File size: 13,943 Bytes
cf070a2
4b899e5
 
 
 
fd6e070
4b899e5
 
 
08d974f
cf070a2
d3a44ea
cf070a2
 
 
d3a44ea
cf070a2
4b899e5
 
 
 
 
 
cf070a2
4b899e5
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
d3a44ea
cf070a2
 
4b899e5
d3a44ea
4b899e5
d3a44ea
4b899e5
cf070a2
4b899e5
08d974f
d3a44ea
4b899e5
cf070a2
 
 
 
 
 
 
 
 
 
 
 
 
 
 
4b899e5
 
 
 
 
 
 
 
 
 
 
fd6e070
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
4b899e5
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
cf070a2
 
 
d3a44ea
cf070a2
 
 
 
 
 
 
 
 
 
d3a44ea
 
cf070a2
 
 
d3a44ea
cf070a2
 
 
 
4de4e3b
d3a44ea
cf070a2
d3a44ea
cf070a2
 
 
 
 
 
 
 
 
 
fd6e070
cf070a2
 
fd6e070
 
 
 
cf070a2
 
 
 
fd6e070
 
 
 
 
cf070a2
fd6e070
 
cf070a2
4b899e5
 
 
 
 
 
 
 
 
 
 
fd6e070
4b899e5
 
 
c9a3e86
 
 
 
 
 
fd6e070
 
 
 
c9a3e86
 
 
 
4b899e5
 
fd6e070
4b899e5
 
fd6e070
c9a3e86
4b899e5
 
 
 
cf070a2
 
 
c9a3e86
4de4e3b
cf070a2
 
4de4e3b
c9a3e86
fd6e070
 
 
 
cf070a2
 
fd6e070
 
4de4e3b
fd6e070
 
4de4e3b
cf070a2
fd6e070
 
 
 
cf070a2
 
 
 
 
 
 
4de4e3b
4b899e5
 
 
 
 
 
 
 
 
 
fd6e070
 
8250165
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
fd6e070
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
import os
from ast import List
from fastapi.middleware.cors import CORSMiddleware
from pydantic import BaseModel
import io
#import fitz
import traceback
import pandas as pd

import base64
import json
import re
import asyncio
import functools
from typing import Any, Optional

import google.generativeai as genai
from fastapi import FastAPI, UploadFile, File, Form, HTTPException, APIRouter, Request
from fastapi.responses import JSONResponse, RedirectResponse
from fastapi.staticfiles import StaticFiles
import firebase_admin
from firebase_admin import credentials, firestore
from google.generativeai import generative_models
from pydantic import BaseModel
from past_reports import router as reports_router, db_fetch_reports

from api_key import GEMINI_API_KEY 

app = FastAPI()
api = APIRouter(prefix="/api")
app.include_router(api)

EXTRACTED_TEXT_CACHE = ""

app.mount("/app", StaticFiles(directory="web", html=True), name="web")
app.include_router(reports_router)

app.add_middleware(
    CORSMiddleware,
    allow_origins=["*"],
    allow_credentials=True,
    allow_methods=["*"],
    allow_headers=["*"],
)

@app.get("/")
def root():
    return RedirectResponse(url="/app/")

class AnalyzeRequest(BaseModel):
    image_base64: str
    prompt: Optional[str] = None

GEMINI_API_KEY = os.getenv("GEMINI_API_KEY", GEMINI_API_KEY)

if not GEMINI_API_KEY:
    raise RuntimeError(
        "No Gemini API key found. Put it in api_key.py as `GEMINI_API_KEY = '...'` or set env var GEMINI_API_KEY."
    )

genai.configure(api_key=GEMINI_API_KEY)

generation_config = {
    "temperature": 0.2,    
    "top_p": 0.95,
    "top_k": 40,
    "max_output_tokens": 2048,
}

safety_settings = [
    {"category": "HARM_CATEGORY_HARASSMENT", "threshold": "BLOCK_MEDIUM_AND_ABOVE"},
    {"category": "HARM_CATEGORY_HATE_SPEECH", "threshold": "BLOCK_MEDIUM_AND_ABOVE"},
    {"category": "HARM_CATEGORY_SEXUALLY_EXPLICIT", "threshold": "BLOCK_MEDIUM_AND_ABOVE"},
    {"category": "HARM_CATEGORY_DANGEROUS_CONTENT", "threshold": "BLOCK_MEDIUM_AND_ABOVE"},
]

