push backend

Backend/Dockerfile

+# 
+FROM python:3.8.16
+
+# 
+WORKDIR /code
+
+COPY . .
+
+RUN apt-get update && apt-get install -y libgl1-mesa-glx
+
+# 
+RUN pip install --no-cache-dir --upgrade -r newreq.txt
+
+EXPOSE 8000
+# 
+CMD ["uvicorn", "main:app", "--host", "0.0.0.0", "--port", "8000"]

Backend/main.py

+from fastapi import FastAPI, File, UploadFile
+from fastapi.middleware.cors import CORSMiddleware
+import pandas as pd
+import cv2
+import numpy as np
+import base64
+from pathlib import Path
+import os
+import pickle
+from pydantic import BaseModel
+import tensorflow as tf
+from tensorflow.keras.models import load_model
+
+
+class Component1Type(BaseModel):
+    base64Image: str
+
+
+class Component2Type(BaseModel):
+    menopause: int
+    agegrp: int
+    density: int
+    race: int
+    Hispanic: int
+    bmi: int
+    agefirst: int
+    nrelbc: int
+    brstproc: int
+    surgmeno: int
+    hrt: int
+
+
+class Component3Type(BaseModel):
+    ageAtDiagnosis: int
+    typeOfBreastSurgery: int
+    cancerType: int
+    cancerTypeDetailed: int
+    Cellularity: int
+    Chemotherapy: int
+    Pam50ClaudinLowSubtype: int
+    cohort: float
+    erStatusMeasuredByIHC: int
+    erStatus: int
+    neoplasmHistologicGrade: float
+    her2StatusMeasuredBySNP6: int
+    her2Status: int
+    tumorOtherHistologicSubtype: int
+    hormoneTherapy: int
+    inferredMenopausalState: int
+    integrativeCluster: int
+    primaryTumorLaterality: int
+    lymphNodesExaminedPositive: float
+    mutationCount: float
+    nottinghamPrognosticIndex: float
+    oncotreeCode: int
+    overallSurvivalMonths: int
+    overallSurvivalStatus: int
+    prStatus: int
+    radioTherapy: int
+    relapseFreeStatusMonths: int
+    relapseFreeStatus: int
+    numberOfSamplesPerPatient: int
+    sex: int
+    threeGeneClassifierSubtype: int
+    tmbNonsynonymous: float
+    tumorSize: float
+    tumorStage: float
+
+
+app = FastAPI()
+
+origins = [
+    "http://localhost",
+    "http://localhost:8080",
+    "http://localhost:3000",
+]
+
+app.add_middleware(
+    CORSMiddleware,
+    allow_origins=origins,
+    allow_credentials=True,
+    allow_methods=["*"],
+    allow_headers=["*"],
+)
+
+
+@app.get("/")
+def get_root():
+    return {"message": "Hello World"}
+
+
+model1_path = Path("assets/model1.h5")
+
+
+@app.post("/component/1")
+async def get_component1(file: UploadFile = File(...)):
+
+    with open(file.filename, "wb") as buffer:
+        buffer.write(await file.read())
+    
+    with open(file.filename, "rb") as image_file:
+
+        img_bytes = image_file.read()
+    
+        if os.path.isfile(model1_path):
+            new_model = load_model(model1_path)
+            img_array = np.frombuffer(img_bytes, np.uint8)
+            img = cv2.imdecode(img_array, cv2.IMREAD_COLOR)
+            resize = tf.image.resize(img, (256, 256))
+            yhat = new_model.predict(np.expand_dims(resize/255, 0))
+            type = ''
+            if yhat > 0.3:
+                type = 'Malignant'
+            else:
+                type = 'Benign'
+            return {"type": type}
+        else:
+            return ({"model_loaded": False, "error": "Model file not found"})
+
+
+model2_path = Path("assets/model2.pkl")
+
+
+@app.post("/component/2")
+def get_component2(data: Component2Type):
+    data_fields = {
+        "menopause": data.menopause,
+        "agegrp": data.agegrp,
+        "density": data.density,
+        "race": data.race,
+        "Hispanic": data.Hispanic,
+        "bmi": data.bmi,
+        "agefirst": data.agefirst,
+        "nrelbc": data.nrelbc,
+        "brstproc": data.brstproc,
+        "surgmeno": data.surgmeno,
+        "hrt": data.hrt
+    }
+    if os.path.isfile(model2_path):
+        with open(model2_path, 'rb') as f:
+            loaded_model = pickle.load(f)
+            dfToCheck = pd.DataFrame(data_fields, index=[0])
+            return {"probabilityOfSurvice": loaded_model.predict_proba(dfToCheck)[:, 1][0]}
+    else:
+        return ({"model_loaded": False, "error": "Model file not found"})
+
+
+model3_path = Path("assets/model3.pkl")
+
+
+@app.post("/component/3")
+def get_component3(data: Component3Type):
+    dict_to_predict = {
+        'Age at Diagnosis': data.ageAtDiagnosis,
+        'Type of Breast Surgery': data.typeOfBreastSurgery,
+        'Cancer Type': data.cancerType,
+        'Cancer Type Detailed': data.cancerTypeDetailed,
+        'Cellularity': data.Cellularity,
+        'Chemotherapy': data.Chemotherapy,
+        'Pam50 + Claudin-low subtype': data.Pam50ClaudinLowSubtype,
+        'Cohort': data.cohort,
+        'ER status measured by IHC': data.erStatusMeasuredByIHC,
+        'ER Status': data.erStatus,
+        'Neoplasm Histologic Grade': data.neoplasmHistologicGrade,
+        'HER2 status measured by SNP6': data.her2StatusMeasuredBySNP6,
+        'HER2 Status': data.her2Status,
+        'Tumor Other Histologic Subtype': data.tumorOtherHistologicSubtype,
+        'Hormone Therapy': data.hormoneTherapy,
+        'Inferred Menopausal State': data.inferredMenopausalState,
+        'Integrative Cluster': data.integrativeCluster,
+        'Primary Tumor Laterality': data.primaryTumorLaterality,
+        'Lymph nodes examined positive': data.lymphNodesExaminedPositive,
+        'Mutation Count': data.mutationCount,
+        'Nottingham prognostic index': data.nottinghamPrognosticIndex,
+        'Oncotree Code': data.oncotreeCode,
+        'Overall Survival (Months)': data.overallSurvivalMonths,
+        'Overall Survival Status': data.overallSurvivalStatus,
+        'PR Status': data.prStatus,
+        'Radio Therapy': data.radioTherapy,
+        'Relapse Free Status (Months)': data.relapseFreeStatusMonths,
+        'Relapse Free Status': data.relapseFreeStatus,
+        'Number of Samples Per Patient': data.numberOfSamplesPerPatient,
+        'Sex': data.sex,
+        '3-Gene classifier subtype': data.threeGeneClassifierSubtype,
+        'TMB (nonsynonymous)': data.tmbNonsynonymous,
+        'Tumor Size': data.tumorSize,
+        'Tumor Stage': data.tumorStage
+    }
+    if os.path.isfile(model3_path):
+        with open(model3_path, 'rb') as f:
+            loaded_model = pickle.load(f)
+            dfToCheck = pd.DataFrame(dict_to_predict, index=[0])
+            return {"prob": 1-loaded_model.predict_proba(dfToCheck)[:, 1][0]}
+    else:
+        return ({"model_loaded": False, "error": "Model file not found"})

Backend/readme.md

+## Run the Backend
+pip install --no-cache-dir --upgrade -r newreq.txt
+
+uvicorn main:app --reload
\ No newline at end of file