Act as a professional optha doctor and tell me proper prescription after listening the problem, proper format
In ward sundari majhi, she is admitted in female surgical ward, her right eye is swelled,she didn't even open the ,the upper eyelid also swelled , she tell me suddenly this happened, no trauma nothing, write proper prescription
Does Sundari Majhi have fever or raised temperature? · Is there redness/erythema over the swollen eyelid and surrounding area? · Can she perceive light or count fingers in the right eye (even with difficulty opening)?
Is there any proptosis (forward bulging of the eyeball)? · Does she have pain on eye movement (even passive)? · What is her approximate age group?
| # | Investigation | Purpose |
|---|---|---|
| 1 | CT Orbit + Paranasal Sinuses with contrast (URGENT) | Confirm orbital vs preseptal, rule out subperiosteal/orbital abscess |
| 2 | CBC with differential | Check WBC, neutrophilia |
| 3 | Blood culture x2 (before antibiotics if possible) | Identify causative organism |
| 4 | ESR, CRP | Infection markers |
| 5 | Random blood sugar / HbA1c | Rule out diabetic/immunocompromised state |
| 6 | Blood urea, serum creatinine | Baseline before IV antibiotics |
| 7 | ENT consultation | Rule out sinusitis (most common source) |
| Drug | Dose | Route | Frequency | Duration |
|---|---|---|---|---|
| Inj. Ceftriaxone | 2g | IV | BD (every 12 hrs) | 7-14 days (review) |
| Inj. Metronidazole | 500mg | IV | TDS (every 8 hrs) | 7-14 days (review) |
De-escalate to oral antibiotics after 48-72 hrs clinical improvement, guided by culture sensitivity.
| Drug | Dose | Route | Frequency | Note |
|---|---|---|---|---|
| Inj. Dexamethasone | 0.1 mg/kg | IV | OD (morning) | Start after 24-48 hrs of antibiotics, to reduce orbital edema |
| Tab. Paracetamol | 500mg | Oral | TDS (if fever/pain) | Symptomatic |
| Inj. Pantoprazole | 40mg | IV | OD (before food) | GI protection with steroids |
| Treatment | Detail |
|---|---|
| Eye moist pad / saline-soaked gauze | Cover right eye gently - do NOT apply pressure |
| Lubricating eye drops (Carboxymethylcellulose 0.5%) | RE, 2 hourly - prevent exposure keratopathy |
| No eye patching with pressure | Risk of increased orbital pressure |
| Parameter | Frequency |
|---|---|
| Vitals (Temp, Pulse, BP, SpO2) | Every 4 hours |
| Visual acuity / light perception check (RE) | Every 6 hours - URGENT if worsens |
| Pupil reaction (RE vs LE) | Every 6 hours |
| Proptosis measurement (Hertel if available) | Daily |
| Pain on eye movement (scale 0-10) | Every 6 hours |
IMPORTANT NOTE: This is a sight-threatening emergency. IV antibiotics must be started immediately. CT orbit must be done on priority. Senior ophthalmologist must be informed urgently. Do NOT delay treatment awaiting investigation results if clinical picture is clear.