# --- Pydantic Models for API Endpoints ---
class ChatRequest(BaseModel):
    user_id: Optional[str] = "anonymous"
    question: str
    
class ChatResponse(BaseModel):
    answer: str
    
class TextRequest(BaseModel):
    text: str

system_prompt = """ You are a highly skilled medical practitioner specializing in medical image and document analysis. You will be given either a medical image or a PDF. 

Your responsibilities are:

1. **Extract Text**: If the input is a PDF or image, first extract all the text content (lab values, notes, measurements, etc.). Do not summarize — keep the extracted text verbatim.  

2. **Detailed Analysis**: Use both the extracted text and the visual features of the image to identify any anomalies, diseases, or health issues.  

3. **Finding Report**: Document all observed anomalies or signs of disease.  
   - Include any measurements (e.g., triglycerides, HBa1c, HDL) in the format:  
     `{"findings": "Condition only if risky: measurement type -- value with unit(current range)"}`  
   - Simplify the finding in **3 words** at the beginning when helpful.  

4. **Checking for Past**: If a disease is family history or previously recovered, mark severity as:  
   `"severity": "severity of anomaly (Past Anomaly but Still Under Risk)"`

5. **Recommendations and Next Steps**: Provide detailed recommendations (tests, follow-ups, consultations).  

6. **Treatment Suggestions**: Offer preliminary treatments or interventions.  

