New Feature : Prescription Digitization

.gitignore

@@ -0,0 +1 @@
+configs

CONTRIBUTING.md

@@ -15,9 +15,14 @@ If you have a suggestion that would make this better, please fork the repo and c
 ## Setup - Front End
 - Change directory into frontend `cd Sahaay/frontEnd`
 - Install dependencies `npm i`
+- Copy your configuration files into configs folder
+    - It contains 2 files.
+    - Rename to `google-services.json` for Android.
+    - Rename to `GoogleService-Info.plist` for IOS.
+
 
 ### Development Build
-- For Android, download development build apk from this [link](https://expo.dev/artifacts/eas/46C15VEcskVvueo11M3TA5.apk).<div align='center'>OR</div>Build your own development apk using `npm run build-dev-apk` & download it from [Expo dev](https://expo.dev/).
+- For Android, download development build apk from this [link](https://expo.dev/artifacts/eas/f1XcZvGjKjD47iykSFr7xt.apk).<div align='center'>OR</div>Build your own development apk using `npm run build-dev-apk` & download it from [Expo dev](https://expo.dev/).
 - Install development build apk in your Android device.
 - Start metro bundler `npm start`
 - Start metro bundler with empty cache `npm run start-c`
@@ -30,16 +35,21 @@ If you have a suggestion that would make this better, please fork the repo and c
 
 ## Setup - Back End
 - Change directory into backend `cd Sahaay/backEnd`
+- Copy your configuration files into configs folder
+    - It contains 2 files.
+    - Rename to `gCloud.json` for Google's Vision API
+    - Rename to `serviceAccount.json` for rest applications
 - Start Server using
 ```powershell
-uvicorn --app-dir src server:app --reload --host [your-host-ip or 127.0.0.1]
+# If all packages are already installed in your system
+uvicorn --app-dir src server:app --reload --host [your-host-ip]
 ```
 
 <p align='center'>or</p> 
 
 ```powershell
 pipenv install
-pipenv run python src/server.py
+pipenv run uvicorn --app-dir src server:app --reload --host [your-host-ip]
 ```
 
 ### Deploy Server to [Platform.sh](https://platform.sh/)

README.md

@@ -27,12 +27,12 @@
 Sahaay is an application which Digitize, Analyze & Predict Medical Reports using Deep Learning. It provides a solution to digitize the handwritten prescriptions, lab reports and detect various diseases.
 
 ## Main Features
-- Prescription Digitization (in development)
+- Prescription Digitization
 - Brain Tumor Classification
 - Kidney Stone Classification
 - Skin Disease Classification
 - Calories Estimation
-- Digital Prescription (in development)
+- Digital Prescription
 - Search About Medicine
 
 ## App Features
@@ -42,7 +42,7 @@ Sahaay is an application which Digitize, Analyze & Predict Medical Reports using
 - Share your profile & reports with other users
 
 ## ⬇️ Download Android App
-- Download Sahaay app using this [link](https://expo.dev/artifacts/eas/qfrEemJyBC8d9EYU9bvyHX.apk).
+- Download Sahaay app using this [link](https://expo.dev/artifacts/eas/bPMpoC94ycbrSUm9U9LgfF.apk).
 - See screenshots of this app [here](./SCREENSHOTS.md).
 ## 🔧 Tools/Technologies Used
 

backEnd/.gitignore

@@ -104,4 +104,5 @@ venv.bak/
 # mypy
 .mypy_cache/
 .platform/local
-Pipfile.lock
+Pipfile.lock
+configs

backEnd/Pipfile

@@ -6,12 +6,14 @@ verify_ssl = true
 [dev-packages]
 
 [packages]
+google-cloud-vision = "*"
 fastapi = "*"
-uvicorn = "*"
 firebase-admin = "*"
+openai = "*"
 opencv-python = "*"
 scikit-image = "*"
 tensorflow = "*"
+uvicorn = "*"
 
 [requires]
-python_version = "3.11"
+python_version = "3.11"

backEnd/src/routers/brain_tumor_detection.py

@@ -0,0 +1,63 @@
+from fastapi import APIRouter
+from pydantic import BaseModel
+
+from util.predictions_analyzer import predictions_analyzer
+import util.myfirebase as myfirebase
+
+from keras.models import load_model
+from skimage import io
+import cv2
+
+router = APIRouter(prefix="/brain-tumor-detection", tags=["brain"])
+
+
+class Item(BaseModel):
+    uid: str
+    img_url: str
+
+
+global graph, sess
+
+model = load_model("src/models/brain_tumor.h5")
+
+img_size = 200
+classes = {0: "Glioma", 1: "Meningioma", 2: "Tumor Not Detected", 3: "Pituitary"}
+
+
+def predict_class(image):
+    img = io.imread(image)
+    img = cv2.resize(img, (img_size, img_size)) / 255.0
+
+    predictions = (model.predict(img.reshape(1, img_size, img_size, 3)) * 100.0).round(
+        2
+    )
+
+    return predictions_analyzer(predictions=predictions, classes=classes)
+
+
+@router.post("/")
+async def main(item: Item):
+    try:
+        uid = item.uid
+        img_url = item.img_url
+
+        label, accuracy = predict_class(img_url)
+
+        doc_id = myfirebase.saveReport(
+            uid,
+            {
+                "title": "Brain Tumor Detection",
+                "result": label,
+                "img_url": img_url,
+                "isDetected": True
+                if label != "Tumor Not Detected"
+                else False,
+                "accuracy": accuracy,
+            },
+        )
+
+        return {"data": {"id": doc_id}}
+
+    except Exception as e:
+        print(e)
+        return {"error": {"message": e.__str__()}}

backEnd/src/routers/calorieEstimation.py → backEnd/src/routers/calorie_estimation.py

@@ -63,6 +63,7 @@ def main(item: Item):
 
         nut = json["results"][0]["items"][0]["nutrition"]
 
+        # convert SI Unit to human readable units & concat unit
         doc_id = myfirebase.saveReport(
             uid,
             {

backEnd/src/routers/kidney_stone_detection.py

@@ -0,0 +1,58 @@
+from fastapi import APIRouter
+from pydantic import BaseModel
+
+from util.predictions_analyzer import predictions_analyzer
+import util.myfirebase as myfirebase
+
+from keras.models import load_model
+from skimage import io
+import cv2
+
+router = APIRouter(prefix="/kidney-stone-detection", tags=["kidney"])
+
+
+class Item(BaseModel):
+    uid: str
+    img_url: str
+
+
+model = load_model("src/models/kidney_stone.h5")
+img_size = 200
+classes = {0: "Cyst", 1: "Normal", 2: "Stone", 3: "Tumor"}
+
+
+def predict_class(image):
+    img = io.imread(image)
+    img = cv2.resize(img, (img_size, img_size)) / 255.0
+
+    predictions = (model.predict(img.reshape(1, img_size, img_size, 3)) * 100.0).round(
+        2
+    )
+
+    return predictions_analyzer(predictions=predictions, classes=classes)
+
+
+@router.post("/")
+async def main(item: Item):
+    try:
+        uid = item.uid
+        img_url = item.img_url
+
+        label, accuracy = predict_class(img_url)
+
+        doc_id = myfirebase.saveReport(
+            uid,
+            {
+                "title": "Kidney Stone Detection",
+                "result": label,
+                "img_url": img_url,
+                "isDetected": True if label != "Normal" else False,
+                "accuracy": accuracy,
+            },
+        )
+
+        return {"data": {"id": doc_id}}
+
+    except Exception as e:
+        print(e)
+        return {"error": {"message": e.__str__()}}