Write proper prescription format like C/O, H/O, Provisional diagnosis etc
| History | Details |
|---|---|
| Similar complaints in past | Nil |
| Diabetes Mellitus | Not known (to be investigated) |
| Hypertension | Not known |
| Sinusitis / Rhinitis | Not elicited |
| Previous eye surgery / injury | Nil |
| Tuberculosis | Not known |
| Any immunocompromised state | Not known |
| Parameter | Finding |
|---|---|
| General condition | Moderately ill-looking |
| Consciousness | Conscious, oriented |
| Temperature | Low-grade fever (to be recorded) |
| Pulse | ___ bpm |
| BP | ___/___ mmHg |
| SPO2 | ___% |
| Pallor / Icterus / Cyanosis | Not present |
| Lymphadenopathy | Not noted |
| Parameter | Finding |
|---|---|
| Periorbital region | Diffuse swelling of upper and lower lids, more upper lid |
| Skin over lids | Mildly erythematous, warm to touch |
| Lid opening | Unable to open (Complete ptosis / mechanical) |
| Proptosis | Present - right eye appears pushed forward |
| Conjunctiva | Chemosis likely (unable to assess fully) |
| Cornea | Unable to assess (eye closed) |
| Pupil | Unable to assess |
| Eye movements | Restricted, pain on movement |
| Vision | Light Perception Only (OD) |
| IOP | Deferred (eye closed) |
| Parameter | Finding |
|---|---|
| All findings | Within normal limits |
| Vision | 6/6 (to be confirmed) |
Right Eye (OD) - Orbital Cellulitis (Post-septal Cellulitis)
| # | Investigation | Priority |
|---|---|---|
| 1 | CT Orbit + Paranasal Sinuses (with contrast) | URGENT |
| 2 | CBC with differential count | URGENT |
| 3 | Blood culture x2 (before antibiotics) | URGENT |
| 4 | ESR, CRP | Same day |
| 5 | Random blood sugar, HbA1c | Same day |
| 6 | Blood urea, serum creatinine | Same day |
| 7 | LFT | Same day |
| 8 | Urine routine & microscopy | Same day |
| Drug | Dose | Route | Frequency | Duration |
|---|---|---|---|---|
| Inj. Ceftriaxone | 2g | IV slow (30 min) | BD | 7-14 days |
| Inj. Metronidazole | 500mg | IV | TDS | 7-14 days |
| Drug | Dose | Route | Frequency | Remarks |
|---|---|---|---|---|
| Inj. Dexamethasone | 0.1 mg/kg | IV | OD (morning) | Start after 24-48 hrs of antibiotics |
| Inj. Pantoprazole | 40 mg | IV | OD (before food) | GI protection |
| Tab. Paracetamol | 500 mg | Oral | TDS (PRN) | Fever/pain |
| IV fluids | NS/RL | IV | As per fluid chart | Maintain hydration |
| Treatment | Instruction |
|---|---|
| Carboxymethylcellulose 0.5% eye drops | 2 hourly in Right Eye - prevent corneal dryness |
| Saline-soaked moist pad over right eye | Gentle cover only - NO pressure dressing |
| DO NOT patch or apply pressure | Risk of worsening orbital pressure |
| Parameter | Frequency |
|---|---|
| Vitals (Temp / Pulse / BP / SPO2) | Every 4 hourly |
| Visual acuity / Light perception (Right Eye) | Every 6 hourly |
| Pupil reaction & RAPD check | Every 6 hourly |
| Proptosis assessment | Daily |
| Eye movement restriction | Daily |
| Fever chart | Continuous |
| Specialty | Reason |
|---|---|
| ENT (Urgent) | Rule out sinusitis as primary source |
| Medicine | Systemic workup, diabetes, immunosuppression |
| Neurosurgery (standby) | If cavernous sinus thrombosis suspected |
Hear this and tell me what she is saying
ls "/home/daytona/workspace/attachments/e2c4130c-d5a0-4efc-860d-055f1d13a10d/" && file "/home/daytona/workspace/attachments/e2c4130c-d5a0-4efc-860d-055f1d13a10d/telegram_audio 4.ogg" && file "/home/daytona/workspace/attachments/e2c4130c-d5a0-4efc-860d-055f1d13a10d/telegram_audio 3.ogg"
which whisper || pip show openai-whisper 2>/dev/null | head -5; which ffmpeg; python3 -c "import whisper; print('whisper ok')" 2>/dev/null || echo "whisper not available"
pip install --user openai-whisper 2>&1 | tail -5