7. **Output Format**: Always return a JSON object containing both the raw extracted text and the structured analysis, like this:

```json
{
  "ocr_text": "<<<FULL VERBATIM TEXT FROM THE PDF/IMAGE>>>",
  "analysis": [
    {
      "findings": "UPPERCASE MAIN CONCERN. Description of the first disease or condition.",
      "severity": "MILD/SEVERE/CRITICAL",
      "recommendations": ["Follow-up test 1", "Follow-up test 2"],
      "treatment_suggestions": ["Treatment 1", "Treatment 2"],
      "home_care_guidance": ["Care tip 1", "Care tip 2"]
    },
    {
      "findings": "UPPERCASE MAIN CONCERN. Description of the second disease or condition.",
      "severity": "MILD/SEVERE/CRITICAL",
      "recommendations": ["Follow-up test A", "Follow-up test B"],
      "treatment_suggestions": ["Treatment A", "Treatment B"],
      "home_care_guidance": ["Care tip A", "Care tip B"]
    }
  ]
}

Important Notes: 
    1. Scope of Response: Only respond if the image pertains to a human health issue. 
    2. Clarity of Image: Ensure the image is clear and suitable for accurate analysis. 
    3. As an AI, don't produce any output than doesn't follow this format and also remember that we need your input on the medical report for a team of renowned doctors to recheck your readings and determine AI's role in the medical industry. 
    4. Your Insights are Invaluable: Your insights play a crucial role in guiding clinical decisions. Please proceed with your analysis, adhering to the structured approach outlined above.
    5. Completely UPPERCASE the main concern in the finding """

system_prompt_chat = """
*** Role: Medical Guidance Facilitator
*** Objective:
Analyze medical data, provide concise, evidence-based insights, and recommend actionable next steps for patient care. This includes suggesting local physicians or specialists within a user-specified mile radius, prioritizing in-network options when insurance information is available, and maintaining strict safety compliance with appropriate disclaimers.
*** Capabilities:
1. Report Analysis – Review and interpret findings in uploaded medical reports.
2. Historical Context – Compare current findings with any available previous reports.
3. Medical Q&A – Answer specific questions about the report using trusted medical sources.
4. Specialist Matching – Recommend relevant physician specialties for identified conditions.
5. Local Physician Recommendations – List at least two real physician or clinic options within the user-specified mile radius (include name, specialty, address, distance from user, and contact info) based on the patient’s location and clinical need.
6. Insurance Guidance – If insurance/network information is provided, prioritize in-network physicians.
7. Safety Protocols – Include a brief disclaimer encouraging users to verify information, confirm insurance coverage, and consult providers directly.
*** Response Structure:
Start with a direct answer to the user’s primary question (maximum 4 concise sentences, each on a new line).
If a physician/specialist is needed, recommend at least two local providers within the requested radius (include name, specialty, address, distance, and contact info).
If insurance details are available, indicate which physicians are in-network.
End with a short safety disclaimer.
***Input Fields:
Provided Document Text: {document_text}
User Question: {user_question}
Assistant Answer:
"""

# Initialize model
model = genai.GenerativeModel(model_name="gemini-2.5-flash-lite")

async def _call_model_blocking(request_inputs, generation_cfg, safety_cfg):
    """Run blocking model call in threadpool (so uvicorn's event loop isn't blocked)."""
    fn = functools.partial(
        model.generate_content,
        request_inputs,
        generation_config=generation_cfg,
        safety_settings=safety_cfg,
    )
    loop = asyncio.get_event_loop()
    return await loop.run_in_executor(None, fn)


async def analyze_image(image_bytes: bytes, mime_type: str, prompt: Optional[str] = None) -> Any:
    base64_img = base64.b64encode(image_bytes).decode("utf-8")
    text_prompt = (prompt or system_prompt).strip()

    request_inputs = [
        {"inline_data": {"mime_type": mime_type, "data": base64_img}},
        {"text": text_prompt},
    ]

    try:
        response = await _call_model_blocking(request_inputs, generation_config, safety_settings)
    except Exception as e:
        raise RuntimeError(f"Model call failed: {e}")
    text = getattr(response, "text", None)
    if not text and isinstance(response, dict):
        candidates = response.get("candidates") or []
        if candidates:
            text = candidates[0].get("content") or candidates[0].get("text")
    if not text:
        text = str(response)

    clean = re.sub(r"```(?:json)?", "", text).strip()
    print(f"Cleaned text: {clean}")
    try:
        parsed = json.loads(clean)
        ocr_text = parsed["ocr_text"]
        analysis = parsed["analysis"]
        print(f"Parsed JSON: {parsed}")
        return analysis,ocr_text
    except json.JSONDecodeError:
        match = re.search(r"(\[.*\]|\{.*\})", clean, re.DOTALL)
        if match:
            try:
                 
