When I had my annual physical review a few days ago, my family doctor introduced me to Heidi. Heidi is not a person. Heidi is an artificial intelligence (AI) medical scribe developed by the Australian technology company Heidi Health. My doctor told me that he turns Heidi on for it to listen when conversing with a patient.
Heidi utilizes ambient listening technology to transcribe conversations in real-time. It understands different dialects and accents and medical terminology with minimal error rates.
Before we started the appointment, my doctor asked if I was okay having Heidi listen in to our conversation. Being the computer nerd I am, I said absolutely. He then turned the AI on, and we began to discuss the latest medical results from my recent bloodwork. We went on to discuss other medical concerns. Heidi listened.
This AI is a transcriber of spoken conversations. It has clinical awareness of medical conditions. If information is not discussed in a conversation, my doctor can add these details to Heidi’s clinical notes after the medical appointment, simply by dictating to the AI, which then generates a clinical report.
As a transcriber, Heidi generates results in seconds. The AI can be integrated with hospital and patient electronic health records. It can generate referral letters that my doctor can send to specialists.
He told me it is saving him time, reducing paperwork, and lessening the administrative nightmare that accompanies general practice.
Heidi works on or offline. Transcriptions and final clinical reports are encrypted.
Relatively new, Heidi is already making a difference in medical practices across the globe, handling over 2 million patient consults per week in more than 50 countries and covering over 300 specialties.
Heidi Has Competition
It is no wonder that physicians are looking to AI to help them in their medical practices. As my family doctor showed me what Heidi was doing for him, I must admit I was impressed.
Is Heidi a one-off, or are there other AI software developers chasing the same market? It didn’t take me long to find out that AI scribes are a burgeoning software market niche.
Heidi is offered free or in a Pro version that costs US$99 per month. How does it compare to other AI scribes? I did a little digging to come up with this list and links for you to check out:
- Abridge describes itself as an “enterprise-grade AI for clinical conversations.” Focuses on powerful multilingual clinical note generation and medical coding integration, though pricing is on the higher end. A subscription costs US$208 per month.
- Augmedix has three versions: Go, Assist and Live. All three assist in creating medical clinician notes. Go pricing starts at US$9 per user per month.
- Chartnote provides four different subscriptions. Basic provides a minimal scribe for free. Subscriptions are monthly with Premium at US$4.99, Professional at US$15.99 and Max at US$99.99.
- DAX Copilot pricing is US$369 per month per user with a US$700 one-time setup fee.
- DeepScribe is priced at US$750 per month.
- Empathia offers EMMA as its AI scribe with a free trial, Essential version at US$84 per month and Premium at US$127 per month.
- Fathom is priced beginning at US$14 per month.
- Freed offers a free trial subscription with pricing for practices between 2 and 50 at US$84 per month.
- JotPsych is a medical scribe built for behavioural health. You can try it for free with individual user pricing per month at US$150, and group practice rates from 7 to 50 at US$130 monthly per user.
- Nabla offers a free trial for Copilot and AI Scribe, with the latter subscriptions starting at US$119 per provider.
- NoteMD offers a free trial and a Pro version for small and medium medical practices at US$99 per month.
- Scribeberry offers its dictation and transcription tool in a free trial or Pro version for US$99 per user per month.
- ScribeHealth offers a free 20-session one-week trial with pricing per user at US$49 monthly.
- Sonix can be tried for free with subscriptions beginning at US$22 per month.
- Suki offers two products: Assistant with full electronic health record integration at US$399 per user per month, and Compose medical dictation at US$299 per user per month.
- Tali is free a AI scrbie for personal use with a Premium subscription at US$45 per month, and Pro for US$100 per month.
- Twofold Health offers its scribe for free with Personal and Custom subscriptions beginning at US$49 per user per month.
- Vero Scribe offers its application for US$59.99 per user per month.
Transcription Isn’t The Only Thing Medical AI Can Do
- Radiology to interpret images,
- Pathology to identify disease agents,
- Wearable devices for continuous patient monitoring,
- Research to help with breakthrough gene therapies and new pharmaceutical discoveries using tools like AlphaFold.
AI is diagnosing diseases faster than specialists and general practitioners. It anticipates needed interventions before medical crises unfold by detecting subtle changes in patients that a doctor may not recognize. In administration, besides being a scribe, AI tools are streamlining schedules, intake, and triage activities.
It can do all these things because the Large Language Models (LLMs) that train use tens of millions of medical records containing imaging, clinical notes, and test results from doctors, researchers, and clinicians.
Where Medical AI is Heading
With all that AI can do presently, what is expected of it in the future? The following areas of medical practice that AI will impact the most include:
- AI Virtual Assistants: An AI will continuously monitor patient data and alert healthcare providers to potential problems. It will send personal reminders to patients to take medications, to pharmacists for automatic refilling of prescriptions. It will act as a coach, provide tailored health education, explain lab results, and guide patients through rehabilitation, assist with mental health issues by identifying patients in crisis and alerting physicians, while providing basic therapy support. It will cross language and disability barriers (hearing and sight) in support of patients through speech-to-text and translation.
- Predictive Medicine: Using advanced algorithms and machine learning, AIs will use predictive analytics for early identification and intervention in at-risk populations, addressing disease predictions and assessing individual risks to create personalized treatment plans.
- Digital twins: AI will help physicians create virtual online twins of patients incorporating their health record, wearable data, blood, ultrasound, MRI, CT scan, X-ray and genetic tests, and hospital and clinical notes, viewable in real-time.
- Precision Medicine: Analyzing large, complex data records will allow AI to find subtle patterns in patients to help with more accurate diagnosis, and proactive personalized treatment.
- Provider Empowerment: Besides clinical notes, AI will streamline the administrative side of medical practices, dealing with scheduling, billing, insurance, and the maintenance of electronic health records.
-
AI Ecosystems: An AI-driven ecosystem will facilitate seamless data sharing among all members of the medical community, including doctors, pharmaceutical companies, druggists, researchers, outpatient diagnostic clinics, laboratories, and hospitals.
-
Improved Healthcare Access: AI will extend healthcare reach to manage global workforce shortages and serve remote patients more effectively than current medical delivery systems.
