, then went about their life. Carlene put it better than we could: "It really is mind boggling how this is surfacing the impact of interventions so elegantly." For interventional psychiatry, where every week of response matters, watching the curve respond in near real time changes the conversation with the patient — and the decision about what to do next. Thank you, Dr. MacMillan, for sharing this one. Keep posted for a series of patient cases !

60% of relapses in interventional psychiatry are never caught between visits. Last week, another clinic decided that was unacceptable. The team at Hope Therapeutics went live with Emobot across their TMS, IV Ketamine patients. Not for the technology — for the math: 50% of TRD patients relapse within 6–12 months, and most of those relapses happen in the silence between appointments. A PHQ-9 every 6–8 weeks is a snapshot. These patients needed a thermometer. What kickoff actually looked like: → 3-minute install per patient → On-device AI. Zero surveys. Zero daily burden. → 75% of patients activated — because their physician simply recommended it. The clinic now sees objective signal on every patient, in real time, between visits. Huge thanks to the team who made it real: Rebecca Cohen, MD, FAPA, FCTMSS, Mark Bennett, Courtney Olive, Daniella Indenbaum, Kimberly J., and Emily Ordover LPC LMHC MS— and Jane Green, Zachary Javitt for backing the rollout. Watching a team move this fast tells you where IP care is heading. Here's my question for the psychiatrists reading: how many relapses do you think you're not seeing right now? If you'd rather not guess — comment "demo" and I'll show you the AI live. 15 minutes. #InterventionalPsychiatry #TMS #Spravato #DigitalMentalHealth #Emobot

Late-breaking Emobot and NRx Pharmaceuticals, Inc. Clinical Validation Poster at ASCP 2026 today, 11:45 AM ET.... First public pooled validation of EMOCARE — our investigational, passive, multimodal "Depression Thermometer" — across three prospective observational studies (n = 45, MDD + Bipolar Disorder): → r = 0.895 within-person concordance with clinician-rated MADRS (p = .016) → ρ = 0.834 sensitivity to symptom change vs. PHQ-9 (p < .001) → ρ = 0.61–0.83 across MADRS, HAM-D₁₇, PHQ-9 and GAD-7 What this means — concretely — for interventional psychiatry: Today. The Emobot app is already in patients' hands across the Hope Therapeutics network. Continuous, passive mood, sleep and activity monitoring in the background of the patient's own phone. When the signal detects a dip, the app nudges the patient directly — surfacing the change and prompting them to reschedule with their provider, and when clinically appropriate, to initiate a retreatment. A safety net between visits, with zero patient burden. Tomorrow. Take that same validated signal, surface it to the clinical team, and you get something interventional psychiatry has never had at scale: a true measurement-based, personalized feedback loop. Daily, patient-specific evidence of how a TMS course, an esketamine series, or an emerging psychedelic protocol is actually performing — between visits, not only at them. The intent is concrete: lift response and remission rates, and cut the patient drop-out that is well documented during the acute phase of intensive interventional programs. Huge thanks to my co-authors — Antony Perzo, Michael Sapko, M.D., Ph.D., Pr. Renaud Seguier, Ph.D., and Jonathan C. Javitt, M.D., M.P.H. — and to the team at NRx Pharmaceuticals, Inc. (Nasdaq: NRXP) and Hope Therapeutics. The speed at which we went from partnership announcement in March to ASCP late-breaker today has been remarkable. Full poster attached, study details and PR in comment... #ASCP2026 #InterventionalPsychiatry #TMS #Esketamine #Psychedelics #DigitalPsychiatry #DepressionMonitoring #MeasurementBasedCare