python3 -c " import whisper model = whisper.load_model('base') print('Model loaded') # Transcribe audio 3 result3 = model.transcribe('/home/daytona/workspace/attachments/e2c4130c-d5a0-4efc-860d-055f1d13a10d/telegram_audio 3.ogg') print('=== AUDIO 3 ===') print('Language:', result3.get('language')) print('Text:', result3['text']) # Transcribe audio 4 result4 = model.transcribe('/home/daytona/workspace/attachments/e2c4130c-d5a0-4efc-860d-055f1d13a10d/telegram_audio 4.ogg') print('=== AUDIO 4 ===') print('Language:', result4.get('language')) print('Text:', result4['text']) " 2>/dev/null
# Convert ogg to wav first and try tiny model ffmpeg -i "/home/daytona/workspace/attachments/e2c4130c-d5a0-4efc-860d-055f1d13a10d/telegram_audio 3.ogg" /home/daytona/workspace/audio3.wav -y 2>/dev/null && echo "converted 3" && ffmpeg -i "/home/daytona/workspace/attachments/e2c4130c-d5a0-4efc-860d-055f1d13a10d/telegram_audio 4.ogg" /home/daytona/workspace/audio4.wav -y 2>/dev/null && echo "converted 4" && ls -lh /home/daytona/workspace/audio*.wav
ffmpeg -i "/home/daytona/workspace/attachments/e2c4130c-d5a0-4efc-860d-055f1d13a10d/telegram_audio 3.ogg" /home/daytona/workspace/audio3.wav -y 2>&1 | tail -10
# Use mp3 format instead ffmpeg -i "/home/daytona/workspace/attachments/e2c4130c-d5a0-4efc-860d-055f1d13a10d/telegram_audio 3.ogg" -ar 16000 -ac 1 /home/daytona/workspace/audio3.mp3 -y 2>&1 | tail -3 && ffmpeg -i "/home/daytona/workspace/attachments/e2c4130c-d5a0-4efc-860d-055f1d13a10d/telegram_audio 4.ogg" -ar 16000 -ac 1 /home/daytona/workspace/audio4.mp3 -y 2>&1 | tail -3 && ls -lh /home/daytona/workspace/audio*.mp3
python3 << 'EOF' import whisper import warnings warnings.filterwarnings("ignore") print("Loading tiny model...") model = whisper.load_model("tiny") print("Model ready. Transcribing audio 3...") r3 = model.transcribe("/home/daytona/workspace/audio3.mp3", fp16=False) print("=== AUDIO 3 ===") print("Language detected:", r3.get("language")) print("Text:", r3["text"]) EOF
# Check if faster-whisper is available, or try vosk python3 -c "import faster_whisper; print('faster_whisper ok')" 2>/dev/null || echo "not available" python3 -c "import vosk; print('vosk ok')" 2>/dev/null || echo "vosk not available" python3 -c "import speech_recognition; print('speech_recognition ok')" 2>/dev/null || echo "sr not available"
pip install --user faster-whisper 2>&1 | tail -3
pip install --user faster-whisper 2>&1 | grep -i "successfully\|error\|already"
timeout 100 python3 << 'EOF' from faster_whisper import WhisperModel import warnings warnings.filterwarnings("ignore") print("Loading tiny model via faster-whisper...") model = WhisperModel("tiny", device="cpu", compute_type="int8") print("Done. Transcribing audio 3...") segments, info = model.transcribe("/home/daytona/workspace/audio3.mp3", beam_size=1) print(f"Language: {info.language} (prob: {info.language_probability:.2f})") text3 = "" for seg in segments: text3 += seg.text print("AUDIO 3 TEXT:", text3) EOF
timeout 100 python3 << 'EOF' from faster_whisper import WhisperModel import warnings warnings.filterwarnings("ignore") model = WhisperModel("tiny", device="cpu", compute_type="int8") # Try with Bengali language hint print("Transcribing audio 3 with Bengali hint...") segments, info = model.transcribe("/home/daytona/workspace/audio3.mp3", beam_size=1, language="bn") text3 = "".join(seg.text for seg in segments) print("AUDIO 3 (bn):", text3) print("\nTranscribing audio 4 with Bengali hint...") segments, info = model.transcribe("/home/daytona/workspace/audio4.mp3", beam_size=1, language="bn") text4 = "".join(seg.text for seg in segments) print("AUDIO 4 (bn):", text4) EOF