                parsed =  json.loads(match.group(1)), None
                ocr_text = parsed["ocr_text"]
                analysis = parsed["analysis"]
                return analysis, ocr_text
            except json.JSONDecodeError:
                return {"raw_found_json": match.group(1)}, None
        return {"raw_output": clean}, None

def get_past_reports_from_sqllite(user_id: str):
    try:
        reports = db_fetch_reports(user_id=user_id, limit=10, offset=0)
        
        history_text = ""
        for report in reports:
            history_text += f"Report from {report.get('report_date', 'N/A')}:\n{report.get('ocr_text', 'No OCR text found')}\n\n"
    except Exception as e:
        history_text = "No past reports found for this user."
    return history_text


@app.post("/chat/", response_model=ChatResponse)
async def chat_endpoint(request: ChatRequest):
    global result
    print(f"Received chat request for user: {request.user_id}")
    """
    Chatbot endpoint that answers questions based on the last analyzed document and user history.
    """
   
    #history_text = get_past_reports_from_firestore(request.user_id)
    full_document_text = get_past_reports_from_sqllite(request.user_id.strip())
    
    full_document_text =  EXTRACTED_TEXT_CACHE+"\n\n" + "PAST REPORTS:\n" + full_document_text
    print(f"Full document text: {full_document_text}")
    if not full_document_text:
        raise HTTPException(status_code=400, detail="No past reports or current data exists for this user")
    
    
    try:
        full_prompt = system_prompt_chat.format(
            document_text=full_document_text,
            user_question=request.question
        )
        print(f"Full prompt: {full_prompt}")
       
        response = model.generate_content(full_prompt)
        return ChatResponse(answer=response.text)
    except Exception as e:
        raise HTTPException(status_code=500, detail=f"Chat error: {e}")

@app.post("/analyze")
async def analyze_endpoint(file: UploadFile = File(...), prompt: str = Form(None)):
    
    """
    Upload an image file (field name `file`) and optional text `prompt`.
    Returns parsed JSON (or raw model output if JSON couldn't be parsed).
    """

    global result,EXTRACTED_TEXT_CACHE
    
    filename = file.filename.lower()
    print(f"Received analyze request for file {filename}")
    contents = await file.read()  # <-- this gets the uploaded file bytes
    mime = file.content_type or "image/png"
    
    #result = await analyze_image(contents, mime, prompt)
    try:
        result, ocr_text = await analyze_image(contents, mime, prompt)
        EXTRACTED_TEXT_CACHE = ocr_text
    except Exception as e:
        raise HTTPException(status_code=500, detail=str(e))
    return JSONResponse(content={
        "ocr_text": ocr_text,
        "Detected_Anomolies": result
    })

@app.post("/analyze_json")
async def analyze_json(req: AnalyzeRequest):
    import base64
    image_bytes = base64.b64decode(req.image_base64)
    result = await analyze_image(image_bytes, "image/png", req.prompt)
    return {"result": result}
   
@app.get("/health/")
def health():
    return {"response": "ok"}

@app.on_event("startup")
def _log_routes():
    from fastapi.routing import APIRoute
    print("Mounted routes:")
    for r in app.routes:
        if isinstance(r, APIRoute):
            print(" ", r.path, r.methods)

@app.get("/UserResults/")
def get_user_results(user_id: str):
    cursor = conn.cursor()
    cursor.execute(
        "SELECT ocr_text, anomalies, created_at FROM reports WHERE user_id = ? ORDER BY created_at ASC",
        (user_id,)
    )
    rows = cursor.fetchall()

    # On-the-fly "latest anomaly per measurement"
    latest_anomalies = {}  # measurement -> anomaly dict
    combined_reports = []  # optional: keep full report data if needed

    for ocr_text, anomalies_json, created_at in rows:
        anomalies = json.loads(anomalies_json or "[]")
        for a in anomalies:
            meas = a.get("measurement")
            # Always keep the latest anomaly (rows are ordered oldest → newest)
            latest_anomalies[meas] = a
        combined_reports.append({
            "ocr_text": ocr_text,
            "created_at": created_at,
            "anomalies": anomalies
        })

    return {
        "user_id": user_id,
        "latest_anomalies": list(latest_anomalies.values()),
        "all_reports": combined_reports
    }


def main():
    """Run the application."""
    try:
        logger.info(f"Starting server on 8000")
        logger.info(f"Debug mode: true")

        if Config.DEBUG:
            # Use import string for reload mode
            uvicorn.run(
                "main:app",
                host="localhost",
                port="8000",
                reload=True,
                log_level="debug"
            )
        else:
            # Use app instance for production
            uvicorn.run(
                app,
                host="localhost",
                port="8000",
                reload=False,
                log_level="info"
            )

    except Exception as e:
        logger.error(f"Failed to start server: {e}")
        raise


if __name__ == "__main__":
    main()