Telemedicine Session Audio Transcripts

This dataset provides detailed, time-stamped transcripts of telemedicine consultations, capturing speaker roles, utterance sequences, and session metadata for each interaction. Designed for training and evaluating healthcare NLP assistants, it enables research in medical dialogue understanding, intent recognition, and clinical documentation automation while supporting privacy and compliance requirements.

  • last updated 21 Jul 2025
  • by GoMask
The brief that made it

Training conversational AI and NLP assistants for healthcare applications

Sample rows

preview · 8 of 200 rows · all 19 columns
session_idstringprovider_rolestringutterance_confidencefloatis_sensitivebooleanmedical_specialtystringtranscript_idstringsession_start_datetimedatetimesession_end_datetimedatetimepatient_idstringprovider_idstringlanguagestringutterance_orderintegerspeaker_typestringspeaker_idstringutterance_start_timefloatutterance_end_timefloatutterance_textstringsession_notesstringtranscript_sourcestring
sess001physician0.97falsefamily medicinetrans0012023-10-01T09:05:00Z2023-10-01T09:38:00Zpat8a1e9dprovb21k7men1providerprovb21k7m04.5Good morning, how are you feeling today?Patient reports mild headache.manual
sess001physician0.95falsefamily medicinetrans0022023-10-01T09:05:00Z2023-10-01T09:38:00Zpat8a1e9dprovb21k7men2patientpat8a1e9d4.58.7I am okay, just a bit of a headache since yesterday.Patient reports mild headache.manual
sess002nurse0.89falseendocrinologytrans0032023-10-02T13:20:00Z2023-10-02T14:05:00Zpatf49k2vprovx31q7bes1providerprovx31q7b03Buenos días, ¿en qué puedo ayudarle?Consulta sobre control de glucosa.asr
sess002nurse0.82falseendocrinologytrans0042023-10-02T13:20:00Z2023-10-02T14:05:00Zpatf49k2vprovx31q7bes2patientpatf49k2v35.6Me cuesta mantener el azúcar estable.Consulta sobre control de glucosa.asr
sess003therapist0.91truepsychiatrytrans0052023-10-05T15:45:00Z2023-10-05T16:30:00Zpatx29yj4provp67q2nfr1patientpatx29yj406.2Je me sens très stressé ces derniers temps.Anxiété accrue signalée par le patient.hybrid
sess003therapist0.95truepsychiatrytrans0062023-10-05T15:45:00Z2023-10-05T16:30:00Zpatx29yj4provp67q2nfr2providerprovp67q2n6.210Pouvez-vous me dire ce qui provoque ce stress?Anxiété accrue signalée par le patient.hybrid
sess004physician0.98falsecardiologytrans0072023-10-07T10:15:00Zblankpatq93r2pprovw97h2sen1providerprovw97h2s02.3Your test results look normal.Review of ECG and lipid panel.manual
sess004physician0.96falsecardiologytrans0082023-10-07T10:15:00Zblankpatq93r2pprovw97h2sen2patientpatq93r2p2.36.1That's a relief, thank you doctor.Review of ECG and lipid panel.manual

What the 200 rows show

from the 200-row sample

Therapist (provider role) stands out: 40 of its 40 rows have is_sensitive = true, against 7 of 160 for the rest.

  • 24%is_sensitive = true
  • 0.92median utterance_confidence
  • 2speaker types
  • 3transcript sources
  • 10languages
  • 0.0median utterance_start_time
Is sensitive rate by provider_roleis_sensitive = true
0%50%100%11%physician7 of 640%nurse0 of 62100%therapist40 of 400%other0 of 34
utterance_confidence200 rows, in bands of 0.1
0651304000000106312300.51utterance_confidence →

Median 0.92, from 0.0 to 1.0.

medical_specialty200 rows · top 10 of 16 values
  1. emergency26
  2. psychology21
  3. psychiatry19
  4. internal medicine17
  5. geriatrics16
  6. pediatrics14
  7. radiology12
  8. immunology12
  9. endocrinology11
  10. cardiology11
19 columns by typefrom the column list below
  • string 12
  • integer 1
  • float 3
  • datetime 2
  • boolean 1