# Try with Hindi and Odia language hints - patient is likely from Odisha (name: Sundari Majhi) timeout 100 python3 << 'EOF' from faster_whisper import WhisperModel import warnings warnings.filterwarnings("ignore") model = WhisperModel("tiny", device="cpu", compute_type="int8") for lang in ["hi", "or"]: print(f"\n--- Language: {lang} ---") segments, info = model.transcribe("/home/daytona/workspace/audio3.mp3", beam_size=1, language=lang) text = "".join(seg.text for seg in segments) print("AUDIO 3:", text[:300]) segments, info = model.transcribe("/home/daytona/workspace/audio4.mp3", beam_size=1, language=lang) text = "".join(seg.text for seg in segments) print("AUDIO 4:", text[:300]) EOF
# Try auto-detection with small model for better accuracy timeout 100 python3 << 'EOF' from faster_whisper import WhisperModel import warnings warnings.filterwarnings("ignore") model = WhisperModel("tiny", device="cpu", compute_type="int8") # Auto detect language print("=== AUDIO 3 - Auto detect ===") segments, info = model.transcribe("/home/daytona/workspace/audio3.mp3", beam_size=5) print(f"Detected language: {info.language} (prob: {info.language_probability:.2f})") for seg in segments: print(f"[{seg.start:.1f}s - {seg.end:.1f}s]: {seg.text}") print("\n=== AUDIO 4 - Auto detect ===") segments, info = model.transcribe("/home/daytona/workspace/audio4.mp3", beam_size=5) print(f"Detected language: {info.language} (prob: {info.language_probability:.2f})") for seg in segments: print(f"[{seg.start:.1f}s - {seg.end:.1f}s]: {seg.text}") EOF
tiny Whisper model has poor accuracy for Bengali/Odia - it needs the medium or large model for reliable regional Indian language transcriptionA new patient , see his condition and write prescription
| H/O similar complaints | Not known |
| H/O Diabetes / HTN | Not known |
| H/O Lumbar disc disease | Not known |
| H/O Trauma / Fall | Denied |
| Drug allergy | NKDA |
| Parameter | Finding |
|---|---|
| Site | Posterior leg, below knee (popliteal fossa / gastrocnemius area) |
| Tenderness | To be elicited on palpation |
| Swelling | To be noted |
| Muscle spasm | Likely present |
| Straight Leg Raise (SLR) test | To be performed |
| Knee range of motion | To be assessed |
| Ankle dorsiflexion | To be assessed (Simmonds / Thompson test if Achilles suspected) |
| Neurological exam | Power, reflexes, sensation in L4-L5-S1 distribution |
Hamstring / Gastrocnemius muscle strain with possible radiculopathy (L5-S1)
| # | Test | Purpose |
|---|---|---|
| 1 | X-Ray Knee (AP + Lateral) | Rule out bony pathology |
| 2 | Doppler Ultrasound Leg | Rule out DVT |
| 3 | MRI Lumbar Spine (if radiculopathy suspected) | Disc herniation / nerve compression |
| 4 | CBC, ESR, CRP | Rule out inflammatory/infective cause |
| 5 | Random Blood Sugar | Baseline |
| Drug | Dose | Route | Frequency | Duration |
|---|---|---|---|---|
| Tab. Diclofenac + Paracetamol (Voveran Plus / Combiflam) | 1 tab | Oral | BD after food | 5 days |
| Tab. Thiocolchicoside (Myoril 4mg) | 4 mg | Oral | BD | 5 days (muscle relaxant) |
| Tab. Pantoprazole 40mg | 1 tab | Oral | OD before breakfast | 5 days (GI protection) |
| Tab. Methylcobalamin 500mcg | 1 tab | Oral | OD | 15 days (nerve support) |
| Drug | Application | Frequency |
|---|---|---|
| Diclofenac Gel 1% (Voveran / Voltaren) | Apply over painful area gently | TDS, avoid vigorous massage |
| Ice pack (NOT hot fomentation for acute strain) | 15-20 min, 3-4x/day for first 48 hrs |
| Rest from aggravating activities | Strictly advised |
| Gentle hamstring / calf stretching | After 3-4 days when acute pain settles |
| TENS / Ultrasonic therapy | By physiotherapist |
Note: Stop hot fomentation and massage in acute phase - it can worsen swelling and inflammation.
Note: This is a musculoskeletal complaint, not an ophthalmic case. It has been assessed and managed as a general/orthopedic consultation based on the symptoms described. Please ensure the patient is seen by an Orthopedic or General Physician for proper physical examination.