Columns

19 columns in four groups
blueprint · 19 columns
columntypedescriptionexample
Text 12 columns
session_idstringUnique identifier for the telemedicine sessionuniquesess001
transcript_idstringUnique identifier for the transcript entry (utterance or segment)uniquetrans001
patient_idstringAnonymized unique identifier for the patientpat8a1e9d
provider_idstringAnonymized unique identifier for the healthcare providerprovb21k7m
provider_rolestringRole of the healthcare provider (e.g., physician, nurse, therapist)physician · nurse · therapist · other · optionalphysician
languagestringPrimary language used in the session (ISO 639-1 code, e.g., 'en', 'es')10 languages · optionalen
speaker_typestringIndicates whether the speaker is the provider or the patientprovider · patientprovider
speaker_idstringAnonymized unique identifier for the speaker (matches patient_id or provider_id)provb21k7m
utterance_textstringVerbatim transcript text of the utterance or segment你今天感觉怎么样？
medical_specialtystringMedical specialty relevant to the session (e.g., cardiology, dermatology)optionalfamily medicine
session_notesstringOptional notes or summary about the session (e.g., chief complaint, diagnosis)optionalSystem check.
transcript_sourcestringSource of the transcript (e.g., manual, ASR, hybrid)manual · asr · hybrid · optionalmanual
Numbers 4 columns
utterance_orderintegerSequence number of the utterance within the session1 or more1
utterance_start_timefloatStart time of the utterance in seconds from the beginning of the session audio0 or more · optional0
utterance_end_timefloatEnd time of the utterance in seconds from the beginning of the session audio0 or more · optional4.5
utterance_confidencefloatConfidence score (0-1) for the transcript accuracy of this utterance0 to 1 · optional0.97
Dates and times 2 columns
session_start_datetimedatetimeDate and time when the telemedicine session started2023-10-01T09:05:00Z
session_end_datetimedatetimeDate and time when the telemedicine session endedoptional2023-10-01T09:38:00Z
True or false 1 column
is_sensitivebooleanIndicates if the utterance contains sensitive information (e.g., PII, PHI)optionalfalse

Use it for

  • is sensitive24%47 of 200 rowsmean utterance confid…0.92phys…0.87nurse0.92ther…0.82other

    A healthcare dashboard

    The is_sensitive rate, utterance_confidence by provider_role and a breakdown of medical_specialty. Excel, Power BI or Tableau.

  • Why do 47 of 200 rows have is_sensitive = true?

    A root-cause class exercise

    Hand out the rows and one question. The answer is in the data, not in the brief.

  • A software demo

    Believable sessions with transcript_id, session_start_datetime and session_end_datetime to fill a screen in front of a buyer.

Not quite right?

Make it yours.

Same 19 columns, your size and your rules. See 20 rows before you pay.

Preview 20 rows free

1,000 rows: $1.99 · 10,000: $12.99One-time. No subscription. All prices

This dataset200 rows19 columns
Yours1,000 rows19 columns

blueprint · telemedicine-session-audio-transcripts

Behind this dataset

Same schema. As many rows as you need.

These 200 rows came out of a blueprint — 19 columns with generation rules behind each one. Open it in Data Factory to retune a column, add your own, wire in foreign keys, and run it at the size you actually need.

Rules it was built with
  • JSON record for each call session.
  • Speaker separation (doctor/patient) marked.
  • Anonymized sensitive information.
  • Includes summary of consultation outcomes.
Rows
Open the blueprint in Data Factory

1 credit per row. Preview 20 rows free. Buy rows for the full run.

Exports
CSV, JSON, JSONL, Parquet, SQL, Excel, TSV, XML
Licence
yours to use, including commercially
API slug
telemedicine-session-audio-transcripts

More in Healthcare

Browse all
Also in HealthcareHealthcare Chatbot Intent DatasetPharmacy Inventory Movement DatasetPatient Medication Adherence LogsAll 1,966 datasets

What should your data show?

Preview 20 rows free
No signup. No card.